krb5: Import MIT 1.21.3

Import MIT KRB5 1.21.3, which will gracefully replace KTH Heimdal.

The tarball used in this import is the same tarball used in
ports/krb5-121.

The Makefiles to add MIT KRB5 to the build will be added later.

Obtained from:	http://web.mit.edu/kerberos/dist/

Sponsored by:	The FreeBSD Foundation

Add 'crypto/krb5/' from commit '8f7d3ef26dec89a92ec0665de84a5936310a5574'

git-subtree-dir: crypto/krb5
git-subtree-mainline: bf6873c578
git-subtree-split: 8f7d3ef26d
This commit is contained in:
Cy Schubert 2025-06-04 19:47:44 -07:00
commit 7f2fe78b9d
4459 changed files with 977490 additions and 0 deletions

104
crypto/krb5/.github/workflows/build.yml vendored Normal file
View file

@ -0,0 +1,104 @@
name: Build
on:
push: {paths: [src/**, .github/workflows/build.yml]}
pull_request: {paths: [src/**, .github/workflows/build.yml]}
jobs:
unix:
runs-on: ${{ matrix.os }}
strategy:
fail-fast: false
matrix:
name: [linux-clang, linux-clang-openssl, linux-gcc]
include:
- name: linux-clang
os: ubuntu-latest
compiler: clang
makevars: CPPFLAGS=-Werror
configureopts: --enable-asan
- name: linux-clang-openssl
os: ubuntu-latest
compiler: clang
makevars: CPPFLAGS=-Werror
configureopts: --with-crypto-impl=openssl
- name: linux-gcc
os: ubuntu-latest
compiler: gcc
steps:
- name: Checkout repository
uses: actions/checkout@v1
- name: Linux setup
if: startsWith(matrix.os, 'ubuntu')
run: |
sudo apt-get update -qq
sudo apt-get install -y bison gettext keyutils ldap-utils libcmocka-dev libldap2-dev libkeyutils-dev libsasl2-dev libssl-dev python3-kdcproxy python3-pip slapd tcsh
pip3 install pyrad
- name: Build
env:
CC: ${{ matrix.compiler }}
MAKEVARS: ${{ matrix.makevars }}
CONFIGURE_OPTS: ${{ matrix.configureopts }}
run: |
cd src
autoreconf
./configure --enable-maintainer-mode --with-ldap $CONFIGURE_OPTS --prefix=$HOME/inst
make $MAKEVARS
make check
make install
- name: Display skipped tests
run: cat src/skiptests
- name: Check for files unexpectedly not removed by make distclean
run: |
cd src
make distclean
rm -rf autom4te.cache configure include/autoconf.h.in
if [ -n "$(git ls-files -o)" ]; then
echo "Files not removed by make distclean:"
git ls-files -o
exit 1
fi
windows:
runs-on: windows-latest
env:
KRB_INSTALL_DIR: C:\kfw
steps:
- name: Checkout repository
uses: actions/checkout@v1
- name: Setup
shell: cmd
run: |
mkdir %KRB_INSTALL_DIR%
- uses: ilammy/msvc-dev-cmd@v1
with:
arch: x86
- name: Build 32-bit
shell: cmd
run: |
cd src
set
set PATH=%PATH%;%wix%bin
nmake -f Makefile.in prep-windows
nmake
nmake install
cd windows\installer\wix
nmake
rename kfw.msi kfw32.msi
- uses: ilammy/msvc-dev-cmd@v1
with:
arch: x64
- name: Build 64-bit
shell: cmd
run: |
cd src
set
set PATH=%PATH%;%wix%bin;"%WindowsSdkVerBinPath%"\x86
nmake clean
nmake
nmake install
cd windows\installer\wix
nmake clean
nmake
rename kfw.msi kfw64.msi

41
crypto/krb5/.github/workflows/doc.yml vendored Normal file
View file

@ -0,0 +1,41 @@
name: Doc
on:
push: {paths: [doc/**, src/doc/*, src/include/krb5/krb5.hin, .github/workflows/doc.yml]}
pull_request: {paths: [doc/**, src/doc/*, src/include/krb5/krb5.hin, .github/workflows/doc.yml]}
jobs:
doc-older-sphinx:
runs-on: ubuntu-22.04
steps:
- name: Checkout repository
uses: actions/checkout@v1
- name: Linux setup
run: |
sudo apt-get update -qq
sudo apt-get install -y doxygen python3-lxml python3-pip python3-sphinx
pip3 install Cheetah3
- name: Build documentation
run: |
cd src/doc
make -f Makefile.in SPHINX_ARGS=-W htmlsrc
doc-newest-sphinx:
runs-on: ubuntu-latest
steps:
- name: Checkout repository
uses: actions/checkout@v1
- name: Linux setup
run: |
sudo apt-get update -qq
sudo apt-get install -y doxygen python3-lxml python3-pip
pip3 install Cheetah3 sphinx
- name: Build documentation
run: |
cd src/doc
make -f Makefile.in SPHINX_ARGS=-W htmlsrc
- name: Upload HTML
uses: actions/upload-artifact@v2
with:
name: html
path: doc/html
retention-days: 7

1384
crypto/krb5/NOTICE Normal file

File diff suppressed because it is too large Load diff

578
crypto/krb5/README Normal file
View file

@ -0,0 +1,578 @@
Kerberos Version 5, Release 1.21
Release Notes
The MIT Kerberos Team
Copyright and Other Notices
---------------------------
Copyright (C) 1985-2024 by the Massachusetts Institute of Technology
and its contributors. All rights reserved.
Please see the file named NOTICE for additional notices.
Documentation
-------------
Unified documentation for Kerberos V5 is available in both HTML and
PDF formats. The table of contents of the HTML format documentation
is at doc/html/index.html, and the PDF format documentation is in the
doc/pdf directory.
Additionally, you may find copies of the HTML format documentation
online at
https://web.mit.edu/kerberos/krb5-latest/doc/
for the most recent supported release, or at
https://web.mit.edu/kerberos/krb5-devel/doc/
for the release under development.
More information about Kerberos may be found at
https://web.mit.edu/kerberos/
and at the MIT Kerberos Consortium web site
https://kerberos.org/
Building and Installing Kerberos 5
----------------------------------
Build documentation is in doc/html/build/index.html or
doc/pdf/build.pdf.
The installation guide is in doc/html/admin/install.html or
doc/pdf/install.pdf.
If you are attempting to build under Windows, please see the
src/windows/README file.
Reporting Bugs
--------------
Please report any problems/bugs/comments by sending email to
krb5-bugs@mit.edu.
You may view bug reports by visiting
https://krbdev.mit.edu/rt/
and using the "Guest Login" button. Please note that the web
interface to our bug database is read-only for guests, and the primary
way to interact with our bug database is via email.
PAC transitions
---------------
Beginning with release 1.20, the KDC will include minimal PACs in
tickets instead of AD-SIGNEDPATH authdata. S4U requests (protocol
transition and constrained delegation) must now contain valid PACs in
the incoming tickets. Beginning with release 1.21, service ticket
PACs will contain a new KDC checksum buffer, to mitigate a hash
collision attack against the old KDC checksum. If only some KDCs in a
realm have been upgraded across versions 1.20 or 1.21, the upgraded
KDCs will reject S4U requests containing tickets from non-upgraded
KDCs and vice versa.
Triple-DES and RC4 transitions
------------------------------
Beginning with the krb5-1.21 release, the KDC will not issue tickets
with triple-DES or RC4 session keys unless explicitly configured using
the new allow_des3 and allow_rc4 variables in [libdefaults]. To
facilitate the negotiation of session keys, the KDC will assume that
all services can handle aes256-sha1 session keys unless the service
principal has a session_enctypes string attribute.
Beginning with the krb5-1.19 release, a warning will be issued if
initial credentials are acquired using the des3-cbc-sha1 encryption
type. Beginning with the krb5-1.21 release, a warning will also be
issued for the arcfour-hmac encryption type. In future releases,
these encryption types will be disabled by default and eventually
removed.
Beginning with the krb5-1.18 release, all support for single-DES
encryption types has been removed.
Major changes in 1.21.3 (2024-06-26)
------------------------------------
This is a bug fix release.
* Fix vulnerabilities in GSS message token handling [CVE-2024-37370,
CVE-2024-37371].
* Fix a potential bad pointer free in krb5_cccol_have_contents().
* Fix a memory leak in the macOS ccache type.
krb5-1.21.2 changes by ticket ID
--------------------------------
9102 Eliminate sim_client include of getopt.h
9103 segfault trying to free a garbage pointer
9104 Work around Doxygen 1.9.7 change
9107 In PKINIT, check for null PKCS7 enveloped fields
9109 memory leak on macos
9115 Fix leak in KDC NDR encoding
9125 Formatting error in realm_config.rst
9128 Fix vulnerabilities in GSS message token handling
Major changes in 1.21.2 (2023-08-14)
------------------------------------
This is a bug fix release.
* Fix double-free in KDC TGS processing [CVE-2023-39975].
krb5-1.21.2 changes by ticket ID
--------------------------------
9101 Fix double-free in KDC TGS processing
Major changes in 1.21.1 (2023-07-10)
------------------------------------
This is a bug fix release.
* Fix potential uninitialized pointer free in kadm5 XDR parsing
[CVE-2023-36054].
krb5-1.21.1 changes by ticket ID
--------------------------------
9099 Ensure array count consistency in kadm5 RPC
Major changes in 1.21 (2023-06-05)
----------------------------------
User experience:
* Added a credential cache type providing compatibility with the macOS
11 native credential cache.
Developer experience:
* libkadm5 will use the provided krb5_context object to read
configuration values, instead of creating its own.
* Added an interface to retrieve the ticket session key from a GSS
context.
Protocol evolution:
* The KDC will no longer issue tickets with RC4 or triple-DES session
keys unless explicitly configured with the new allow_rc4 or
allow_des3 variables respectively.
* The KDC will assume that all services can handle aes256-sha1 session
keys unless the service principal has a session_enctypes string
attribute.
* Support for PAC full KDC checksums has been added to mitigate an
S4U2Proxy privilege escalation attack.
* The PKINIT client will advertise a more modern set of supported CMS
algorithms.
Code quality:
* Removed unused code in libkrb5, libkrb5support, and the PKINIT
module.
* Modernized the KDC code for processing TGS requests, the code for
encrypting and decrypting key data, the PAC handling code, and the
GSS library packet parsing and composition code.
* Improved the test framework's detection of memory errors in daemon
processes when used with asan.
krb5-1.21 changes by ticket ID
------------------------------
9052 Support macOS 11 native credential cache
9053 Make kprop work for dump files larger than 4GB
9054 Replace macros with typedefs in gssrpc types.h
9055 Use SHA-256 instead of SHA-1 for PKINIT CMS digest
9057 Omit LDFLAGS from krb5-config --libs output
9058 Add configure variable for default PKCS#11 module
9059 Use context profile for libkadm5 configuration
9066 Set reasonable supportedCMSTypes in PKINIT
9069 Update error checking for OpenSSL CMS_verify
9071 Add and use ts_interval() helper
9072 Avoid small read overrun in UTF8 normalization
9076 Use memmove() in Unicode functions
9077 Fix aclocal.m4 syntax error for autoconf 2.72
9078 Fix profile crash on memory exhaustion
9079 Fix preauth crash on memory exhaustion
9080 Fix gic_keytab crash on memory exhaustion
9082 Fix policy DB fallback error handling
9083 Fix kpropd crash with unrecognized option
9084 Add PAC full checksums
9085 Fix read overruns in SPNEGO parsing
9086 Fix possible double-free during KDB creation
9087 Fix meridian type in getdate.y
9088 Use control flow guard flag in Windows builds
9089 Add pac_privsvr_enctype string attribute
9090 Convey realm names to certauth modules
9091 Add GSS_C_INQ_ODBC_SESSION_KEY
9092 Fix maintainer-mode build for binutils 2.37
9093 Add PA-REDHAT-PASSKEY padata type
Acknowledgements
----------------
Past Sponsors of the MIT Kerberos Consortium:
Apple
Carnegie Mellon University
Centrify Corporation
Columbia University
Cornell University
The Department of Defense of the United States of America (DoD)
Fidelity Investments
Google
Iowa State University
MIT
Michigan State University
Microsoft
MITRE Corporation
Morgan-Stanley
The National Aeronautics and Space Administration
of the United States of America (NASA)
Network Appliance (NetApp)
Nippon Telephone and Telegraph (NTT)
US Government Office of the National Coordinator for Health
Information Technology (ONC)
Oracle
Pennsylvania State University
Red Hat
Stanford University
TeamF1, Inc.
The University of Alaska
The University of Michigan
The University of Pennsylvania
Past and present members of the Kerberos Team at MIT:
Danilo Almeida
Jeffrey Altman
Justin Anderson
Richard Basch
Mitch Berger
Jay Berkenbilt
Andrew Boardman
Bill Bryant
Steve Buckley
Joe Calzaretta
John Carr
Mark Colan
Don Davis
Sarah Day
Alexandra Ellwood
Carlos Garay
Dan Geer
Nancy Gilman
Matt Hancher
Thomas Hardjono
Sam Hartman
Paul Hill
Marc Horowitz
Eva Jacobus
Miroslav Jurisic
Barry Jaspan
Benjamin Kaduk
Geoffrey King
Kevin Koch
John Kohl
HaoQi Li
Jonathan Lin
Peter Litwack
Scott McGuire
Steve Miller
Kevin Mitchell
Cliff Neuman
Paul Park
Ezra Peisach
Chris Provenzano
Ken Raeburn
Jon Rochlis
Jeff Schiller
Jen Selby
Robert Silk
Bill Sommerfeld
Jennifer Steiner
Ralph Swick
Brad Thompson
Harry Tsai
Zhanna Tsitkova
Ted Ts'o
Marshall Vale
Taylor Yu
The following external contributors have provided code, patches, bug
reports, suggestions, and valuable resources:
Ian Abbott
Daniel Albers
Brandon Allbery
Russell Allbery
Brian Almeida
Michael B Allen
Pooja Anil
Jeffrey Arbuckle
Heinz-Ado Arnolds
Derek Atkins
Mark Bannister
David Bantz
Alex Baule
Nikhil Benesch
David Benjamin
Thomas Bernard
Adam Bernstein
Arlene Berry
Jeff Blaine
Toby Blake
Radoslav Bodo
Alexander Bokovoy
Sumit Bose
Emmanuel Bouillon
Isaac Boukris
Ulf Bremer
Pavel Březina
Philip Brown
Samuel Cabrero
Michael Calmer
Andrea Campi
Julien Chaffraix
Jacob Champion
Puran Chand
Ravi Channavajhala
Srinivas Cheruku
Leonardo Chiquitto
Rachit Chokshi
Seemant Choudhary
Howard Chu
Andrea Cirulli
Christopher D. Clausen
Kevin Coffman
Simon Cooper
Sylvain Cortes
Ian Crowther
Arran Cudbard-Bell
Adam Dabrowski
Jeff D'Angelo
Nalin Dahyabhai
Mark Davies
Dennis Davis
Alex Dehnert
Misty De Meo
Mark Deneen
Günther Deschner
John Devitofranceschi
Marc Dionne
Roland Dowdeswell
Ken Dreyer
Dorian Ducournau
Viktor Dukhovni
Jason Edgecombe
Mark Eichin
Shawn M. Emery
Douglas E. Engert
Peter Eriksson
Juha Erkkilä
Gilles Espinasse
Sergey Fedorov
Ronni Feldt
Bill Fellows
JC Ferguson
Remi Ferrand
Paul Fertser
Fabiano Fidêncio
Frank Filz
William Fiveash
Jacques Florent
Oliver Freyermuth
Ákos Frohner
Sebastian Galiano
Marcus Granado
Dylan Gray
Norm Green
Scott Grizzard
Helmut Grohne
Steve Grubb
Philip Guenther
Timo Gurr
Dominic Hargreaves
Robbie Harwood
John Hascall
Jakob Haufe
Matthieu Hautreux
Jochen Hein
Paul B. Henson
Kihong Heo
Jeff Hodges
Christopher Hogan
Love Hörnquist Åstrand
Ken Hornstein
Henry B. Hotz
Luke Howard
Jakub Hrozek
Shumon Huque
Jeffrey Hutzelman
Sergey Ilinykh
Wyllys Ingersoll
Holger Isenberg
Spencer Jackson
Diogenes S. Jesus
Mike Jetzer
Pavel Jindra
Brian Johannesmeyer
Joel Johnson
Lutz Justen
Ganesh Kamath
Alexander Karaivanov
Anders Kaseorg
Bar Katz
Zentaro Kavanagh
Mubashir Kazia
W. Trevor King
Patrik Kis
Martin Kittel
Thomas Klausner
Tomasz Kłoczko
Matthew Krupcale
Mikkel Kruse
Reinhard Kugler
Harshawardhan Kulkarni
Tomas Kuthan
Pierre Labastie
Andreas Ladanyi
Chris Leick
Volker Lendecke
Jan iankko Lieskovsky
Todd Lipcon
Oliver Loch
Chris Long
Kevin Longfellow
Frank Lonigro
Jon Looney
Nuno Lopes
Todd Lubin
Ryan Lynch
Glenn Machin
Roland Mainz
Sorin Manolache
Robert Marshall
Andrei Maslennikov
Michael Mattioli
Nathaniel McCallum
Greg McClement
Cameron Meadors
Vipul Mehta
Alexey Melnikov
Ivan A. Melnikov
Franklyn Mendez
Mantas Mikulėnas
Markus Moeller
Kyle Moffett
Jon Moore
Paul Moore
Keiichi Mori
Michael Morony
Robert Morris
Sam Morris
Zbysek Mraz
Edward Murrell
Joshua Neuheisel
Nikos Nikoleris
Demi Obenour
Felipe Ortega
Michael Osipov
Andrej Ota
Dmitri Pal
Javier Palacios
Dilyan Palauzov
Tom Parker
Eric Pauly
Leonard Peirce
Ezra Peisach
Alejandro Perez
Zoran Pericic
W. Michael Petullo
Mark Phalan
Sharwan Ram
Brett Randall
Jonathan Reams
Jonathan Reed
Robert Relyea
Tony Reix
Martin Rex
Pat Riehecky
Julien Rische
Jason Rogers
Matt Rogers
Nate Rosenblum
Solly Ross
Mike Roszkowski
Guillaume Rousse
Joshua Schaeffer
Alexander Scheel
Jens Schleusener
Ryan Schmidt
Andreas Schneider
Paul Seyfert
Tom Shaw
Jim Shi
Jerry Shipman
Peter Shoults
Richard Silverman
Cel Skeggs
Simo Sorce
Anthony Sottile
Michael Spang
Michael Ströder
Bjørn Tore Sund
Ondřej Surý
Joseph Sutton
Joe Travaglini
Sergei Trofimovich
Greg Troxel
Fraser Tweedale
Tim Uglow
Rathor Vipin
Denis Vlasenko
Thomas Wagner
Jorgen Wahlsten
Stef Walter
Max (Weijun) Wang
John Washington
Stef Walter
Xi Wang
Nehal J Wani
Kevin Wasserman
Margaret Wasserman
Marcus Watts
Andreas Wiese
Simon Wilkinson
Nicolas Williams
Ross Wilper
Augustin Wolf
Garrett Wollman
David Woodhouse
Tsu-Phong Wu
Xu Qiang
Neng Xue
Zhaomo Yang
Tianjiao Yin
Nickolai Zeldovich
Bean Zhang
ChenChen Zhou
Hanz van Zijst
Gertjan Zwartjes
The above is not an exhaustive list; many others have contributed in
various ways to the MIT Kerberos development effort over the years.

56
crypto/krb5/doc/README Normal file
View file

@ -0,0 +1,56 @@
BUILDING
========
See doc/build_this.rst for details about how to build the
documentation.
CONVENTIONS
===========
We use the following conventions:
* Use four-space indentation where indentation levels are arbitrary.
Do not use tabs anywhere. Avoid trailing whitespace at the end of
lines or files.
* Fill lines to 70 columns (the emacs default) where lines can be
wrapped.
* For section headers, use === underlines for page titles, --- for
sections, ~~~ for subsections, and ### for sub-subsections. Make
underlines exactly as long as titles. Do not include trailing
punctuation in section headers. Do not capitalize section headers
(except for the first word) except in source files intended to
generate man pages.
* For bullet lists, use * for top-level bullets and - for sub-bullets.
Do not indent bullet or enumerated lists relative to the surrounding
text.
* Use italics (*word*) for words representing variables or parameters.
Use boldface (**word**) for command options, subcommands of programs
like kadmin, and krb5.conf/kdc.conf parameter names. Use literal
text (``text``) for examples and multi-component pathnames. For
command names, single-component filenames, and krb5.conf/kdc.conf
section names, use references (like :ref:`kadmin(1)`) if introducing
them, or just use them unadorned otherwise.
* In man pages for commands with subcommands, make a subsection for
each subcommand. Start the subcommand with an indented synopsis,
then follow with non-indented text describing the subcommand and its
options. See kadmin_local.rst for an example.
* In man page synopses, put a newline in the RST source before each
option. Put all parts of the synopsis at the same indentation
level. Ideally we would want a hanging indent to the width of the
command or subcommand name, but RST doesn't support that. Use
boldface for literal text in the synopsis, italics for variable
text, and unadorned text for syntax symbols (such as square brackets
to indicate optional parameters). If immediately following one kind
of inline markup with another or putting inline markup next to
punctuation, you may need to use "\ " as a dummy separator.
* For directives that take a content block (e.g., note, error, and
warning), leave a blank line before the content block (after any
arguments or options that may be present).

169
crypto/krb5/doc/_static/kerb.css_t vendored Normal file
View file

@ -0,0 +1,169 @@
/*
* kerb.css
* ~~~~~~~~~~~
*
* Sphinx stylesheet -- modification to agogo theme.
*
*/
div.body {
padding-right: .5em;
text-align: left;
overflow-x: hidden;
}
/* Page layout */
div.header, div.content, div.footer {
margin-left: auto;
margin-right: auto;
padding-left: 1em;
padding-right: 1em;
max-width: 60em;
}
div.header-wrapper {
background: white;
border-bottom: 3px solid #2e3436;
border-top: 13px solid #5d1509;
}
/* Header */
div.header {
padding-top: 10px;
padding-bottom: 0px;
}
div.header h1 {
font-family: "Georgia", "Times New Roman", serif, black;
font-weight: normal;
}
div.header h1 a {
color: {{ theme_bgcolor }};
font-size: 120%;
padding-top: 10px;
}
div.header div.right a {
color: #fcaf3e;
letter-spacing: .1em;
text-transform: lowercase;
float: right;
}
div.header div.rel {
font-family: "Georgia", "Times New Roman", serif, black;
font-weight: normal;
margin-bottom: 1.6em;
}
/* Content */
div.document {
width: 80%;
float: left;
margin: 0;
background-color: white;
padding-top: 20px;
padding-bottom: 20px;
}
div.document div.section h1 {
margin-bottom: 20px;
padding: 1px;
line-height: 130%;
}
div.document div.section dl {
margin-top: 15px;
margin-bottom: 5px;
padding: 1px;
text-align: left;
}
/* Sidebar */
div.sidebar {
float: right;
font-size: .9em;
width: 20%;
margin: 0;
padding: 0;
background-color: #F9F9F9;
}
div.sidebar ul {
list-style-type: none;
margin-left: .5em;
}
div.sidebar li.toctree-l1 a {
margin-left: .5em;
}
div.sidebar li.toctree-l2 a {
margin-left: .5em;
}
div.sidebar li.toctree-l3 a {
margin-left: .5em;
}
div.sidebar li.toctree-l2.current a {
border-right: 2px solid #fcaf3e !important;
}
div.sidebar li.toctree-l3.current a {
font-weight: bold;
}
div.sidebar li.toctree-l4 a {
display: none;
}
div.sidebar input[type=text] {
width: auto;
}
/* Other body styles */
dt:target, .highlighted {
background-color: #c1c1c1;
}
/* Code displays */
pre {
overflow: auto;
overflow-y: hidden;
}
td.linenos pre {
padding: 5px 0px;
border: 0;
background-color: transparent;
color: #aaa;
}
/* ordered lists */
ol.arabic {
list-style: decimal;
}
ol.loweralpha {
list-style: lower-alpha;
}
ol.upperalpha {
list-style: upper-alpha;
}
ol.lowerroman {
list-style-type: lower-roman;
}
ol.upperroman {
list-style-type: upper-roman;
}

73
crypto/krb5/doc/_templates/layout.html vendored Normal file
View file

@ -0,0 +1,73 @@
{% extends "!layout.html" %}
{% set rellinks = [('search', 'Enter search criteria', 'S', 'Search')] +
rellinks +
[('index', 'Full Table of Contents', 'C', 'Contents')] %}
{# Add a "feedback" button to the rellinks #}
{%- macro feedback_rellinks() %}
{%- for rellink in rellinks|reverse %}
<a href="{{ pathto(rellink[0]) }}" title="{{ rellink[1]|striptags|e }}"
{{ accesskey(rellink[2]) }}>{{ rellink[3] }}</a>{{ reldelim2 }}
{%- endfor %}
<a href="mailto:krb5-bugs@mit.edu?subject=Documentation__{{ title }}">feedback</a>
{%- endmacro %}
{% block footer %}
<div class="footer-wrapper">
<div class="footer" >
<div class="right" ><i>Release: {{ release }}</i><br />
{%- if show_copyright %}
{%- if hasdoc('copyright') %}
{% trans path=pathto('copyright'), copyright=copyright|e %}&copy; <a href="{{ path }}">Copyright</a> {{ copyright }}.{% endtrans %}
{%- else %}
{% trans copyright=copyright|e %}&copy; Copyright {{ copyright }}.{% endtrans %}
{%- endif %}
{%- endif %}
</div>
<div class="left">
{{ feedback_rellinks() }}
</div>
</div>
</div>
{% endblock %}
{% block header %}
<div class="header-wrapper">
<div class="header">
{% if logo %}
<p class="logo">
{# Link logo to kerberos.org #}
<a href="https://kerberos.org"> <img class="logo"
src="{{ pathto('_static/' + logo, 1) }}" alt="Logo" /></a>
</p>
{% endif %}
{% block headertitle %}
<h1><a href="{{ pathto(master_doc) }}">{{ shorttitle|e }}</a></h1>
{% endblock %}
<div class="rel">
{{ feedback_rellinks() }}
</div>
</div>
</div>
{% endblock %}
{%- block sidebartoc %}
<h2>{{ _('On this page') }}</h2>
{{ toc }}
<br/>
<h2>{{ _('Table of contents') }}</h2>
{{ toctree(collapse=true, maxdepth=3, titles_only=true,
includehidden=false) }}
<br/>
<h4><a href="{{ pathto('index') }}">Full Table of Contents</a></h4>
{%- endblock %}
{%- block sidebarsearch %}
<h4>{{ _('Search') }}</h4>
<form class="search" action="{{ pathto('search') }}" method="get">
<input type="text" name="q" size="18" />
<input type="submit" value="{{ _('Go') }}" />
<input type="hidden" name="check_keywords" value="yes" />
<input type="hidden" name="area" value="default" />
</form>
{%- endblock %}

35
crypto/krb5/doc/about.rst Normal file
View file

@ -0,0 +1,35 @@
Contributing to the MIT Kerberos Documentation
==============================================
We are looking for documentation writers and editors who could contribute
towards improving the MIT KC documentation content. If you are an experienced
Kerberos developer and/or administrator, please consider sharing your knowledge
and experience with the Kerberos Community. You can suggest your own topic or
write about any of the topics listed
`here <https://k5wiki.kerberos.org/wiki/Projects/Documentation_Tasks>`__.
If you have any questions, comments, or suggestions on the existing documents,
please send your feedback via email to krb5-bugs@mit.edu. The HTML version of
this documentation has a "FEEDBACK" link to the krb5-bugs@mit.edu email
address with a pre-constructed subject line.
Background
----------
Starting with release 1.11, the Kerberos documentation set is
unified in a central form. Man pages, HTML documentation, and PDF
documents are compiled from reStructuredText sources, and the application
developer documentation incorporates Doxygen markup from the source
tree. This project was undertaken along the outline described
`here <https://k5wiki.kerberos.org/wiki/Projects/Kerberos_Documentation>`__.
Previous versions of Kerberos 5 attempted to maintain separate documentation
in the texinfo format, with separate groff manual pages. Having the API
documentation disjoint from the source code implementing that API
resulted in the documentation becoming stale, and over time the documentation
ceased to match reality. With a fresh start and a source format that is
easier to use and maintain, reStructuredText-based documents should provide
an improved experience for the user. Consolidating all the documentation
formats into a single source document makes the documentation set easier
to maintain.

View file

@ -0,0 +1,17 @@
Administration programs
========================
.. toctree::
:maxdepth: 1
kadmin_local.rst
kadmind.rst
kdb5_util.rst
kdb5_ldap_util.rst
krb5kdc.rst
kprop.rst
kpropd.rst
kproplog.rst
ktutil.rst
k5srvutil.rst
sserver.rst

View file

@ -0,0 +1,69 @@
.. _k5srvutil(1):
k5srvutil
=========
SYNOPSIS
--------
**k5srvutil** *operation*
[**-i**]
[**-f** *filename*]
[**-e** *keysalts*]
DESCRIPTION
-----------
k5srvutil allows an administrator to list keys currently in
a keytab, to obtain new keys for a principal currently in a keytab,
or to delete non-current keys from a keytab.
*operation* must be one of the following:
**list**
Lists the keys in a keytab, showing version number and principal
name.
**change**
Uses the kadmin protocol to update the keys in the Kerberos
database to new randomly-generated keys, and updates the keys in
the keytab to match. If a key's version number doesn't match the
version number stored in the Kerberos server's database, then the
operation will fail. If the **-i** flag is given, k5srvutil will
prompt for confirmation before changing each key. If the **-k**
option is given, the old and new keys will be displayed.
Ordinarily, keys will be generated with the default encryption
types and key salts. This can be overridden with the **-e**
option. Old keys are retained in the keytab so that existing
tickets continue to work, but **delold** should be used after
such tickets expire, to prevent attacks against the old keys.
**delold**
Deletes keys that are not the most recent version from the keytab.
This operation should be used some time after a change operation
to remove old keys, after existing tickets issued for the service
have expired. If the **-i** flag is given, then k5srvutil will
prompt for confirmation for each principal.
**delete**
Deletes particular keys in the keytab, interactively prompting for
each key.
In all cases, the default keytab is used unless this is overridden by
the **-f** option.
k5srvutil uses the :ref:`kadmin(1)` program to edit the keytab in
place.
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kadmin(1)`, :ref:`ktutil(1)`, :ref:`kerberos(7)`

View file

@ -0,0 +1,985 @@
.. _kadmin(1):
kadmin
======
SYNOPSIS
--------
.. _kadmin_synopsis:
**kadmin**
[**-O**\|\ **-N**]
[**-r** *realm*]
[**-p** *principal*]
[**-q** *query*]
[[**-c** *cache_name*]\|[**-k** [**-t** *keytab*]]\|\ **-n**]
[**-w** *password*]
[**-s** *admin_server*\ [:*port*]]
[command args...]
**kadmin.local**
[**-r** *realm*]
[**-p** *principal*]
[**-q** *query*]
[**-d** *dbname*]
[**-e** *enc*:*salt* ...]
[**-m**]
[**-x** *db_args*]
[command args...]
DESCRIPTION
-----------
kadmin and kadmin.local are command-line interfaces to the Kerberos V5
administration system. They provide nearly identical functionalities;
the difference is that kadmin.local directly accesses the KDC
database, while kadmin performs operations using :ref:`kadmind(8)`.
Except as explicitly noted otherwise, this man page will use "kadmin"
to refer to both versions. kadmin provides for the maintenance of
Kerberos principals, password policies, and service key tables
(keytabs).
The remote kadmin client uses Kerberos to authenticate to kadmind
using the service principal ``kadmin/admin`` or ``kadmin/ADMINHOST``
(where *ADMINHOST* is the fully-qualified hostname of the admin
server). If the credentials cache contains a ticket for one of these
principals, and the **-c** credentials_cache option is specified, that
ticket is used to authenticate to kadmind. Otherwise, the **-p** and
**-k** options are used to specify the client Kerberos principal name
used to authenticate. Once kadmin has determined the principal name,
it requests a service ticket from the KDC, and uses that service
ticket to authenticate to kadmind.
Since kadmin.local directly accesses the KDC database, it usually must
be run directly on the primary KDC with sufficient permissions to read
the KDC database. If the KDC database uses the LDAP database module,
kadmin.local can be run on any host which can access the LDAP server.
OPTIONS
-------
.. _kadmin_options:
**-r** *realm*
Use *realm* as the default database realm.
**-p** *principal*
Use *principal* to authenticate. Otherwise, kadmin will append
``/admin`` to the primary principal name of the default ccache,
the value of the **USER** environment variable, or the username as
obtained with getpwuid, in order of preference.
**-k**
Use a keytab to decrypt the KDC response instead of prompting for
a password. In this case, the default principal will be
``host/hostname``. If there is no keytab specified with the
**-t** option, then the default keytab will be used.
**-t** *keytab*
Use *keytab* to decrypt the KDC response. This can only be used
with the **-k** option.
**-n**
Requests anonymous processing. Two types of anonymous principals
are supported. For fully anonymous Kerberos, configure PKINIT on
the KDC and configure **pkinit_anchors** in the client's
:ref:`krb5.conf(5)`. Then use the **-n** option with a principal
of the form ``@REALM`` (an empty principal name followed by the
at-sign and a realm name). If permitted by the KDC, an anonymous
ticket will be returned. A second form of anonymous tickets is
supported; these realm-exposed tickets hide the identity of the
client but not the client's realm. For this mode, use ``kinit
-n`` with a normal principal name. If supported by the KDC, the
principal (but not realm) will be replaced by the anonymous
principal. As of release 1.8, the MIT Kerberos KDC only supports
fully anonymous operation.
**-c** *credentials_cache*
Use *credentials_cache* as the credentials cache. The cache
should contain a service ticket for the ``kadmin/admin`` or
``kadmin/ADMINHOST`` (where *ADMINHOST* is the fully-qualified
hostname of the admin server) service; it can be acquired with the
:ref:`kinit(1)` program. If this option is not specified, kadmin
requests a new service ticket from the KDC, and stores it in its
own temporary ccache.
**-w** *password*
Use *password* instead of prompting for one. Use this option with
care, as it may expose the password to other users on the system
via the process list.
**-q** *query*
Perform the specified query and then exit.
**-d** *dbname*
Specifies the name of the KDC database. This option does not
apply to the LDAP database module.
**-s** *admin_server*\ [:*port*]
Specifies the admin server which kadmin should contact.
**-m**
If using kadmin.local, prompt for the database master password
instead of reading it from a stash file.
**-e** "*enc*:*salt* ..."
Sets the keysalt list to be used for any new keys created. See
:ref:`Keysalt_lists` in :ref:`kdc.conf(5)` for a list of possible
values.
**-O**
Force use of old AUTH_GSSAPI authentication flavor.
**-N**
Prevent fallback to AUTH_GSSAPI authentication flavor.
**-x** *db_args*
Specifies the database specific arguments. See the next section
for supported options.
Starting with release 1.14, if any command-line arguments remain after
the options, they will be treated as a single query to be executed.
This mode of operation is intended for scripts and behaves differently
from the interactive mode in several respects:
* Query arguments are split by the shell, not by kadmin.
* Informational and warning messages are suppressed. Error messages
and query output (e.g. for **get_principal**) will still be
displayed.
* Confirmation prompts are disabled (as if **-force** was given).
Password prompts will still be issued as required.
* The exit status will be non-zero if the query fails.
The **-q** option does not carry these behavior differences; the query
will be processed as if it was entered interactively. The **-q**
option cannot be used in combination with a query in the remaining
arguments.
.. _dboptions:
DATABASE OPTIONS
----------------
Database options can be used to override database-specific defaults.
Supported options for the DB2 module are:
**-x dbname=**\ \*filename*
Specifies the base filename of the DB2 database.
**-x lockiter**
Make iteration operations hold the lock for the duration of
the entire operation, rather than temporarily releasing the
lock while handling each principal. This is the default
behavior, but this option exists to allow command line
override of a [dbmodules] setting. First introduced in
release 1.13.
**-x unlockiter**
Make iteration operations unlock the database for each
principal, instead of holding the lock for the duration of the
entire operation. First introduced in release 1.13.
Supported options for the LDAP module are:
**-x host=**\ *ldapuri*
Specifies the LDAP server to connect to by a LDAP URI.
**-x binddn=**\ *bind_dn*
Specifies the DN used to bind to the LDAP server.
**-x bindpwd=**\ *password*
Specifies the password or SASL secret used to bind to the LDAP
server. Using this option may expose the password to other
users on the system via the process list; to avoid this,
instead stash the password using the **stashsrvpw** command of
:ref:`kdb5_ldap_util(8)`.
**-x sasl_mech=**\ *mechanism*
Specifies the SASL mechanism used to bind to the LDAP server.
The bind DN is ignored if a SASL mechanism is used. New in
release 1.13.
**-x sasl_authcid=**\ *name*
Specifies the authentication name used when binding to the
LDAP server with a SASL mechanism, if the mechanism requires
one. New in release 1.13.
**-x sasl_authzid=**\ *name*
Specifies the authorization name used when binding to the LDAP
server with a SASL mechanism. New in release 1.13.
**-x sasl_realm=**\ *realm*
Specifies the realm used when binding to the LDAP server with
a SASL mechanism, if the mechanism uses one. New in release
1.13.
**-x debug=**\ *level*
sets the OpenLDAP client library debug level. *level* is an
integer to be interpreted by the library. Debugging messages
are printed to standard error. New in release 1.12.
COMMANDS
--------
When using the remote client, available commands may be restricted
according to the privileges specified in the :ref:`kadm5.acl(5)` file
on the admin server.
.. _add_principal:
add_principal
~~~~~~~~~~~~~
**add_principal** [*options*] *newprinc*
Creates the principal *newprinc*, prompting twice for a password. If
no password policy is specified with the **-policy** option, and the
policy named ``default`` is assigned to the principal if it exists.
However, creating a policy named ``default`` will not automatically
assign this policy to previously existing principals. This policy
assignment can be suppressed with the **-clearpolicy** option.
This command requires the **add** privilege.
Aliases: **addprinc**, **ank**
Options:
**-expire** *expdate*
(:ref:`getdate` string) The expiration date of the principal.
**-pwexpire** *pwexpdate*
(:ref:`getdate` string) The password expiration date.
**-maxlife** *maxlife*
(:ref:`duration` or :ref:`getdate` string) The maximum ticket life
for the principal.
**-maxrenewlife** *maxrenewlife*
(:ref:`duration` or :ref:`getdate` string) The maximum renewable
life of tickets for the principal.
**-kvno** *kvno*
The initial key version number.
**-policy** *policy*
The password policy used by this principal. If not specified, the
policy ``default`` is used if it exists (unless **-clearpolicy**
is specified).
**-clearpolicy**
Prevents any policy from being assigned when **-policy** is not
specified.
{-\|+}\ **allow_postdated**
**-allow_postdated** prohibits this principal from obtaining
postdated tickets. **+allow_postdated** clears this flag.
{-\|+}\ **allow_forwardable**
**-allow_forwardable** prohibits this principal from obtaining
forwardable tickets. **+allow_forwardable** clears this flag.
{-\|+}\ **allow_renewable**
**-allow_renewable** prohibits this principal from obtaining
renewable tickets. **+allow_renewable** clears this flag.
{-\|+}\ **allow_proxiable**
**-allow_proxiable** prohibits this principal from obtaining
proxiable tickets. **+allow_proxiable** clears this flag.
{-\|+}\ **allow_dup_skey**
**-allow_dup_skey** disables user-to-user authentication for this
principal by prohibiting others from obtaining a service ticket
encrypted in this principal's TGT session key.
**+allow_dup_skey** clears this flag.
{-\|+}\ **requires_preauth**
**+requires_preauth** requires this principal to preauthenticate
before being allowed to kinit. **-requires_preauth** clears this
flag. When **+requires_preauth** is set on a service principal,
the KDC will only issue service tickets for that service principal
if the client's initial authentication was performed using
preauthentication.
{-\|+}\ **requires_hwauth**
**+requires_hwauth** requires this principal to preauthenticate
using a hardware device before being allowed to kinit.
**-requires_hwauth** clears this flag. When **+requires_hwauth** is
set on a service principal, the KDC will only issue service tickets
for that service principal if the client's initial authentication was
performed using a hardware device to preauthenticate.
{-\|+}\ **ok_as_delegate**
**+ok_as_delegate** sets the **okay as delegate** flag on tickets
issued with this principal as the service. Clients may use this
flag as a hint that credentials should be delegated when
authenticating to the service. **-ok_as_delegate** clears this
flag.
{-\|+}\ **allow_svr**
**-allow_svr** prohibits the issuance of service tickets for this
principal. In release 1.17 and later, user-to-user service
tickets are still allowed unless the **-allow_dup_skey** flag is
also set. **+allow_svr** clears this flag.
{-\|+}\ **allow_tgs_req**
**-allow_tgs_req** specifies that a Ticket-Granting Service (TGS)
request for a service ticket for this principal is not permitted.
**+allow_tgs_req** clears this flag.
{-\|+}\ **allow_tix**
**-allow_tix** forbids the issuance of any tickets for this
principal. **+allow_tix** clears this flag.
{-\|+}\ **needchange**
**+needchange** forces a password change on the next initial
authentication to this principal. **-needchange** clears this
flag.
{-\|+}\ **password_changing_service**
**+password_changing_service** marks this principal as a password
change service principal.
{-\|+}\ **ok_to_auth_as_delegate**
**+ok_to_auth_as_delegate** allows this principal to acquire
forwardable tickets to itself from arbitrary users, for use with
constrained delegation.
{-\|+}\ **no_auth_data_required**
**+no_auth_data_required** prevents PAC or AD-SIGNEDPATH data from
being added to service tickets for the principal.
{-\|+}\ **lockdown_keys**
**+lockdown_keys** prevents keys for this principal from leaving
the KDC via kadmind. The chpass and extract operations are denied
for a principal with this attribute. The chrand operation is
allowed, but will not return the new keys. The delete and rename
operations are also denied if this attribute is set, in order to
prevent a malicious administrator from replacing principals like
krbtgt/* or kadmin/* with new principals without the attribute.
This attribute can be set via the network protocol, but can only
be removed using kadmin.local.
**-randkey**
Sets the key of the principal to a random value.
**-nokey**
Causes the principal to be created with no key. New in release
1.12.
**-pw** *password*
Sets the password of the principal to the specified string and
does not prompt for a password. Note: using this option in a
shell script may expose the password to other users on the system
via the process list.
**-e** *enc*:*salt*,...
Uses the specified keysalt list for setting the keys of the
principal. See :ref:`Keysalt_lists` in :ref:`kdc.conf(5)` for a
list of possible values.
**-x** *db_princ_args*
Indicates database-specific options. The options for the LDAP
database module are:
**-x dn=**\ *dn*
Specifies the LDAP object that will contain the Kerberos
principal being created.
**-x linkdn=**\ *dn*
Specifies the LDAP object to which the newly created Kerberos
principal object will point.
**-x containerdn=**\ *container_dn*
Specifies the container object under which the Kerberos
principal is to be created.
**-x tktpolicy=**\ *policy*
Associates a ticket policy to the Kerberos principal.
.. note::
- The **containerdn** and **linkdn** options cannot be
specified with the **dn** option.
- If the *dn* or *containerdn* options are not specified while
adding the principal, the principals are created under the
principal container configured in the realm or the realm
container.
- *dn* and *containerdn* should be within the subtrees or
principal container configured in the realm.
Example::
kadmin: addprinc jennifer
No policy specified for "jennifer@ATHENA.MIT.EDU";
defaulting to no policy.
Enter password for principal jennifer@ATHENA.MIT.EDU:
Re-enter password for principal jennifer@ATHENA.MIT.EDU:
Principal "jennifer@ATHENA.MIT.EDU" created.
kadmin:
.. _modify_principal:
modify_principal
~~~~~~~~~~~~~~~~
**modify_principal** [*options*] *principal*
Modifies the specified principal, changing the fields as specified.
The options to **add_principal** also apply to this command, except
for the **-randkey**, **-pw**, and **-e** options. In addition, the
option **-clearpolicy** will clear the current policy of a principal.
This command requires the *modify* privilege.
Alias: **modprinc**
Options (in addition to the **addprinc** options):
**-unlock**
Unlocks a locked principal (one which has received too many failed
authentication attempts without enough time between them according
to its password policy) so that it can successfully authenticate.
.. _rename_principal:
rename_principal
~~~~~~~~~~~~~~~~
**rename_principal** [**-force**] *old_principal* *new_principal*
Renames the specified *old_principal* to *new_principal*. This
command prompts for confirmation, unless the **-force** option is
given.
This command requires the **add** and **delete** privileges.
Alias: **renprinc**
.. _delete_principal:
delete_principal
~~~~~~~~~~~~~~~~
**delete_principal** [**-force**] *principal*
Deletes the specified *principal* from the database. This command
prompts for deletion, unless the **-force** option is given.
This command requires the **delete** privilege.
Alias: **delprinc**
.. _change_password:
change_password
~~~~~~~~~~~~~~~
**change_password** [*options*] *principal*
Changes the password of *principal*. Prompts for a new password if
neither **-randkey** or **-pw** is specified.
This command requires the **changepw** privilege, or that the
principal running the program is the same as the principal being
changed.
Alias: **cpw**
The following options are available:
**-randkey**
Sets the key of the principal to a random value.
**-pw** *password*
Set the password to the specified string. Using this option in a
script may expose the password to other users on the system via
the process list.
**-e** *enc*:*salt*,...
Uses the specified keysalt list for setting the keys of the
principal. See :ref:`Keysalt_lists` in :ref:`kdc.conf(5)` for a
list of possible values.
**-keepold**
Keeps the existing keys in the database. This flag is usually not
necessary except perhaps for ``krbtgt`` principals.
Example::
kadmin: cpw systest
Enter password for principal systest@BLEEP.COM:
Re-enter password for principal systest@BLEEP.COM:
Password for systest@BLEEP.COM changed.
kadmin:
.. _purgekeys:
purgekeys
~~~~~~~~~
**purgekeys** [**-all**\|\ **-keepkvno** *oldest_kvno_to_keep*] *principal*
Purges previously retained old keys (e.g., from **change_password
-keepold**) from *principal*. If **-keepkvno** is specified, then
only purges keys with kvnos lower than *oldest_kvno_to_keep*. If
**-all** is specified, then all keys are purged. The **-all** option
is new in release 1.12.
This command requires the **modify** privilege.
.. _get_principal:
get_principal
~~~~~~~~~~~~~
**get_principal** [**-terse**] *principal*
Gets the attributes of principal. With the **-terse** option, outputs
fields as quoted tab-separated strings.
This command requires the **inquire** privilege, or that the principal
running the the program to be the same as the one being listed.
Alias: **getprinc**
Examples::
kadmin: getprinc tlyu/admin
Principal: tlyu/admin@BLEEP.COM
Expiration date: [never]
Last password change: Mon Aug 12 14:16:47 EDT 1996
Password expiration date: [never]
Maximum ticket life: 0 days 10:00:00
Maximum renewable life: 7 days 00:00:00
Last modified: Mon Aug 12 14:16:47 EDT 1996 (bjaspan/admin@BLEEP.COM)
Last successful authentication: [never]
Last failed authentication: [never]
Failed password attempts: 0
Number of keys: 1
Key: vno 1, aes256-cts-hmac-sha384-192
MKey: vno 1
Attributes:
Policy: [none]
kadmin: getprinc -terse systest
systest@BLEEP.COM 3 86400 604800 1
785926535 753241234 785900000
tlyu/admin@BLEEP.COM 786100034 0 0
kadmin:
.. _list_principals:
list_principals
~~~~~~~~~~~~~~~
**list_principals** [*expression*]
Retrieves all or some principal names. *expression* is a shell-style
glob expression that can contain the wild-card characters ``?``,
``*``, and ``[]``. All principal names matching the expression are
printed. If no expression is provided, all principal names are
printed. If the expression does not contain an ``@`` character, an
``@`` character followed by the local realm is appended to the
expression.
This command requires the **list** privilege.
Alias: **listprincs**, **get_principals**, **getprincs**
Example::
kadmin: listprincs test*
test3@SECURE-TEST.OV.COM
test2@SECURE-TEST.OV.COM
test1@SECURE-TEST.OV.COM
testuser@SECURE-TEST.OV.COM
kadmin:
.. _get_strings:
get_strings
~~~~~~~~~~~
**get_strings** *principal*
Displays string attributes on *principal*.
This command requires the **inquire** privilege.
Alias: **getstrs**
.. _set_string:
set_string
~~~~~~~~~~
**set_string** *principal* *name* *value*
Sets a string attribute on *principal*. String attributes are used to
supply per-principal configuration to the KDC and some KDC plugin
modules. The following string attribute names are recognized by the
KDC:
**require_auth**
Specifies an authentication indicator which is required to
authenticate to the principal as a service. Multiple indicators
can be specified, separated by spaces; in this case any of the
specified indicators will be accepted. (New in release 1.14.)
**session_enctypes**
Specifies the encryption types supported for session keys when the
principal is authenticated to as a server. See
:ref:`Encryption_types` in :ref:`kdc.conf(5)` for a list of the
accepted values.
**otp**
Enables One Time Passwords (OTP) preauthentication for a client
*principal*. The *value* is a JSON string representing an array
of objects, each having optional ``type`` and ``username`` fields.
**pkinit_cert_match**
Specifies a matching expression that defines the certificate
attributes required for the client certificate used by the
principal during PKINIT authentication. The matching expression
is in the same format as those used by the **pkinit_cert_match**
option in :ref:`krb5.conf(5)`. (New in release 1.16.)
**pac_privsvr_enctype**
Forces the encryption type of the PAC KDC checksum buffers to the
specified encryption type for tickets issued to this server, by
deriving a key from the local krbtgt key if it is of a different
encryption type. It may be necessary to set this value to
"aes256-sha1" on the cross-realm krbtgt entry for an Active
Directory realm when using aes-sha2 keys on the local krbtgt
entry.
This command requires the **modify** privilege.
Alias: **setstr**
Example::
set_string host/foo.mit.edu session_enctypes aes128-cts
set_string user@FOO.COM otp "[{""type"":""hotp"",""username"":""al""}]"
.. _del_string:
del_string
~~~~~~~~~~
**del_string** *principal* *key*
Deletes a string attribute from *principal*.
This command requires the **delete** privilege.
Alias: **delstr**
.. _add_policy:
add_policy
~~~~~~~~~~
**add_policy** [*options*] *policy*
Adds a password policy named *policy* to the database.
This command requires the **add** privilege.
Alias: **addpol**
The following options are available:
**-maxlife** *time*
(:ref:`duration` or :ref:`getdate` string) Sets the maximum
lifetime of a password.
**-minlife** *time*
(:ref:`duration` or :ref:`getdate` string) Sets the minimum
lifetime of a password.
**-minlength** *length*
Sets the minimum length of a password.
**-minclasses** *number*
Sets the minimum number of character classes required in a
password. The five character classes are lower case, upper case,
numbers, punctuation, and whitespace/unprintable characters.
**-history** *number*
Sets the number of past keys kept for a principal. This option is
not supported with the LDAP KDC database module.
.. _policy_maxfailure:
**-maxfailure** *maxnumber*
Sets the number of authentication failures before the principal is
locked. Authentication failures are only tracked for principals
which require preauthentication. The counter of failed attempts
resets to 0 after a successful attempt to authenticate. A
*maxnumber* value of 0 (the default) disables lockout.
.. _policy_failurecountinterval:
**-failurecountinterval** *failuretime*
(:ref:`duration` or :ref:`getdate` string) Sets the allowable time
between authentication failures. If an authentication failure
happens after *failuretime* has elapsed since the previous
failure, the number of authentication failures is reset to 1. A
*failuretime* value of 0 (the default) means forever.
.. _policy_lockoutduration:
**-lockoutduration** *lockouttime*
(:ref:`duration` or :ref:`getdate` string) Sets the duration for
which the principal is locked from authenticating if too many
authentication failures occur without the specified failure count
interval elapsing. A duration of 0 (the default) means the
principal remains locked out until it is administratively unlocked
with ``modprinc -unlock``.
**-allowedkeysalts**
Specifies the key/salt tuples supported for long-term keys when
setting or changing a principal's password/keys. See
:ref:`Keysalt_lists` in :ref:`kdc.conf(5)` for a list of the
accepted values, but note that key/salt tuples must be separated
with commas (',') only. To clear the allowed key/salt policy use
a value of '-'.
Example::
kadmin: add_policy -maxlife "2 days" -minlength 5 guests
kadmin:
.. _modify_policy:
modify_policy
~~~~~~~~~~~~~
**modify_policy** [*options*] *policy*
Modifies the password policy named *policy*. Options are as described
for **add_policy**.
This command requires the **modify** privilege.
Alias: **modpol**
.. _delete_policy:
delete_policy
~~~~~~~~~~~~~
**delete_policy** [**-force**] *policy*
Deletes the password policy named *policy*. Prompts for confirmation
before deletion. The command will fail if the policy is in use by any
principals.
This command requires the **delete** privilege.
Alias: **delpol**
Example::
kadmin: del_policy guests
Are you sure you want to delete the policy "guests"?
(yes/no): yes
kadmin:
.. _get_policy:
get_policy
~~~~~~~~~~
**get_policy** [ **-terse** ] *policy*
Displays the values of the password policy named *policy*. With the
**-terse** flag, outputs the fields as quoted strings separated by
tabs.
This command requires the **inquire** privilege.
Alias: **getpol**
Examples::
kadmin: get_policy admin
Policy: admin
Maximum password life: 180 days 00:00:00
Minimum password life: 00:00:00
Minimum password length: 6
Minimum number of password character classes: 2
Number of old keys kept: 5
Reference count: 17
kadmin: get_policy -terse admin
admin 15552000 0 6 2 5 17
kadmin:
The "Reference count" is the number of principals using that policy.
With the LDAP KDC database module, the reference count field is not
meaningful.
.. _list_policies:
list_policies
~~~~~~~~~~~~~
**list_policies** [*expression*]
Retrieves all or some policy names. *expression* is a shell-style
glob expression that can contain the wild-card characters ``?``,
``*``, and ``[]``. All policy names matching the expression are
printed. If no expression is provided, all existing policy names are
printed.
This command requires the **list** privilege.
Aliases: **listpols**, **get_policies**, **getpols**.
Examples::
kadmin: listpols
test-pol
dict-only
once-a-min
test-pol-nopw
kadmin: listpols t*
test-pol
test-pol-nopw
kadmin:
.. _ktadd:
ktadd
~~~~~
| **ktadd** [options] *principal*
| **ktadd** [options] **-glob** *princ-exp*
Adds a *principal*, or all principals matching *princ-exp*, to a
keytab file. Each principal's keys are randomized in the process.
The rules for *princ-exp* are described in the **list_principals**
command.
This command requires the **inquire** and **changepw** privileges.
With the **-glob** form, it also requires the **list** privilege.
The options are:
**-k[eytab]** *keytab*
Use *keytab* as the keytab file. Otherwise, the default keytab is
used.
**-e** *enc*:*salt*,...
Uses the specified keysalt list for setting the new keys of the
principal. See :ref:`Keysalt_lists` in :ref:`kdc.conf(5)` for a
list of possible values.
**-q**
Display less verbose information.
**-norandkey**
Do not randomize the keys. The keys and their version numbers stay
unchanged. This option cannot be specified in combination with the
**-e** option.
An entry for each of the principal's unique encryption types is added,
ignoring multiple keys with the same encryption type but different
salt types.
Alias: **xst**
Example::
kadmin: ktadd -k /tmp/foo-new-keytab host/foo.mit.edu
Entry for principal host/foo.mit.edu@ATHENA.MIT.EDU with kvno 3,
encryption type aes256-cts-hmac-sha1-96 added to keytab
FILE:/tmp/foo-new-keytab
kadmin:
.. _ktremove:
ktremove
~~~~~~~~
**ktremove** [options] *principal* [*kvno* | *all* | *old*]
Removes entries for the specified *principal* from a keytab. Requires
no permissions, since this does not require database access.
If the string "all" is specified, all entries for that principal are
removed; if the string "old" is specified, all entries for that
principal except those with the highest kvno are removed. Otherwise,
the value specified is parsed as an integer, and all entries whose
kvno match that integer are removed.
The options are:
**-k[eytab]** *keytab*
Use *keytab* as the keytab file. Otherwise, the default keytab is
used.
**-q**
Display less verbose information.
Alias: **ktrem**
Example::
kadmin: ktremove kadmin/admin all
Entry for principal kadmin/admin with kvno 3 removed from keytab
FILE:/etc/krb5.keytab
kadmin:
lock
~~~~
Lock database exclusively. Use with extreme caution! This command
only works with the DB2 KDC database module.
unlock
~~~~~~
Release the exclusive database lock.
list_requests
~~~~~~~~~~~~~
Lists available for kadmin requests.
Aliases: **lr**, **?**
quit
~~~~
Exit program. If the database was locked, the lock is released.
Aliases: **exit**, **q**
HISTORY
-------
The kadmin program was originally written by Tom Yu at MIT, as an
interface to the OpenVision Kerberos administration program.
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kpasswd(1)`, :ref:`kadmind(8)`, :ref:`kerberos(7)`

View file

@ -0,0 +1,129 @@
.. _kadmind(8):
kadmind
=======
SYNOPSIS
--------
**kadmind**
[**-x** *db_args*]
[**-r** *realm*]
[**-m**]
[**-nofork**]
[**-proponly**]
[**-port** *port-number*]
[**-P** *pid_file*]
[**-p** *kdb5_util_path*]
[**-K** *kprop_path*]
[**-k** *kprop_port*]
[**-F** *dump_file*]
DESCRIPTION
-----------
kadmind starts the Kerberos administration server. kadmind typically
runs on the primary Kerberos server, which stores the KDC database.
If the KDC database uses the LDAP module, the administration server
and the KDC server need not run on the same machine. kadmind accepts
remote requests from programs such as :ref:`kadmin(1)` and
:ref:`kpasswd(1)` to administer the information in these database.
kadmind requires a number of configuration files to be set up in order
for it to work:
:ref:`kdc.conf(5)`
The KDC configuration file contains configuration information for
the KDC and admin servers. kadmind uses settings in this file to
locate the Kerberos database, and is also affected by the
**acl_file**, **dict_file**, **kadmind_port**, and iprop-related
settings.
:ref:`kadm5.acl(5)`
kadmind's ACL (access control list) tells it which principals are
allowed to perform administration actions. The pathname to the
ACL file can be specified with the **acl_file** :ref:`kdc.conf(5)`
variable; by default, it is |kdcdir|\ ``/kadm5.acl``.
After the server begins running, it puts itself in the background and
disassociates itself from its controlling terminal.
kadmind can be configured for incremental database propagation.
Incremental propagation allows replica KDC servers to receive
principal and policy updates incrementally instead of receiving full
dumps of the database. This facility can be enabled in the
:ref:`kdc.conf(5)` file with the **iprop_enable** option. Incremental
propagation requires the principal ``kiprop/PRIMARY\@REALM`` (where
PRIMARY is the primary KDC's canonical host name, and REALM the realm
name). In release 1.13, this principal is automatically created and
registered into the datebase.
OPTIONS
-------
**-r** *realm*
specifies the realm that kadmind will serve; if it is not
specified, the default realm of the host is used.
**-m**
causes the master database password to be fetched from the
keyboard (before the server puts itself in the background, if not
invoked with the **-nofork** option) rather than from a file on
disk.
**-nofork**
causes the server to remain in the foreground and remain
associated to the terminal.
**-proponly**
causes the server to only listen and respond to Kerberos replica
incremental propagation polling requests. This option can be used
to set up a hierarchical propagation topology where a replica KDC
provides incremental updates to other Kerberos replicas.
**-port** *port-number*
specifies the port on which the administration server listens for
connections. The default port is determined by the
**kadmind_port** configuration variable in :ref:`kdc.conf(5)`.
**-P** *pid_file*
specifies the file to which the PID of kadmind process should be
written after it starts up. This file can be used to identify
whether kadmind is still running and to allow init scripts to stop
the correct process.
**-p** *kdb5_util_path*
specifies the path to the kdb5_util command to use when dumping the
KDB in response to full resync requests when iprop is enabled.
**-K** *kprop_path*
specifies the path to the kprop command to use to send full dumps
to replicas in response to full resync requests.
**-k** *kprop_port*
specifies the port by which the kprop process that is spawned by
kadmind connects to the replica kpropd, in order to transfer the
dump file during an iprop full resync request.
**-F** *dump_file*
specifies the file path to be used for dumping the KDB in response
to full resync requests when iprop is enabled.
**-x** *db_args*
specifies database-specific arguments. See :ref:`Database Options
<dboptions>` in :ref:`kadmin(1)` for supported arguments.
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kpasswd(1)`, :ref:`kadmin(1)`, :ref:`kdb5_util(8)`,
:ref:`kdb5_ldap_util(8)`, :ref:`kadm5.acl(5)`, :ref:`kerberos(7)`

View file

@ -0,0 +1,449 @@
.. _kdb5_ldap_util(8):
kdb5_ldap_util
===============
SYNOPSIS
--------
.. _kdb5_ldap_util_synopsis:
**kdb5_ldap_util**
[**-D** *user_dn* [**-w** *passwd*]]
[**-H** *ldapuri*]
**command**
[*command_options*]
.. _kdb5_ldap_util_synopsis_end:
DESCRIPTION
-----------
kdb5_ldap_util allows an administrator to manage realms, Kerberos
services and ticket policies.
COMMAND-LINE OPTIONS
--------------------
.. _kdb5_ldap_util_options:
**-r** *realm*
Specifies the realm to be operated on.
**-D** *user_dn*
Specifies the Distinguished Name (DN) of the user who has
sufficient rights to perform the operation on the LDAP server.
**-w** *passwd*
Specifies the password of *user_dn*. This option is not
recommended.
**-H** *ldapuri*
Specifies the URI of the LDAP server.
By default, kdb5_ldap_util operates on the default realm (as specified
in :ref:`krb5.conf(5)`) and connects and authenticates to the LDAP
server in the same manner as :ref:kadmind(8)` would given the
parameters in :ref:`dbdefaults` in :ref:`kdc.conf(5)`.
.. _kdb5_ldap_util_options_end:
COMMANDS
--------
create
~~~~~~
.. _kdb5_ldap_util_create:
**create**
[**-subtrees** *subtree_dn_list*]
[**-sscope** *search_scope*]
[**-containerref** *container_reference_dn*]
[**-k** *mkeytype*]
[**-kv** *mkeyVNO*]
[**-M** *mkeyname*]
[**-m|-P** *password*\|\ **-sf** *stashfilename*]
[**-s**]
[**-maxtktlife** *max_ticket_life*]
[**-maxrenewlife** *max_renewable_ticket_life*]
[*ticket_flags*]
Creates realm in directory. Options:
**-subtrees** *subtree_dn_list*
Specifies the list of subtrees containing the principals of a
realm. The list contains the DNs of the subtree objects separated
by colon (``:``).
**-sscope** *search_scope*
Specifies the scope for searching the principals under the
subtree. The possible values are 1 or one (one level), 2 or sub
(subtrees).
**-containerref** *container_reference_dn*
Specifies the DN of the container object in which the principals
of a realm will be created. If the container reference is not
configured for a realm, the principals will be created in the
realm container.
**-k** *mkeytype*
Specifies the key type of the master key in the database. The
default is given by the **master_key_type** variable in
:ref:`kdc.conf(5)`.
**-kv** *mkeyVNO*
Specifies the version number of the master key in the database;
the default is 1. Note that 0 is not allowed.
**-M** *mkeyname*
Specifies the principal name for the master key in the database.
If not specified, the name is determined by the
**master_key_name** variable in :ref:`kdc.conf(5)`.
**-m**
Specifies that the master database password should be read from
the TTY rather than fetched from a file on the disk.
**-P** *password*
Specifies the master database password. This option is not
recommended.
**-sf** *stashfilename*
Specifies the stash file of the master database password.
**-s**
Specifies that the stash file is to be created.
**-maxtktlife** *max_ticket_life*
(:ref:`getdate` string) Specifies maximum ticket life for
principals in this realm.
**-maxrenewlife** *max_renewable_ticket_life*
(:ref:`getdate` string) Specifies maximum renewable life of
tickets for principals in this realm.
*ticket_flags*
Specifies global ticket flags for the realm. Allowable flags are
documented in the description of the **add_principal** command in
:ref:`kadmin(1)`.
Example::
kdb5_ldap_util -D cn=admin,o=org -H ldaps://ldap-server1.mit.edu
-r ATHENA.MIT.EDU create -subtrees o=org -sscope SUB
Password for "cn=admin,o=org":
Initializing database for realm 'ATHENA.MIT.EDU'
You will be prompted for the database Master Password.
It is important that you NOT FORGET this password.
Enter KDC database master key:
Re-enter KDC database master key to verify:
.. _kdb5_ldap_util_create_end:
modify
~~~~~~
.. _kdb5_ldap_util_modify:
**modify**
[**-subtrees** *subtree_dn_list*]
[**-sscope** *search_scope*]
[**-containerref** *container_reference_dn*]
[**-maxtktlife** *max_ticket_life*]
[**-maxrenewlife** *max_renewable_ticket_life*]
[*ticket_flags*]
Modifies the attributes of a realm. Options:
**-subtrees** *subtree_dn_list*
Specifies the list of subtrees containing the principals of a
realm. The list contains the DNs of the subtree objects separated
by colon (``:``). This list replaces the existing list.
**-sscope** *search_scope*
Specifies the scope for searching the principals under the
subtrees. The possible values are 1 or one (one level), 2 or sub
(subtrees).
**-containerref** *container_reference_dn* Specifies the DN of the
container object in which the principals of a realm will be
created.
**-maxtktlife** *max_ticket_life*
(:ref:`getdate` string) Specifies maximum ticket life for
principals in this realm.
**-maxrenewlife** *max_renewable_ticket_life*
(:ref:`getdate` string) Specifies maximum renewable life of
tickets for principals in this realm.
*ticket_flags*
Specifies global ticket flags for the realm. Allowable flags are
documented in the description of the **add_principal** command in
:ref:`kadmin(1)`.
Example::
shell% kdb5_ldap_util -r ATHENA.MIT.EDU -D cn=admin,o=org -H
ldaps://ldap-server1.mit.edu modify +requires_preauth
Password for "cn=admin,o=org":
shell%
.. _kdb5_ldap_util_modify_end:
view
~~~~
.. _kdb5_ldap_util_view:
**view**
Displays the attributes of a realm.
Example::
kdb5_ldap_util -D cn=admin,o=org -H ldaps://ldap-server1.mit.edu
-r ATHENA.MIT.EDU view
Password for "cn=admin,o=org":
Realm Name: ATHENA.MIT.EDU
Subtree: ou=users,o=org
Subtree: ou=servers,o=org
SearchScope: ONE
Maximum ticket life: 0 days 01:00:00
Maximum renewable life: 0 days 10:00:00
Ticket flags: DISALLOW_FORWARDABLE REQUIRES_PWCHANGE
.. _kdb5_ldap_util_view_end:
destroy
~~~~~~~
.. _kdb5_ldap_util_destroy:
**destroy** [**-f**]
Destroys an existing realm. Options:
**-f**
If specified, will not prompt the user for confirmation.
Example::
shell% kdb5_ldap_util -r ATHENA.MIT.EDU -D cn=admin,o=org -H
ldaps://ldap-server1.mit.edu destroy
Password for "cn=admin,o=org":
Deleting KDC database of 'ATHENA.MIT.EDU', are you sure?
(type 'yes' to confirm)? yes
OK, deleting database of 'ATHENA.MIT.EDU'...
shell%
.. _kdb5_ldap_util_destroy_end:
list
~~~~
.. _kdb5_ldap_util_list:
**list**
Lists the names of realms under the container.
Example::
shell% kdb5_ldap_util -D cn=admin,o=org -H
ldaps://ldap-server1.mit.edu list
Password for "cn=admin,o=org":
ATHENA.MIT.EDU
OPENLDAP.MIT.EDU
MEDIA-LAB.MIT.EDU
shell%
.. _kdb5_ldap_util_list_end:
stashsrvpw
~~~~~~~~~~
.. _kdb5_ldap_util_stashsrvpw:
**stashsrvpw**
[**-f** *filename*]
*name*
Allows an administrator to store the password for service object in a
file so that KDC and Administration server can use it to authenticate
to the LDAP server. Options:
**-f** *filename*
Specifies the complete path of the service password file. By
default, ``/usr/local/var/service_passwd`` is used.
*name*
Specifies the name of the object whose password is to be stored.
If :ref:`krb5kdc(8)` or :ref:`kadmind(8)` are configured for
simple binding, this should be the distinguished name it will
use as given by the **ldap_kdc_dn** or **ldap_kadmind_dn**
variable in :ref:`kdc.conf(5)`. If the KDC or kadmind is
configured for SASL binding, this should be the authentication
name it will use as given by the **ldap_kdc_sasl_authcid** or
**ldap_kadmind_sasl_authcid** variable.
Example::
kdb5_ldap_util stashsrvpw -f /home/andrew/conf_keyfile
cn=service-kdc,o=org
Password for "cn=service-kdc,o=org":
Re-enter password for "cn=service-kdc,o=org":
.. _kdb5_ldap_util_stashsrvpw_end:
create_policy
~~~~~~~~~~~~~
.. _kdb5_ldap_util_create_policy:
**create_policy**
[**-maxtktlife** *max_ticket_life*]
[**-maxrenewlife** *max_renewable_ticket_life*]
[*ticket_flags*]
*policy_name*
Creates a ticket policy in the directory. Options:
**-maxtktlife** *max_ticket_life*
(:ref:`getdate` string) Specifies maximum ticket life for
principals.
**-maxrenewlife** *max_renewable_ticket_life*
(:ref:`getdate` string) Specifies maximum renewable life of
tickets for principals.
*ticket_flags*
Specifies the ticket flags. If this option is not specified, by
default, no restriction will be set by the policy. Allowable
flags are documented in the description of the **add_principal**
command in :ref:`kadmin(1)`.
*policy_name*
Specifies the name of the ticket policy.
Example::
kdb5_ldap_util -D cn=admin,o=org -H ldaps://ldap-server1.mit.edu
-r ATHENA.MIT.EDU create_policy -maxtktlife "1 day"
-maxrenewlife "1 week" -allow_postdated +needchange
-allow_forwardable tktpolicy
Password for "cn=admin,o=org":
.. _kdb5_ldap_util_create_policy_end:
modify_policy
~~~~~~~~~~~~~
.. _kdb5_ldap_util_modify_policy:
**modify_policy**
[**-maxtktlife** *max_ticket_life*]
[**-maxrenewlife** *max_renewable_ticket_life*]
[*ticket_flags*]
*policy_name*
Modifies the attributes of a ticket policy. Options are same as for
**create_policy**.
Example::
kdb5_ldap_util -D cn=admin,o=org -H
ldaps://ldap-server1.mit.edu -r ATHENA.MIT.EDU modify_policy
-maxtktlife "60 minutes" -maxrenewlife "10 hours"
+allow_postdated -requires_preauth tktpolicy
Password for "cn=admin,o=org":
.. _kdb5_ldap_util_modify_policy_end:
view_policy
~~~~~~~~~~~
.. _kdb5_ldap_util_view_policy:
**view_policy**
*policy_name*
Displays the attributes of the named ticket policy.
Example::
kdb5_ldap_util -D cn=admin,o=org -H ldaps://ldap-server1.mit.edu
-r ATHENA.MIT.EDU view_policy tktpolicy
Password for "cn=admin,o=org":
Ticket policy: tktpolicy
Maximum ticket life: 0 days 01:00:00
Maximum renewable life: 0 days 10:00:00
Ticket flags: DISALLOW_FORWARDABLE REQUIRES_PWCHANGE
.. _kdb5_ldap_util_view_policy_end:
destroy_policy
~~~~~~~~~~~~~~
.. _kdb5_ldap_util_destroy_policy:
**destroy_policy**
[**-force**]
*policy_name*
Destroys an existing ticket policy. Options:
**-force**
Forces the deletion of the policy object. If not specified, the
user will be prompted for confirmation before deleting the policy.
*policy_name*
Specifies the name of the ticket policy.
Example::
kdb5_ldap_util -D cn=admin,o=org -H ldaps://ldap-server1.mit.edu
-r ATHENA.MIT.EDU destroy_policy tktpolicy
Password for "cn=admin,o=org":
This will delete the policy object 'tktpolicy', are you sure?
(type 'yes' to confirm)? yes
** policy object 'tktpolicy' deleted.
.. _kdb5_ldap_util_destroy_policy_end:
list_policy
~~~~~~~~~~~
.. _kdb5_ldap_util_list_policy:
**list_policy**
Lists ticket policies.
Example::
kdb5_ldap_util -D cn=admin,o=org -H ldaps://ldap-server1.mit.edu
-r ATHENA.MIT.EDU list_policy
Password for "cn=admin,o=org":
tktpolicy
tmppolicy
userpolicy
.. _kdb5_ldap_util_list_policy_end:
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kadmin(1)`, :ref:`kerberos(7)`

View file

@ -0,0 +1,502 @@
.. _kdb5_util(8):
kdb5_util
=========
SYNOPSIS
--------
.. _kdb5_util_synopsis:
**kdb5_util**
[**-r** *realm*]
[**-d** *dbname*]
[**-k** *mkeytype*]
[**-kv** *mkeyVNO*]
[**-M** *mkeyname*]
[**-m**]
[**-sf** *stashfilename*]
[**-P** *password*]
[**-x** *db_args*]
*command* [*command_options*]
.. _kdb5_util_synopsis_end:
DESCRIPTION
-----------
kdb5_util allows an administrator to perform maintenance procedures on
the KDC database. Databases can be created, destroyed, and dumped to
or loaded from ASCII files. kdb5_util can create a Kerberos master
key stash file or perform live rollover of the master key.
When kdb5_util is run, it attempts to acquire the master key and open
the database. However, execution continues regardless of whether or
not kdb5_util successfully opens the database, because the database
may not exist yet or the stash file may be corrupt.
Note that some KDC database modules may not support all kdb5_util
commands.
COMMAND-LINE OPTIONS
--------------------
.. _kdb5_util_options:
**-r** *realm*
specifies the Kerberos realm of the database.
**-d** *dbname*
specifies the name under which the principal database is stored;
by default the database is that listed in :ref:`kdc.conf(5)`. The
password policy database and lock files are also derived from this
value.
**-k** *mkeytype*
specifies the key type of the master key in the database. The
default is given by the **master_key_type** variable in
:ref:`kdc.conf(5)`.
**-kv** *mkeyVNO*
Specifies the version number of the master key in the database;
the default is 1. Note that 0 is not allowed.
**-M** *mkeyname*
principal name for the master key in the database. If not
specified, the name is determined by the **master_key_name**
variable in :ref:`kdc.conf(5)`.
**-m**
specifies that the master database password should be read from
the keyboard rather than fetched from a file on disk.
**-sf** *stash_file*
specifies the stash filename of the master database password. If
not specified, the filename is determined by the
**key_stash_file** variable in :ref:`kdc.conf(5)`.
**-P** *password*
specifies the master database password. Using this option may
expose the password to other users on the system via the process
list.
**-x** *db_args*
specifies database-specific options. See :ref:`kadmin(1)` for
supported options.
.. _kdb5_util_options_end:
COMMANDS
--------
create
~~~~~~
.. _kdb5_util_create:
**create** [**-s**]
Creates a new database. If the **-s** option is specified, the stash
file is also created. This command fails if the database already
exists. If the command is successful, the database is opened just as
if it had already existed when the program was first run.
.. _kdb5_util_create_end:
destroy
~~~~~~~
.. _kdb5_util_destroy:
**destroy** [**-f**]
Destroys the database, first overwriting the disk sectors and then
unlinking the files, after prompting the user for confirmation. With
the **-f** argument, does not prompt the user.
.. _kdb5_util_destroy_end:
stash
~~~~~
.. _kdb5_util_stash:
**stash** [**-f** *keyfile*]
Stores the master principal's keys in a stash file. The **-f**
argument can be used to override the *keyfile* specified in
:ref:`kdc.conf(5)`.
.. _kdb5_util_stash_end:
dump
~~~~
.. _kdb5_util_dump:
**dump** [**-b7**\|\ **-r13**\|\ **-r18**]
[**-verbose**] [**-mkey_convert**] [**-new_mkey_file**
*mkey_file*] [**-rev**] [**-recurse**] [*filename*
[*principals*...]]
Dumps the current Kerberos and KADM5 database into an ASCII file. By
default, the database is dumped in current format, "kdb5_util
load_dump version 7". If filename is not specified, or is the string
"-", the dump is sent to standard output. Options:
**-b7**
causes the dump to be in the Kerberos 5 Beta 7 format ("kdb5_util
load_dump version 4"). This was the dump format produced on
releases prior to 1.2.2.
**-r13**
causes the dump to be in the Kerberos 5 1.3 format ("kdb5_util
load_dump version 5"). This was the dump format produced on
releases prior to 1.8.
**-r18**
causes the dump to be in the Kerberos 5 1.8 format ("kdb5_util
load_dump version 6"). This was the dump format produced on
releases prior to 1.11.
**-verbose**
causes the name of each principal and policy to be printed as it
is dumped.
**-mkey_convert**
prompts for a new master key. This new master key will be used to
re-encrypt principal key data in the dumpfile. The principal keys
themselves will not be changed.
**-new_mkey_file** *mkey_file*
the filename of a stash file. The master key in this stash file
will be used to re-encrypt the key data in the dumpfile. The key
data in the database will not be changed.
**-rev**
dumps in reverse order. This may recover principals that do not
dump normally, in cases where database corruption has occurred.
**-recurse**
causes the dump to walk the database recursively (btree only).
This may recover principals that do not dump normally, in cases
where database corruption has occurred. In cases of such
corruption, this option will probably retrieve more principals
than the **-rev** option will.
.. versionchanged:: 1.15
Release 1.15 restored the functionality of the **-recurse**
option.
.. versionchanged:: 1.5
The **-recurse** option ceased working until release 1.15,
doing a normal dump instead of a recursive traversal.
.. _kdb5_util_dump_end:
load
~~~~
.. _kdb5_util_load:
**load** [**-b7**\|\ **-r13**\|\ **-r18**] [**-hash**]
[**-verbose**] [**-update**] *filename*
Loads a database dump from the named file into the named database. If
no option is given to determine the format of the dump file, the
format is detected automatically and handled as appropriate. Unless
the **-update** option is given, **load** creates a new database
containing only the data in the dump file, overwriting the contents of
any previously existing database. Note that when using the LDAP KDC
database module, the **-update** flag is required.
Options:
**-b7**
requires the database to be in the Kerberos 5 Beta 7 format
("kdb5_util load_dump version 4"). This was the dump format
produced on releases prior to 1.2.2.
**-r13**
requires the database to be in Kerberos 5 1.3 format ("kdb5_util
load_dump version 5"). This was the dump format produced on
releases prior to 1.8.
**-r18**
requires the database to be in Kerberos 5 1.8 format ("kdb5_util
load_dump version 6"). This was the dump format produced on
releases prior to 1.11.
**-hash**
stores the database in hash format, if using the DB2 database
type. If this option is not specified, the database will be
stored in btree format. This option is not recommended, as
databases stored in hash format are known to corrupt data and lose
principals.
**-verbose**
causes the name of each principal and policy to be printed as it
is dumped.
**-update**
records from the dump file are added to or updated in the existing
database. Otherwise, a new database is created containing only
what is in the dump file and the old one destroyed upon successful
completion.
.. _kdb5_util_load_end:
ark
~~~
**ark** [**-e** *enc*:*salt*,...] *principal*
Adds new random keys to *principal* at the next available key version
number. Keys for the current highest key version number will be
preserved. The **-e** option specifies the list of encryption and
salt types to be used for the new keys.
add_mkey
~~~~~~~~
**add_mkey** [**-e** *etype*] [**-s**]
Adds a new master key to the master key principal, but does not mark
it as active. Existing master keys will remain. The **-e** option
specifies the encryption type of the new master key; see
:ref:`Encryption_types` in :ref:`kdc.conf(5)` for a list of possible
values. The **-s** option stashes the new master key in the stash
file, which will be created if it doesn't already exist.
After a new master key is added, it should be propagated to replica
servers via a manual or periodic invocation of :ref:`kprop(8)`. Then,
the stash files on the replica servers should be updated with the
kdb5_util **stash** command. Once those steps are complete, the key
is ready to be marked active with the kdb5_util **use_mkey** command.
use_mkey
~~~~~~~~
**use_mkey** *mkeyVNO* [*time*]
Sets the activation time of the master key specified by *mkeyVNO*.
Once a master key becomes active, it will be used to encrypt newly
created principal keys. If no *time* argument is given, the current
time is used, causing the specified master key version to become
active immediately. The format for *time* is :ref:`getdate` string.
After a new master key becomes active, the kdb5_util
**update_princ_encryption** command can be used to update all
principal keys to be encrypted in the new master key.
list_mkeys
~~~~~~~~~~
**list_mkeys**
List all master keys, from most recent to earliest, in the master key
principal. The output will show the kvno, enctype, and salt type for
each mkey, similar to the output of :ref:`kadmin(1)` **getprinc**. A
``*`` following an mkey denotes the currently active master key.
purge_mkeys
~~~~~~~~~~~
**purge_mkeys** [**-f**] [**-n**] [**-v**]
Delete master keys from the master key principal that are not used to
protect any principals. This command can be used to remove old master
keys all principal keys are protected by a newer master key.
**-f**
does not prompt for confirmation.
**-n**
performs a dry run, showing master keys that would be purged, but
not actually purging any keys.
**-v**
gives more verbose output.
update_princ_encryption
~~~~~~~~~~~~~~~~~~~~~~~
**update_princ_encryption** [**-f**] [**-n**] [**-v**]
[*princ-pattern*]
Update all principal records (or only those matching the
*princ-pattern* glob pattern) to re-encrypt the key data using the
active database master key, if they are encrypted using a different
version, and give a count at the end of the number of principals
updated. If the **-f** option is not given, ask for confirmation
before starting to make changes. The **-v** option causes each
principal processed to be listed, with an indication as to whether it
needed updating or not. The **-n** option performs a dry run, only
showing the actions which would have been taken.
tabdump
~~~~~~~
**tabdump** [**-H**] [**-c**] [**-e**] [**-n**] [**-o** *outfile*]
*dumptype*
Dump selected fields of the database in a tabular format suitable for
reporting (e.g., using traditional Unix text processing tools) or
importing into relational databases. The data format is tab-separated
(default), or optionally comma-separated (CSV), with a fixed number of
columns. The output begins with a header line containing field names,
unless suppression is requested using the **-H** option.
The *dumptype* parameter specifies the name of an output table (see
below).
Options:
**-H**
suppress writing the field names in a header line
**-c**
use comma separated values (CSV) format, with minimal quoting,
instead of the default tab-separated (unquoted, unescaped) format
**-e**
write empty hexadecimal string fields as empty fields instead of
as "-1".
**-n**
produce numeric output for fields that normally have symbolic
output, such as enctypes and flag names. Also requests output of
time stamps as decimal POSIX time_t values.
**-o** *outfile*
write the dump to the specified output file instead of to standard
output
Dump types:
**keydata**
principal encryption key information, including actual key data
(which is still encrypted in the master key)
**name**
principal name
**keyindex**
index of this key in the principal's key list
**kvno**
key version number
**enctype**
encryption type
**key**
key data as a hexadecimal string
**salttype**
salt type
**salt**
salt data as a hexadecimal string
**keyinfo**
principal encryption key information (as in **keydata** above),
excluding actual key data
**princ_flags**
principal boolean attributes. Flag names print as hexadecimal
numbers if the **-n** option is specified, and all flag positions
are printed regardless of whether or not they are set. If **-n**
is not specified, print all known flag names for each principal,
but only print hexadecimal flag names if the corresponding flag is
set.
**name**
principal name
**flag**
flag name
**value**
boolean value (0 for clear, or 1 for set)
**princ_lockout**
state information used for tracking repeated password failures
**name**
principal name
**last_success**
time stamp of most recent successful authentication
**last_failed**
time stamp of most recent failed authentication
**fail_count**
count of failed attempts
**princ_meta**
principal metadata
**name**
principal name
**modby**
name of last principal to modify this principal
**modtime**
timestamp of last modification
**lastpwd**
timestamp of last password change
**policy**
policy object name
**mkvno**
key version number of the master key that encrypts this
principal's key data
**hist_kvno**
key version number of the history key that encrypts the key
history data for this principal
**princ_stringattrs**
string attributes (key/value pairs)
**name**
principal name
**key**
attribute name
**value**
attribute value
**princ_tktpolicy**
per-principal ticket policy data, including maximum ticket
lifetimes
**name**
principal name
**expiration**
principal expiration date
**pw_expiration**
password expiration date
**max_life**
maximum ticket lifetime
**max_renew_life**
maximum renewable ticket lifetime
Examples::
$ kdb5_util tabdump -o keyinfo.txt keyinfo
$ cat keyinfo.txt
name keyindex kvno enctype salttype salt
K/M@EXAMPLE.COM 0 1 aes256-cts-hmac-sha384-192 normal -1
foo@EXAMPLE.COM 0 1 aes128-cts-hmac-sha1-96 normal -1
bar@EXAMPLE.COM 0 1 aes128-cts-hmac-sha1-96 normal -1
$ sqlite3
sqlite> .mode tabs
sqlite> .import keyinfo.txt keyinfo
sqlite> select * from keyinfo where enctype like 'aes256-%';
K/M@EXAMPLE.COM 1 1 aes256-cts-hmac-sha384-192 normal -1
sqlite> .quit
$ awk -F'\t' '$4 ~ /aes256-/ { print }' keyinfo.txt
K/M@EXAMPLE.COM 1 1 aes256-cts-hmac-sha384-192 normal -1
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kadmin(1)`, :ref:`kerberos(7)`

View file

@ -0,0 +1,60 @@
.. _kprop(8):
kprop
=====
SYNOPSIS
--------
**kprop**
[**-r** *realm*]
[**-f** *file*]
[**-d**]
[**-P** *port*]
[**-s** *keytab*]
*replica_host*
DESCRIPTION
-----------
kprop is used to securely propagate a Kerberos V5 database dump file
from the primary Kerberos server to a replica Kerberos server, which is
specified by *replica_host*. The dump file must be created by
:ref:`kdb5_util(8)`.
OPTIONS
-------
**-r** *realm*
Specifies the realm of the primary server.
**-f** *file*
Specifies the filename where the dumped principal database file is
to be found; by default the dumped database file is normally
|kdcdir|\ ``/replica_datatrans``.
**-P** *port*
Specifies the port to use to contact the :ref:`kpropd(8)` server
on the remote host.
**-d**
Prints debugging information.
**-s** *keytab*
Specifies the location of the keytab file.
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kpropd(8)`, :ref:`kdb5_util(8)`, :ref:`krb5kdc(8)`,
:ref:`kerberos(7)`

View file

@ -0,0 +1,144 @@
.. _kpropd(8):
kpropd
======
SYNOPSIS
--------
**kpropd**
[**-r** *realm*]
[**-A** *admin_server*]
[**-a** *acl_file*]
[**-f** *replica_dumpfile*]
[**-F** *principal_database*]
[**-p** *kdb5_util_prog*]
[**-P** *port*]
[**--pid-file**\ =\ *pid_file*]
[**-D**]
[**-d**]
[**-s** *keytab_file*]
DESCRIPTION
-----------
The *kpropd* command runs on the replica KDC server. It listens for
update requests made by the :ref:`kprop(8)` program. If incremental
propagation is enabled, it periodically requests incremental updates
from the primary KDC.
When the replica receives a kprop request from the primary, kpropd
accepts the dumped KDC database and places it in a file, and then runs
:ref:`kdb5_util(8)` to load the dumped database into the active
database which is used by :ref:`krb5kdc(8)`. This allows the primary
Kerberos server to use :ref:`kprop(8)` to propagate its database to
the replica servers. Upon a successful download of the KDC database
file, the replica Kerberos server will have an up-to-date KDC
database.
Where incremental propagation is not used, kpropd is commonly invoked
out of inetd(8) as a nowait service. This is done by adding a line to
the ``/etc/inetd.conf`` file which looks like this::
kprop stream tcp nowait root /usr/local/sbin/kpropd kpropd
kpropd can also run as a standalone daemon, backgrounding itself and
waiting for connections on port 754 (or the port specified with the
**-P** option if given). Standalone mode is required for incremental
propagation. Starting in release 1.11, kpropd automatically detects
whether it was run from inetd and runs in standalone mode if it is
not. Prior to release 1.11, the **-S** option is required to run
kpropd in standalone mode; this option is now accepted for backward
compatibility but does nothing.
Incremental propagation may be enabled with the **iprop_enable**
variable in :ref:`kdc.conf(5)`. If incremental propagation is
enabled, the replica periodically polls the primary KDC for updates, at
an interval determined by the **iprop_replica_poll** variable. If the
replica receives updates, kpropd updates its log file with any updates
from the primary. :ref:`kproplog(8)` can be used to view a summary of
the update entry log on the replica KDC. If incremental propagation
is enabled, the principal ``kiprop/replicahostname@REALM`` (where
*replicahostname* is the name of the replica KDC host, and *REALM* is
the name of the Kerberos realm) must be present in the replica's
keytab file.
:ref:`kproplog(8)` can be used to force full replication when iprop is
enabled.
OPTIONS
--------
**-r** *realm*
Specifies the realm of the primary server.
**-A** *admin_server*
Specifies the server to be contacted for incremental updates; by
default, the primary admin server is contacted.
**-f** *file*
Specifies the filename where the dumped principal database file is
to be stored; by default the dumped database file is |kdcdir|\
``/from_master``.
**-F** *kerberos_db*
Path to the Kerberos database file, if not the default.
**-p**
Allows the user to specify the pathname to the :ref:`kdb5_util(8)`
program; by default the pathname used is |sbindir|\
``/kdb5_util``.
**-D**
In this mode, kpropd will not detach itself from the current job
and run in the background. Instead, it will run in the
foreground.
**-d**
Turn on debug mode. kpropd will print out debugging messages
during the database propogation and will run in the foreground
(implies **-D**).
**-P**
Allow for an alternate port number for kpropd to listen on. This
is only useful in combination with the **-S** option.
**-a** *acl_file*
Allows the user to specify the path to the kpropd.acl file; by
default the path used is |kdcdir|\ ``/kpropd.acl``.
**--pid-file**\ =\ *pid_file*
In standalone mode, write the process ID of the daemon into
*pid_file*.
**-s** *keytab_file*
Path to a keytab to use for acquiring acceptor credentials.
**-x** *db_args*
Database-specific arguments. See :ref:`Database Options
<dboptions>` in :ref:`kadmin(1)` for supported arguments.
FILES
-----
kpropd.acl
Access file for kpropd; the default location is
``/usr/local/var/krb5kdc/kpropd.acl``. Each entry is a line
containing the principal of a host from which the local machine
will allow Kerberos database propagation via :ref:`kprop(8)`.
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kprop(8)`, :ref:`kdb5_util(8)`, :ref:`krb5kdc(8)`,
:ref:`kerberos(7)`, inetd(8)

View file

@ -0,0 +1,85 @@
.. _kproplog(8):
kproplog
========
SYNOPSIS
--------
**kproplog** [**-h**] [**-e** *num*] [-v]
**kproplog** [-R]
DESCRIPTION
-----------
The kproplog command displays the contents of the KDC database update
log to standard output. It can be used to keep track of incremental
updates to the principal database. The update log file contains the
update log maintained by the :ref:`kadmind(8)` process on the primary
KDC server and the :ref:`kpropd(8)` process on the replica KDC
servers. When updates occur, they are logged to this file.
Subsequently any KDC replica configured for incremental updates will
request the current data from the primary KDC and update their log
file with any updates returned.
The kproplog command requires read access to the update log file. It
will display update entries only for the KDC it runs on.
If no options are specified, kproplog displays a summary of the update
log. If invoked on the primary, kproplog also displays all of the
update entries. If invoked on a replica KDC server, kproplog displays
only a summary of the updates, which includes the serial number of the
last update received and the associated time stamp of the last update.
OPTIONS
-------
**-R**
Reset the update log. This forces full resynchronization. If
used on a replica then that replica will request a full resync.
If used on the primary then all replicas will request full
resyncs.
**-h**
Display a summary of the update log. This information includes
the database version number, state of the database, the number of
updates in the log, the time stamp of the first and last update,
and the version number of the first and last update entry.
**-e** *num*
Display the last *num* update entries in the log. This is useful
when debugging synchronization between KDC servers.
**-v**
Display individual attributes per update. An example of the
output generated for one entry::
Update Entry
Update serial # : 4
Update operation : Add
Update principal : test@EXAMPLE.COM
Update size : 424
Update committed : True
Update time stamp : Fri Feb 20 23:37:42 2004
Attributes changed : 6
Principal
Key data
Password last changed
Modifying principal
Modification time
TL data
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kpropd(8)`, :ref:`kerberos(7)`

View file

@ -0,0 +1,114 @@
.. _krb5kdc(8):
krb5kdc
=======
SYNOPSIS
--------
**krb5kdc**
[**-x** *db_args*]
[**-d** *dbname*]
[**-k** *keytype*]
[**-M** *mkeyname*]
[**-p** *portnum*]
[**-m**]
[**-r** *realm*]
[**-n**]
[**-w** *numworkers*]
[**-P** *pid_file*]
[**-T** *time_offset*]
DESCRIPTION
-----------
krb5kdc is the Kerberos version 5 Authentication Service and Key
Distribution Center (AS/KDC).
OPTIONS
-------
The **-r** *realm* option specifies the realm for which the server
should provide service. This option may be specified multiple times
to serve multiple realms. If no **-r** option is given, the default
realm (as specified in :ref:`krb5.conf(5)`) will be served.
The **-d** *dbname* option specifies the name under which the
principal database can be found. This option does not apply to the
LDAP database.
The **-k** *keytype* option specifies the key type of the master key
to be entered manually as a password when **-m** is given; the default
is |defmkey|.
The **-M** *mkeyname* option specifies the principal name for the
master key in the database (usually ``K/M`` in the KDC's realm).
The **-m** option specifies that the master database password should
be fetched from the keyboard rather than from a stash file.
The **-n** option specifies that the KDC does not put itself in the
background and does not disassociate itself from the terminal.
The **-P** *pid_file* option tells the KDC to write its PID into
*pid_file* after it starts up. This can be used to identify whether
the KDC is still running and to allow init scripts to stop the correct
process.
The **-p** *portnum* option specifies the default UDP and TCP port
numbers which the KDC should listen on for Kerberos version 5
requests, as a comma-separated list. This value overrides the port
numbers specified in the :ref:`kdcdefaults` section of
:ref:`kdc.conf(5)`, but may be overridden by realm-specific values.
If no value is given from any source, the default port is 88.
The **-w** *numworkers* option tells the KDC to fork *numworkers*
processes to listen to the KDC ports and process requests in parallel.
The top level KDC process (whose pid is recorded in the pid file if
the **-P** option is also given) acts as a supervisor. The supervisor
will relay SIGHUP signals to the worker subprocesses, and will
terminate the worker subprocess if the it is itself terminated or if
any other worker process exits.
The **-x** *db_args* option specifies database-specific arguments.
See :ref:`Database Options <dboptions>` in :ref:`kadmin(1)` for
supported arguments.
The **-T** *offset* option specifies a time offset, in seconds, which
the KDC will operate under. It is intended only for testing purposes.
EXAMPLE
-------
The KDC may service requests for multiple realms (maximum 32 realms).
The realms are listed on the command line. Per-realm options that can
be specified on the command line pertain for each realm that follows
it and are superseded by subsequent definitions of the same option.
For example::
krb5kdc -p 2001 -r REALM1 -p 2002 -r REALM2 -r REALM3
specifies that the KDC listen on port 2001 for REALM1 and on port 2002
for REALM2 and REALM3. Additionally, per-realm parameters may be
specified in the :ref:`kdc.conf(5)` file. The location of this file
may be specified by the **KRB5_KDC_PROFILE** environment variable.
Per-realm parameters specified in this file take precedence over
options specified on the command line. See the :ref:`kdc.conf(5)`
description for further details.
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kdb5_util(8)`, :ref:`kdc.conf(5)`, :ref:`krb5.conf(5)`,
:ref:`kdb5_ldap_util(8)`, :ref:`kerberos(7)`

View file

@ -0,0 +1,129 @@
.. _ktutil(1):
ktutil
======
SYNOPSIS
--------
**ktutil**
DESCRIPTION
-----------
The ktutil command invokes a command interface from which an
administrator can read, write, or edit entries in a keytab. (Kerberos
V4 srvtab files are no longer supported.)
COMMANDS
--------
list
~~~~
**list** [**-t**] [**-k**] [**-e**]
Displays the current keylist. If **-t**, **-k**, and/or **-e** are
specified, also display the timestamp, key contents, or enctype
(respectively).
Alias: **l**
read_kt
~~~~~~~
**read_kt** *keytab*
Read the Kerberos V5 keytab file *keytab* into the current keylist.
Alias: **rkt**
write_kt
~~~~~~~~
**write_kt** *keytab*
Write the current keylist into the Kerberos V5 keytab file *keytab*.
Alias: **wkt**
clear_list
~~~~~~~~~~
**clear_list**
Clear the current keylist.
Alias: **clear**
delete_entry
~~~~~~~~~~~~
**delete_entry** *slot*
Delete the entry in slot number *slot* from the current keylist.
Alias: **delent**
add_entry
~~~~~~~~~
**add_entry** {**-key**\|\ **-password**} **-p** *principal*
**-k** *kvno* [**-e** *enctype*] [**-f**\|\ **-s** *salt*]
Add *principal* to keylist using key or password. If the **-f** flag
is specified, salt information will be fetched from the KDC; in this
case the **-e** flag may be omitted, or it may be supplied to force a
particular enctype. If the **-f** flag is not specified, the **-e**
flag must be specified, and the default salt will be used unless
overridden with the **-s** option.
Alias: **addent**
list_requests
~~~~~~~~~~~~~
**list_requests**
Displays a listing of available commands.
Aliases: **lr**, **?**
quit
~~~~
**quit**
Quits ktutil.
Aliases: **exit**, **q**
EXAMPLE
-------
::
ktutil: add_entry -password -p alice@BLEEP.COM -k 1 -e
aes128-cts-hmac-sha1-96
Password for alice@BLEEP.COM:
ktutil: add_entry -password -p alice@BLEEP.COM -k 1 -e
aes256-cts-hmac-sha1-96
Password for alice@BLEEP.COM:
ktutil: write_kt alice.keytab
ktutil:
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kadmin(1)`, :ref:`kdb5_util(8)`, :ref:`kerberos(7)`

View file

@ -0,0 +1,112 @@
.. _sserver(8):
sserver
=======
SYNOPSIS
--------
**sserver**
[ **-p** *port* ]
[ **-S** *keytab* ]
[ *server_port* ]
DESCRIPTION
-----------
sserver and :ref:`sclient(1)` are a simple demonstration client/server
application. When sclient connects to sserver, it performs a Kerberos
authentication, and then sserver returns to sclient the Kerberos
principal which was used for the Kerberos authentication. It makes a
good test that Kerberos has been successfully installed on a machine.
The service name used by sserver and sclient is sample. Hence,
sserver will require that there be a keytab entry for the service
``sample/hostname.domain.name@REALM.NAME``. This keytab is generated
using the :ref:`kadmin(1)` program. The keytab file is usually
installed as |keytab|.
The **-S** option allows for a different keytab than the default.
sserver is normally invoked out of inetd(8), using a line in
``/etc/inetd.conf`` that looks like this::
sample stream tcp nowait root /usr/local/sbin/sserver sserver
Since ``sample`` is normally not a port defined in ``/etc/services``,
you will usually have to add a line to ``/etc/services`` which looks
like this::
sample 13135/tcp
When using sclient, you will first have to have an entry in the
Kerberos database, by using :ref:`kadmin(1)`, and then you have to get
Kerberos tickets, by using :ref:`kinit(1)`. Also, if you are running
the sclient program on a different host than the sserver it will be
connecting to, be sure that both hosts have an entry in /etc/services
for the sample tcp port, and that the same port number is in both
files.
When you run sclient you should see something like this::
sendauth succeeded, reply is:
reply len 32, contents:
You are nlgilman@JIMI.MIT.EDU
COMMON ERROR MESSAGES
---------------------
1) kinit returns the error::
kinit: Client not found in Kerberos database while getting
initial credentials
This means that you didn't create an entry for your username in the
Kerberos database.
2) sclient returns the error::
unknown service sample/tcp; check /etc/services
This means that you don't have an entry in /etc/services for the
sample tcp port.
3) sclient returns the error::
connect: Connection refused
This probably means you didn't edit /etc/inetd.conf correctly, or
you didn't restart inetd after editing inetd.conf.
4) sclient returns the error::
sclient: Server not found in Kerberos database while using
sendauth
This means that the ``sample/hostname@LOCAL.REALM`` service was not
defined in the Kerberos database; it should be created using
:ref:`kadmin(1)`, and a keytab file needs to be generated to make
the key for that service principal available for sclient.
5) sclient returns the error::
sendauth rejected, error reply is:
"No such file or directory"
This probably means sserver couldn't find the keytab file. It was
probably not installed in the proper directory.
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`sclient(1)`, :ref:`kerberos(7)`, services(5), inetd(8)

View file

@ -0,0 +1,8 @@
Advanced topics
===============
.. toctree::
:maxdepth: 1
retiring-des.rst

View file

@ -0,0 +1,422 @@
.. _retiring-des:
Retiring DES
=======================
Version 5 of the Kerberos protocol was originally implemented using
the Data Encryption Standard (DES) as a block cipher for encryption.
While it was considered secure at the time, advancements in computational
ability have rendered DES vulnerable to brute force attacks on its 56-bit
keyspace. As such, it is now considered insecure and should not be
used (:rfc:`6649`).
History
-------
DES was used in the original Kerberos implementation, and was the
only cryptosystem in krb5 1.0. Partial support for triple-DES (3DES) was
added in version 1.1, with full support following in version 1.2.
The Advanced Encryption Standard (AES), which supersedes DES, gained
partial support in version 1.3.0 of krb5 and full support in version 1.3.2.
However, deployments of krb5 using Kerberos databases created with older
versions of krb5 will not necessarily start using strong crypto for
ordinary operation without administrator intervention.
MIT krb5 began flagging deprecated encryption types with release 1.17,
and removed DES (single-DES) support in release 1.18. As a
consequence, a release prior to 1.18 is required to perform these
migrations.
Types of keys
-------------
* The database master key: This key is not exposed to user requests,
but is used to encrypt other key material stored in the kerberos
database. The database master key is currently stored as ``K/M``
by default.
* Password-derived keys: User principals frequently have keys
derived from a password. When a new password is set, the KDC
uses various string2key functions to generate keys in the database
for that principal.
* Keytab keys: Application server principals generally use random
keys which are not derived from a password. When the database
entry is created, the KDC generates random keys of various enctypes
to enter in the database, which are conveyed to the application server
and stored in a keytab.
* Session keys: These are short-term keys generated by the KDC while
processing client requests, with an enctype selected by the KDC.
For details on the various enctypes and how enctypes are selected by the KDC
for session keys and client/server long-term keys, see :ref:`enctypes`.
When using the :ref:`kadmin(1)` interface to generate new long-term keys,
the **-e** argument can be used to force a particular set of enctypes,
overriding the KDC default values.
.. note::
When the KDC is selecting a session key, it has no knowledge about the
kerberos installation on the server which will receive the service ticket,
only what keys are in the database for the service principal.
In order to allow uninterrupted operation to
clients while migrating away from DES, care must be taken to ensure that
kerberos installations on application server machines are configured to
support newer encryption types before keys of those new encryption types
are created in the Kerberos database for those server principals.
Upgrade procedure
-----------------
This procedure assumes that the KDC software has already been upgraded
to a modern version of krb5 that supports non-DES keys, so that the
only remaining task is to update the actual keys used to service requests.
The realm used for demonstrating this procedure, ZONE.MIT.EDU,
is an example of the worst-case scenario, where all keys in the realm
are DES. The realm was initially created with a very old version of krb5,
and **supported_enctypes** in :ref:`kdc.conf(5)` was set to a value
appropriate when the KDC was installed, but was not updated as the KDC
was upgraded:
::
[realms]
ZONE.MIT.EDU = {
[...]
master_key_type = des-cbc-crc
supported_enctypes = des-cbc-crc:normal des:normal des:v4 des:norealm des:onlyrealm des:afs3
}
This resulted in the keys for all principals in the realm being forced
to DES-only, unless specifically requested using :ref:`kadmin(1)`.
Before starting the upgrade, all KDCs were running krb5 1.11,
and the database entries for some "high-value" principals were:
::
[root@casio krb5kdc]# kadmin.local -r ZONE.MIT.EDU -q 'getprinc krbtgt/ZONE.MIT.EDU'
[...]
Number of keys: 1
Key: vno 1, des-cbc-crc:v4
[...]
[root@casio krb5kdc]# kadmin.local -r ZONE.MIT.EDU -q 'getprinc kadmin/admin'
[...]
Number of keys: 1
Key: vno 15, des-cbc-crc
[...]
[root@casio krb5kdc]# kadmin.local -r ZONE.MIT.EDU -q 'getprinc kadmin/changepw'
[...]
Number of keys: 1
Key: vno 14, des-cbc-crc
[...]
The ``krbtgt/REALM`` key appears to have never been changed since creation
(its kvno is 1), and all three database entries have only a des-cbc-crc key.
The krbtgt key and KDC keys
~~~~~~~~~~~~~~~~~~~~~~~~~~~
Perhaps the biggest single-step improvement in the security of the cell
is gained by strengthening the key of the ticket-granting service principal,
``krbtgt/REALM``---if this principal's key is compromised, so is the
entire realm. Since the server that will handle service tickets
for this principal is the KDC itself, it is easy to guarantee that it
will be configured to support any encryption types which might be
selected. However, the default KDC behavior when creating new keys is to
remove the old keys, which would invalidate all existing tickets issued
against that principal, rendering the TGTs cached by clients useless.
Instead, a new key can be created with the old key retained, so that
existing tickets will still function until their scheduled expiry
(see :ref:`changing_krbtgt_key`).
::
[root@casio krb5kdc]# enctypes=aes256-cts-hmac-sha1-96:normal,\
> aes128-cts-hmac-sha1-96:normal,des3-hmac-sha1:normal,des-cbc-crc:normal
[root@casio krb5kdc]# kadmin.local -r ZONE.MIT.EDU -q "cpw -e ${enctypes} -randkey \
> -keepold krbtgt/ZONE.MIT.EDU"
Authenticating as principal root/admin@ZONE.MIT.EDU with password.
Key for "krbtgt/ZONE.MIT.EDU@ZONE.MIT.EDU" randomized.
.. note::
The new ``krbtgt@REALM`` key should be propagated to replica KDCs
immediately so that TGTs issued by the primary KDC can be used to
issue service tickets on replica KDCs. Replica KDCs will refuse
requests using the new TGT kvno until the new krbtgt entry has
been propagated to them.
It is necessary to explicitly specify the enctypes for the new database
entry, since **supported_enctypes** has not been changed. Leaving
**supported_enctypes** unchanged makes a potential rollback operation
easier, since all new keys of new enctypes are the result of explicit
administrator action and can be easily enumerated.
Upgrading the krbtgt key should have minimal user-visible disruption other
than that described in the note above, since only clients which list the
new enctypes as supported will use them, per the procedure
in :ref:`session_key_selection`.
Once the krbtgt key is updated, the session and ticket keys for user
TGTs will be strong keys, but subsequent requests
for service tickets will still get DES keys until the service principals
have new keys generated. Application service
remains uninterrupted due to the key-selection procedure on the KDC.
After the change, the database entry is now:
::
[root@casio krb5kdc]# kadmin.local -r ZONE.MIT.EDU -q 'getprinc krbtgt/ZONE.MIT.EDU'
[...]
Number of keys: 5
Key: vno 2, aes256-cts-hmac-sha1-96
Key: vno 2, aes128-cts-hmac-sha1-96
Key: vno 2, des3-cbc-sha1
Key: vno 2, des-cbc-crc
Key: vno 1, des-cbc-crc:v4
[...]
Since the expected disruptions from rekeying the krbtgt principal are
minor, after a short testing period, it is
appropriate to rekey the other high-value principals, ``kadmin/admin@REALM``
and ``kadmin/changepw@REALM``. These are the service principals used for
changing user passwords and updating application keytabs. The kadmin
and password-changing services are regular kerberized services, so the
session-key-selection algorithm described in :ref:`session_key_selection`
applies. It is particularly important to have strong session keys for
these services, since user passwords and new long-term keys are conveyed
over the encrypted channel.
::
[root@casio krb5kdc]# enctypes=aes256-cts-hmac-sha1-96:normal,\
> aes128-cts-hmac-sha1-96:normal,des3-hmac-sha1:normal
[root@casio krb5kdc]# kadmin.local -r ZONE.MIT.EDU -q "cpw -e ${enctypes} -randkey \
> kadmin/admin"
Authenticating as principal root/admin@ZONE.MIT.EDU with password.
Key for "kadmin/admin@ZONE.MIT.EDU" randomized.
[root@casio krb5kdc]# kadmin.local -r ZONE.MIT.EDU -q "cpw -e ${enctypes} -randkey \
> kadmin/changepw"
Authenticating as principal root/admin@ZONE.MIT.EDU with password.
Key for "kadmin/changepw@ZONE.MIT.EDU" randomized.
It is not necessary to retain a single-DES key for these services, since
password changes are not part of normal daily workflow, and disruption
from a client failure is likely to be minimal. Furthermore, if a kerberos
client experiences failure changing a user password or keytab key,
this indicates that that client will become inoperative once services
are rekeyed to non-DES enctypes. Such problems can be detected early
at this stage, giving more time for corrective action.
Adding strong keys to application servers
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Before switching the default enctypes for new keys over to strong enctypes,
it may be desired to test upgrading a handful of services with the
new configuration before flipping the switch for the defaults. This
still requires using the **-e** argument in :ref:`kadmin(1)` to get non-default
enctypes:
::
[root@casio krb5kdc]# enctypes=aes256-cts-hmac-sha1-96:normal,\
> aes128-cts-hmac-sha1-96:normal,des3-cbc-sha1:normal,des-cbc-crc:normal
[root@casio krb5kdc]# kadmin -r ZONE.MIT.EDU -p zephyr/zephyr@ZONE.MIT.EDU -k -t \
> /etc/zephyr/krb5.keytab -q "ktadd -e ${enctypes} \
> -k /etc/zephyr/krb5.keytab zephyr/zephyr@ZONE.MIT.EDU"
Authenticating as principal zephyr/zephyr@ZONE.MIT.EDU with keytab /etc/zephyr/krb5.keytab.
Entry for principal zephyr/zephyr@ZONE.MIT.EDU with kvno 4, encryption type aes256-cts-hmac-sha1-96 added to keytab WRFILE:/etc/zephyr/krb5.keytab.
Entry for principal zephyr/zephyr@ZONE.MIT.EDU with kvno 4, encryption type aes128-cts-hmac-sha1-96 added to keytab WRFILE:/etc/zephyr/krb5.keytab.
Entry for principal zephyr/zephyr@ZONE.MIT.EDU with kvno 4, encryption type des3-cbc-sha1 added to keytab WRFILE:/etc/zephyr/krb5.keytab.
Entry for principal zephyr/zephyr@ZONE.MIT.EDU with kvno 4, encryption type des-cbc-crc added to keytab WRFILE:/etc/zephyr/krb5.keytab.
Be sure to remove the old keys from the application keytab, per best
practice.
::
[root@casio krb5kdc]# k5srvutil -f /etc/zephyr/krb5.keytab delold
Authenticating as principal zephyr/zephyr@ZONE.MIT.EDU with keytab /etc/zephyr/krb5.keytab.
Entry for principal zephyr/zephyr@ZONE.MIT.EDU with kvno 3 removed from keytab WRFILE:/etc/zephyr/krb5.keytab.
Adding strong keys by default
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Once the high-visibility services have been rekeyed, it is probably
appropriate to change :ref:`kdc.conf(5)` to generate keys with the new
encryption types by default. This enables server administrators to generate
new enctypes with the **change** subcommand of :ref:`k5srvutil(1)`,
and causes user password
changes to add new encryption types for their entries. It will probably
be necessary to implement administrative controls to cause all user
principal keys to be updated in a reasonable period of time, whether
by forcing password changes or a password synchronization service that
has access to the current password and can add the new keys.
::
[realms]
ZONE.MIT.EDU = {
supported_enctypes = aes256-cts-hmac-sha1-96:normal aes128-cts-hmac-sha1-96:normal des3-cbc-sha1:normal des3-hmac-sha1:normal des-cbc-crc:normal
.. note::
The krb5kdc process must be restarted for these changes to take effect.
At this point, all service administrators can update their services and the
servers behind them to take advantage of strong cryptography.
If necessary, the server's krb5 installation should be configured and/or
upgraded to a version supporting non-DES keys. See :ref:`enctypes` for
krb5 version and configuration settings.
Only when the service is configured to accept non-DES keys should
the key version number be incremented and new keys generated
(``k5srvutil change && k5srvutil delold``).
::
root@dr-willy:~# k5srvutil change
Authenticating as principal host/dr-willy.xvm.mit.edu@ZONE.MIT.EDU with keytab /etc/krb5.keytab.
Entry for principal host/dr-willy.xvm.mit.edu@ZONE.MIT.EDU with kvno 3, encryption type AES-256 CTS mode with 96-bit SHA-1 HMAC added to keytab WRFILE:/etc/krb5.keytab.
Entry for principal host/dr-willy.xvm.mit.edu@ZONE.MIT.EDU with kvno 3, encryption type AES-128 CTS mode with 96-bit SHA-1 HMAC added to keytab WRFILE:/etc/krb5.keytab.
Entry for principal host/dr-willy.xvm.mit.edu@ZONE.MIT.EDU with kvno 3, encryption type Triple DES cbc mode with HMAC/sha1 added to keytab WRFILE:/etc/krb5.keytab.
Entry for principal host/dr-willy.xvm.mit.edu@ZONE.MIT.EDU with kvno 3, encryption type DES cbc mode with CRC-32 added to keytab WRFILE:/etc/krb5.keytab.
root@dr-willy:~# klist -e -k -t /etc/krb5.keytab
Keytab name: WRFILE:/etc/krb5.keytab
KVNO Timestamp Principal
---- ----------------- --------------------------------------------------------
2 10/10/12 17:03:59 host/dr-willy.xvm.mit.edu@ZONE.MIT.EDU (DES cbc mode with CRC-32)
3 12/12/12 15:31:19 host/dr-willy.xvm.mit.edu@ZONE.MIT.EDU (AES-256 CTS mode with 96-bit SHA-1 HMAC)
3 12/12/12 15:31:19 host/dr-willy.xvm.mit.edu@ZONE.MIT.EDU (AES-128 CTS mode with 96-bit SHA-1 HMAC)
3 12/12/12 15:31:19 host/dr-willy.xvm.mit.edu@ZONE.MIT.EDU (Triple DES cbc mode with HMAC/sha1)
3 12/12/12 15:31:19 host/dr-willy.xvm.mit.edu@ZONE.MIT.EDU (DES cbc mode with CRC-32)
root@dr-willy:~# k5srvutil delold
Authenticating as principal host/dr-willy.xvm.mit.edu@ZONE.MIT.EDU with keytab /etc/krb5.keytab.
Entry for principal host/dr-willy.xvm.mit.edu@ZONE.MIT.EDU with kvno 2 removed from keytab WRFILE:/etc/krb5.keytab.
When a single service principal is shared by multiple backend servers in
a load-balanced environment, it may be necessary to schedule downtime
or adjust the population in the load-balanced pool in order to propagate
the updated keytab to all hosts in the pool with minimal service interruption.
Removing DES keys from usage
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
This situation remains something of a testing or transitory state,
as new DES keys are still being generated, and will be used if requested
by a client. To make more progress removing DES from the realm, the KDC
should be configured to not generate such keys by default.
.. note::
An attacker posing as a client can implement a brute force attack against
a DES key for any principal, if that key is in the current (highest-kvno)
key list. This attack is only possible if **allow_weak_crypto = true**
is enabled on the KDC. Setting the **+requires_preauth** flag on a
principal forces this attack to be an online attack, much slower than
the offline attack otherwise available to the attacker. However, setting
this flag on a service principal is not always advisable; see the entry in
:ref:`add_principal` for details.
The following KDC configuration will not generate DES keys by default:
::
[realms]
ZONE.MIT.EDU = {
supported_enctypes = aes256-cts-hmac-sha1-96:normal aes128-cts-hmac-sha1-96:normal des3-cbc-sha1:normal des3-hmac-sha1:normal
.. note::
As before, the KDC process must be restarted for this change to take
effect. It is best practice to update kdc.conf on all KDCs, not just the
primary, to avoid unpleasant surprises should the primary fail and a
replica need to be promoted.
It is now appropriate to remove the legacy single-DES key from the
``krbtgt/REALM`` entry:
::
[root@casio krb5kdc]# kadmin.local -r ZONE.MIT.EDU -q "cpw -randkey -keepold \
> krbtgt/ZONE.MIT.EDU"
Authenticating as principal host/admin@ATHENA.MIT.EDU with password.
Key for "krbtgt/ZONE.MIT.EDU@ZONE.MIT.EDU" randomized.
After the maximum ticket lifetime has passed, the old database entry
should be removed.
::
[root@casio krb5kdc]# kadmin.local -r ZONE.MIT.EDU -q 'purgekeys krbtgt/ZONE.MIT.EDU'
Authenticating as principal root/admin@ZONE.MIT.EDU with password.
Old keys for principal "krbtgt/ZONE.MIT.EDU@ZONE.MIT.EDU" purged.
After the KDC is restarted with the new **supported_enctypes**,
all user password changes and application keytab updates will not
generate DES keys by default.
::
contents-vnder-pressvre:~> kpasswd zonetest@ZONE.MIT.EDU
Password for zonetest@ZONE.MIT.EDU: [enter old password]
Enter new password: [enter new password]
Enter it again: [enter new password]
Password changed.
contents-vnder-pressvre:~> kadmin -r ZONE.MIT.EDU -q 'getprinc zonetest'
[...]
Number of keys: 3
Key: vno 9, aes256-cts-hmac-sha1-96
Key: vno 9, aes128-cts-hmac-sha1-96
Key: vno 9, des3-cbc-sha1
[...]
[kaduk@glossolalia ~]$ kadmin -p kaduk@ZONE.MIT.EDU -r ZONE.MIT.EDU -k \
> -t kaduk-zone.keytab -q 'ktadd -k kaduk-zone.keytab kaduk@ZONE.MIT.EDU'
Authenticating as principal kaduk@ZONE.MIT.EDU with keytab kaduk-zone.keytab.
Entry for principal kaduk@ZONE.MIT.EDU with kvno 3, encryption type aes256-cts-hmac-sha1-96 added to keytab WRFILE:kaduk-zone.keytab.
Entry for principal kaduk@ZONE.MIT.EDU with kvno 3, encryption type aes128-cts-hmac-sha1-96 added to keytab WRFILE:kaduk-zone.keytab.
Entry for principal kaduk@ZONE.MIT.EDU with kvno 3, encryption type des3-cbc-sha1 added to keytab WRFILE:kaduk-zone.keytab.
Once all principals have been re-keyed, DES support can be disabled on the
KDC (**allow_weak_crypto = false**), and client machines can remove
**allow_weak_crypto = true** from their :ref:`krb5.conf(5)` configuration
files, completing the migration. **allow_weak_crypto** takes precedence over
all places where DES enctypes could be explicitly configured. DES keys will
not be used, even if they are present, when **allow_weak_crypto = false**.
Support for legacy services
~~~~~~~~~~~~~~~~~~~~~~~~~~~
If there remain legacy services which do not support non-DES enctypes
(such as older versions of AFS), **allow_weak_crypto** must remain
enabled on the KDC. Client machines need not have this setting,
though---applications which require DES can use API calls to allow
weak crypto on a per-request basis, overriding the system krb5.conf.
However, having **allow_weak_crypto** set on the KDC means that any
principals which have a DES key in the database could still use those
keys. To minimize the use of DES in the realm and restrict it to just
legacy services which require DES, it is necessary to remove all other
DES keys. The realm has been configured such that at password and
keytab change, no DES keys will be generated by default. The task
then reduces to requiring user password changes and having server
administrators update their service keytabs. Administrative outreach
will be necessary, and if the desire to eliminate DES is sufficiently
strong, the KDC administrators may choose to randkey any principals
which have not been rekeyed after some timeout period, forcing the
user to contact the helpdesk for access.
The Database Master Key
-----------------------
This procedure does not alter ``K/M@REALM``, the key used to encrypt key
material in the Kerberos database. (This is the key stored in the stash file
on the KDC if stash files are used.) However, the security risk of
a single-DES key for ``K/M`` is minimal, given that access to material
encrypted in ``K/M`` (the Kerberos database) is generally tightly controlled.
If an attacker can gain access to the encrypted database, they likely
have access to the stash file as well, rendering the weak cryptography
broken by non-cryptographic means. As such, upgrading ``K/M`` to a stronger
encryption type is unlikely to be a high-priority task.
Is is possible to upgrade the master key used for the database, if
desired. Using :ref:`kdb5_util(8)`'s **add_mkey**, **use_mkey**, and
**update_princ_encryption** commands, a new master key can be added
and activated for use on new key material, and the existing entries
converted to the new master key.

View file

@ -0,0 +1,171 @@
Application servers
===================
If you need to install the Kerberos V5 programs on an application
server, please refer to the Kerberos V5 Installation Guide. Once you
have installed the software, you need to add that host to the Kerberos
database (see :ref:`principals`), and generate a keytab for that host,
that contains the host's key. You also need to make sure the host's
clock is within your maximum clock skew of the KDCs.
Keytabs
-------
A keytab is a host's copy of its own keylist, which is analogous to a
user's password. An application server that needs to authenticate
itself to the KDC has to have a keytab that contains its own principal
and key. Just as it is important for users to protect their
passwords, it is equally important for hosts to protect their keytabs.
You should always store keytab files on local disk, and make them
readable only by root, and you should never send a keytab file over a
network in the clear. Ideally, you should run the :ref:`kadmin(1)`
command to extract a keytab on the host on which the keytab is to
reside.
.. _add_princ_kt:
Adding principals to keytabs
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
To generate a keytab, or to add a principal to an existing keytab, use
the **ktadd** command from kadmin. Here is a sample session, using
configuration files that enable only AES encryption::
kadmin: ktadd host/daffodil.mit.edu@ATHENA.MIT.EDU
Entry for principal host/daffodil.mit.edu with kvno 2, encryption type aes256-cts-hmac-sha1-96 added to keytab FILE:/etc/krb5.keytab
Entry for principal host/daffodil.mit.edu with kvno 2, encryption type aes128-cts-hmac-sha1-96 added to keytab FILE:/etc/krb5.keytab
Removing principals from keytabs
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
To remove a principal from an existing keytab, use the kadmin
**ktremove** command::
kadmin: ktremove host/daffodil.mit.edu@ATHENA.MIT.EDU
Entry for principal host/daffodil.mit.edu with kvno 2 removed from keytab FILE:/etc/krb5.keytab.
Entry for principal host/daffodil.mit.edu with kvno 2 removed from keytab FILE:/etc/krb5.keytab.
Using a keytab to acquire client credentials
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
While keytabs are ordinarily used to accept credentials from clients,
they can also be used to acquire initial credentials, allowing one
service to authenticate to another.
To manually obtain credentials using a keytab, use the :ref:`kinit(1)`
**-k** option, together with the **-t** option if the keytab is not in
the default location.
Beginning with release 1.11, GSSAPI applications can be configured to
automatically obtain initial credentials from a keytab as needed. The
recommended configuration is as follows:
#. Create a keytab containing a single entry for the desired client
identity.
#. Place the keytab in a location readable by the service, and set the
**KRB5_CLIENT_KTNAME** environment variable to its filename.
Alternatively, use the **default_client_keytab_name** profile
variable in :ref:`libdefaults`, or use the default location of
|ckeytab|.
#. Set **KRB5CCNAME** to a filename writable by the service, which
will not be used for any other purpose. Do not manually obtain
credentials at this location. (Another credential cache type
besides **FILE** can be used if desired, as long the cache will not
conflict with another use. A **MEMORY** cache can be used if the
service runs as a long-lived process. See :ref:`ccache_definition`
for details.)
#. Start the service. When it authenticates using GSSAPI, it will
automatically obtain credentials from the client keytab into the
specified credential cache, and refresh them before they expire.
Clock Skew
----------
A Kerberos application server host must keep its clock synchronized or
it will reject authentication requests from clients. Modern operating
systems typically provide a facility to maintain the correct time;
make sure it is enabled. This is especially important on virtual
machines, where clocks tend to drift more rapidly than normal machine
clocks.
The default allowable clock skew is controlled by the **clockskew**
variable in :ref:`libdefaults`.
Getting DNS information correct
-------------------------------
Several aspects of Kerberos rely on name service. When a hostname is
used to name a service, clients may canonicalize the hostname using
forward and possibly reverse name resolution. The result of this
canonicalization must match the principal entry in the host's keytab,
or authentication will fail. To work with all client canonicalization
configurations, each host's canonical name must be the fully-qualified
host name (including the domain), and each host's IP address must
reverse-resolve to the canonical name.
Configuration of hostnames varies by operating system. On the
application server itself, canonicalization will typically use the
``/etc/hosts`` file rather than the DNS. Ensure that the line for the
server's hostname is in the following form::
IP address fully-qualified hostname aliases
Here is a sample ``/etc/hosts`` file::
# this is a comment
127.0.0.1 localhost localhost.mit.edu
10.0.0.6 daffodil.mit.edu daffodil trillium wake-robin
The output of ``klist -k`` for this example host should look like::
viola# klist -k
Keytab name: /etc/krb5.keytab
KVNO Principal
---- ------------------------------------------------------------
2 host/daffodil.mit.edu@ATHENA.MIT.EDU
If you were to ssh to this host with a fresh credentials cache (ticket
file), and then :ref:`klist(1)`, the output should list a service
principal of ``host/daffodil.mit.edu@ATHENA.MIT.EDU``.
.. _conf_firewall:
Configuring your firewall to work with Kerberos V5
--------------------------------------------------
If you need off-site users to be able to get Kerberos tickets in your
realm, they must be able to get to your KDC. This requires either
that you have a replica KDC outside your firewall, or that you
configure your firewall to allow UDP requests into at least one of
your KDCs, on whichever port the KDC is running. (The default is port
88; other ports may be specified in the KDC's :ref:`kdc.conf(5)`
file.) Similarly, if you need off-site users to be able to change
their passwords in your realm, they must be able to get to your
Kerberos admin server on the kpasswd port (which defaults to 464). If
you need off-site users to be able to administer your Kerberos realm,
they must be able to get to your Kerberos admin server on the
administrative port (which defaults to 749).
If your on-site users inside your firewall will need to get to KDCs in
other realms, you will also need to configure your firewall to allow
outgoing TCP and UDP requests to port 88, and to port 464 to allow
password changes. If your on-site users inside your firewall will
need to get to Kerberos admin servers in other realms, you will also
need to allow outgoing TCP and UDP requests to port 749.
If any of your KDCs are outside your firewall, you will need to allow
kprop requests to get through to the remote KDC. :ref:`kprop(8)` uses
the ``krb5_prop`` service on port 754 (tcp).
The book *UNIX System Security*, by David Curry, is a good starting
point for learning to configure firewalls.

View file

@ -0,0 +1,57 @@
.. _auth_indicator:
Authentication indicators
=========================
As of release 1.14, the KDC can be configured to annotate tickets if
the client authenticated using a stronger preauthentication mechanism
such as :ref:`PKINIT <pkinit>` or :ref:`OTP <otp_preauth>`. These
annotations are called "authentication indicators." Service
principals can be configured to require particular authentication
indicators in order to authenticate to that service. An
authentication indicator value can be any string chosen by the KDC
administrator; there are no pre-set values.
To use authentication indicators with PKINIT or OTP, first configure
the KDC to include an indicator when that preauthentication mechanism
is used. For PKINIT, use the **pkinit_indicator** variable in
:ref:`kdc.conf(5)`. For OTP, use the **indicator** variable in the
token type definition, or specify the indicators in the **otp** user
string as described in :ref:`otp_preauth`.
To require an indicator to be present in order to authenticate to a
service principal, set the **require_auth** string attribute on the
principal to the indicator value to be required. If you wish to allow
one of several indicators to be accepted, you can specify multiple
indicator values separated by spaces.
For example, a realm could be configured to set the authentication
indicator value "strong" when PKINIT is used to authenticate, using a
setting in the :ref:`kdc_realms` subsection::
pkinit_indicator = strong
A service principal could be configured to require the "strong"
authentication indicator value::
$ kadmin setstr host/high.value.server require_auth strong
Password for user/admin@KRBTEST.COM:
A user who authenticates with PKINIT would be able to obtain a ticket
for the service principal::
$ kinit -X X509_user_identity=FILE:/my/cert.pem,/my/key.pem user
$ kvno host/high.value.server
host/high.value.server@KRBTEST.COM: kvno = 1
but a user who authenticates with a password would not::
$ kinit user
Password for user@KRBTEST.COM:
$ kvno host/high.value.server
kvno: KDC policy rejects request while getting credentials for
host/high.value.server@KRBTEST.COM
GSSAPI server applications can inspect authentication indicators
through the :ref:`auth-indicators <gssapi_authind_attr>` name
attribute.

View file

@ -0,0 +1,34 @@
Backups of secure hosts
=======================
When you back up a secure host, you should exclude the host's keytab
file from the backup. If someone obtained a copy of the keytab from a
backup, that person could make any host masquerade as the host whose
keytab was compromised. In many configurations, knowledge of the
host's keytab also allows root access to the host. This could be
particularly dangerous if the compromised keytab was from one of your
KDCs. If the machine has a disk crash and the keytab file is lost, it
is easy to generate another keytab file. (See :ref:`add_princ_kt`.)
If you are unable to exclude particular files from backups, you should
ensure that the backups are kept as secure as the host's root
password.
Backing up the Kerberos database
--------------------------------
As with any file, it is possible that your Kerberos database could
become corrupted. If this happens on one of the replica KDCs, you
might never notice, since the next automatic propagation of the
database would install a fresh copy. However, if it happens to the
primary KDC, the corrupted database would be propagated to all of the
replicas during the next propagation. For this reason, MIT recommends
that you back up your Kerberos database regularly. Because the primary
KDC is continuously dumping the database to a file in order to
propagate it to the replica KDCs, it is a simple matter to have a cron
job periodically copy the dump file to a secure machine elsewhere on
your network. (Of course, it is important to make the host where
these backups are stored as secure as your KDCs, and to encrypt its
transmission across your network.) Then if your database becomes
corrupted, you can load the most recent dump onto the primary KDC.
(See :ref:`restore_from_dump`.)

View file

@ -0,0 +1,20 @@
Configuration Files
===================
Kerberos uses configuration files to allow administrators to specify
settings on a per-machine basis. :ref:`krb5.conf(5)` applies to all
applications using the Kerboros library, on clients and servers.
For KDC-specific applications, additional settings can be specified in
:ref:`kdc.conf(5)`; the two files are merged into a configuration profile
used by applications accessing the KDC database directly. :ref:`kadm5.acl(5)`
is also only used on the KDC, it controls permissions for modifying the
KDC database.
Contents
--------
.. toctree::
:maxdepth: 1
krb5_conf
kdc_conf
kadm5_acl

View file

@ -0,0 +1,163 @@
.. _kadm5.acl(5):
kadm5.acl
=========
DESCRIPTION
-----------
The Kerberos :ref:`kadmind(8)` daemon uses an Access Control List
(ACL) file to manage access rights to the Kerberos database.
For operations that affect principals, the ACL file also controls
which principals can operate on which other principals.
The default location of the Kerberos ACL file is
|kdcdir|\ ``/kadm5.acl`` unless this is overridden by the *acl_file*
variable in :ref:`kdc.conf(5)`.
SYNTAX
------
Empty lines and lines starting with the sharp sign (``#``) are
ignored. Lines containing ACL entries have the format::
principal permissions [target_principal [restrictions] ]
.. note::
Line order in the ACL file is important. The first matching entry
will control access for an actor principal on a target principal.
*principal*
(Partially or fully qualified Kerberos principal name.) Specifies
the principal whose permissions are to be set.
Each component of the name may be wildcarded using the ``*``
character.
*permissions*
Specifies what operations may or may not be performed by a
*principal* matching a particular entry. This is a string of one or
more of the following list of characters or their upper-case
counterparts. If the character is *upper-case*, then the operation
is disallowed. If the character is *lower-case*, then the operation
is permitted.
== ======================================================
a [Dis]allows the addition of principals or policies
c [Dis]allows the changing of passwords for principals
d [Dis]allows the deletion of principals or policies
e [Dis]allows the extraction of principal keys
i [Dis]allows inquiries about principals or policies
l [Dis]allows the listing of all principals or policies
m [Dis]allows the modification of principals or policies
p [Dis]allows the propagation of the principal database (used in :ref:`incr_db_prop`)
s [Dis]allows the explicit setting of the key for a principal
x Short for admcilsp. All privileges (except ``e``)
\* Same as x.
== ======================================================
.. note::
The ``extract`` privilege is not included in the wildcard
privilege; it must be explicitly assigned. This privilege
allows the user to extract keys from the database, and must be
handled with great care to avoid disclosure of important keys
like those of the kadmin/* or krbtgt/* principals. The
**lockdown_keys** principal attribute can be used to prevent
key extraction from specific principals regardless of the
granted privilege.
*target_principal*
(Optional. Partially or fully qualified Kerberos principal name.)
Specifies the principal on which *permissions* may be applied.
Each component of the name may be wildcarded using the ``*``
character.
*target_principal* can also include back-references to *principal*,
in which ``*number`` matches the corresponding wildcard in
*principal*.
*restrictions*
(Optional) A string of flags. Allowed restrictions are:
{+\|-}\ *flagname*
flag is forced to the indicated value. The permissible flags
are the same as those for the **default_principal_flags**
variable in :ref:`kdc.conf(5)`.
*-clearpolicy*
policy is forced to be empty.
*-policy pol*
policy is forced to be *pol*.
-{*expire, pwexpire, maxlife, maxrenewlife*} *time*
(:ref:`getdate` string) associated value will be forced to
MIN(*time*, requested value).
The above flags act as restrictions on any add or modify operation
which is allowed due to that ACL line.
.. warning::
If the kadmind ACL file is modified, the kadmind daemon needs to be
restarted for changes to take effect.
EXAMPLE
-------
Here is an example of a kadm5.acl file::
*/admin@ATHENA.MIT.EDU * # line 1
joeadmin@ATHENA.MIT.EDU ADMCIL # line 2
joeadmin/*@ATHENA.MIT.EDU i */root@ATHENA.MIT.EDU # line 3
*/root@ATHENA.MIT.EDU ci *1@ATHENA.MIT.EDU # line 4
*/root@ATHENA.MIT.EDU l * # line 5
sms@ATHENA.MIT.EDU x * -maxlife 9h -postdateable # line 6
(line 1) Any principal in the ``ATHENA.MIT.EDU`` realm with an
``admin`` instance has all administrative privileges except extracting
keys.
(lines 1-3) The user ``joeadmin`` has all permissions except
extracting keys with his ``admin`` instance,
``joeadmin/admin@ATHENA.MIT.EDU`` (matches line 1). He has no
permissions at all with his null instance, ``joeadmin@ATHENA.MIT.EDU``
(matches line 2). His ``root`` and other non-``admin``, non-null
instances (e.g., ``extra`` or ``dbadmin``) have inquire permissions
with any principal that has the instance ``root`` (matches line 3).
(line 4) Any ``root`` principal in ``ATHENA.MIT.EDU`` can inquire
or change the password of their null instance, but not any other
null instance. (Here, ``*1`` denotes a back-reference to the
component matching the first wildcard in the actor principal.)
(line 5) Any ``root`` principal in ``ATHENA.MIT.EDU`` can generate
the list of principals in the database, and the list of policies
in the database. This line is separate from line 4, because list
permission can only be granted globally, not to specific target
principals.
(line 6) Finally, the Service Management System principal
``sms@ATHENA.MIT.EDU`` has all permissions except extracting keys, but
any principal that it creates or modifies will not be able to get
postdateable tickets or tickets with a life of longer than 9 hours.
MODULE BEHAVIOR
---------------
The ACL file can coexist with other authorization modules in release
1.16 and later, as configured in the :ref:`kadm5_auth` section of
:ref:`krb5.conf(5)`. The ACL file will positively authorize
operations according to the rules above, but will never
authoritatively deny an operation, so other modules can authorize
operations in addition to those authorized by the ACL file.
To operate without an ACL file, set the *acl_file* variable in
:ref:`kdc.conf(5)` to the empty string with ``acl_file = ""``.
SEE ALSO
--------
:ref:`kdc.conf(5)`, :ref:`kadmind(8)`

View file

@ -0,0 +1,976 @@
.. _kdc.conf(5):
kdc.conf
========
The kdc.conf file supplements :ref:`krb5.conf(5)` for programs which
are typically only used on a KDC, such as the :ref:`krb5kdc(8)` and
:ref:`kadmind(8)` daemons and the :ref:`kdb5_util(8)` program.
Relations documented here may also be specified in krb5.conf; for the
KDC programs mentioned, krb5.conf and kdc.conf will be merged into a
single configuration profile.
Normally, the kdc.conf file is found in the KDC state directory,
|kdcdir|. You can override the default location by setting the
environment variable **KRB5_KDC_PROFILE**.
Please note that you need to restart the KDC daemon for any configuration
changes to take effect.
Structure
---------
The kdc.conf file is set up in the same format as the
:ref:`krb5.conf(5)` file.
Sections
--------
The kdc.conf file may contain the following sections:
==================== =================================================
:ref:`kdcdefaults` Default values for KDC behavior
:ref:`kdc_realms` Realm-specific database configuration and settings
:ref:`dbdefaults` Default database settings
:ref:`dbmodules` Per-database settings
:ref:`logging` Controls how Kerberos daemons perform logging
==================== =================================================
.. _kdcdefaults:
[kdcdefaults]
~~~~~~~~~~~~~
Some relations in the [kdcdefaults] section specify default values for
realm variables, to be used if the [realms] subsection does not
contain a relation for the tag. See the :ref:`kdc_realms` section for
the definitions of these relations.
* **host_based_services**
* **kdc_listen**
* **kdc_ports**
* **kdc_tcp_listen**
* **kdc_tcp_ports**
* **no_host_referral**
* **restrict_anonymous_to_tgt**
The following [kdcdefaults] variables have no per-realm equivalent:
**kdc_max_dgram_reply_size**
Specifies the maximum packet size that can be sent over UDP. The
default value is 4096 bytes.
**kdc_tcp_listen_backlog**
(Integer.) Set the size of the listen queue length for the KDC
daemon. The value may be limited by OS settings. The default
value is 5.
**spake_preauth_kdc_challenge**
(String.) Specifies the group for a SPAKE optimistic challenge.
See the **spake_preauth_groups** variable in :ref:`libdefaults`
for possible values. The default is not to issue an optimistic
challenge. (New in release 1.17.)
.. _kdc_realms:
[realms]
~~~~~~~~
Each tag in the [realms] section is the name of a Kerberos realm. The
value of the tag is a subsection where the relations define KDC
parameters for that particular realm. The following example shows how
to define one parameter for the ATHENA.MIT.EDU realm::
[realms]
ATHENA.MIT.EDU = {
max_renewable_life = 7d 0h 0m 0s
}
The following tags may be specified in a [realms] subsection:
**acl_file**
(String.) Location of the access control list file that
:ref:`kadmind(8)` uses to determine which principals are allowed
which permissions on the Kerberos database. To operate without an
ACL file, set this relation to the empty string with ``acl_file =
""``. The default value is |kdcdir|\ ``/kadm5.acl``. For more
information on Kerberos ACL file see :ref:`kadm5.acl(5)`.
**database_module**
(String.) This relation indicates the name of the configuration
section under :ref:`dbmodules` for database-specific parameters
used by the loadable database library. The default value is the
realm name. If this configuration section does not exist, default
values will be used for all database parameters.
**database_name**
(String, deprecated.) This relation specifies the location of the
Kerberos database for this realm, if the DB2 module is being used
and the :ref:`dbmodules` configuration section does not specify a
database name. The default value is |kdcdir|\ ``/principal``.
**default_principal_expiration**
(:ref:`abstime` string.) Specifies the default expiration date of
principals created in this realm. The default value is 0, which
means no expiration date.
**default_principal_flags**
(Flag string.) Specifies the default attributes of principals
created in this realm. The format for this string is a
comma-separated list of flags, with '+' before each flag that
should be enabled and '-' before each flag that should be
disabled. The **postdateable**, **forwardable**, **tgt-based**,
**renewable**, **proxiable**, **dup-skey**, **allow-tickets**, and
**service** flags default to enabled.
There are a number of possible flags:
**allow-tickets**
Enabling this flag means that the KDC will issue tickets for
this principal. Disabling this flag essentially deactivates
the principal within this realm.
**dup-skey**
Enabling this flag allows the KDC to issue user-to-user
service tickets for this principal.
**forwardable**
Enabling this flag allows the principal to obtain forwardable
tickets.
**hwauth**
If this flag is enabled, then the principal is required to
preauthenticate using a hardware device before receiving any
tickets.
**no-auth-data-required**
Enabling this flag prevents PAC or AD-SIGNEDPATH data from
being added to service tickets for the principal.
**ok-as-delegate**
If this flag is enabled, it hints the client that credentials
can and should be delegated when authenticating to the
service.
**ok-to-auth-as-delegate**
Enabling this flag allows the principal to use S4USelf tickets.
**postdateable**
Enabling this flag allows the principal to obtain postdateable
tickets.
**preauth**
If this flag is enabled on a client principal, then that
principal is required to preauthenticate to the KDC before
receiving any tickets. On a service principal, enabling this
flag means that service tickets for this principal will only
be issued to clients with a TGT that has the preauthenticated
bit set.
**proxiable**
Enabling this flag allows the principal to obtain proxy
tickets.
**pwchange**
Enabling this flag forces a password change for this
principal.
**pwservice**
If this flag is enabled, it marks this principal as a password
change service. This should only be used in special cases,
for example, if a user's password has expired, then the user
has to get tickets for that principal without going through
the normal password authentication in order to be able to
change the password.
**renewable**
Enabling this flag allows the principal to obtain renewable
tickets.
**service**
Enabling this flag allows the the KDC to issue service tickets
for this principal. In release 1.17 and later, user-to-user
service tickets are still allowed if the **dup-skey** flag is
set.
**tgt-based**
Enabling this flag allows a principal to obtain tickets based
on a ticket-granting-ticket, rather than repeating the
authentication process that was used to obtain the TGT.
**dict_file**
(String.) Location of the dictionary file containing strings that
are not allowed as passwords. The file should contain one string
per line, with no additional whitespace. If none is specified or
if there is no policy assigned to the principal, no dictionary
checks of passwords will be performed.
**disable_pac**
(Boolean value.) If true, the KDC will not issue PACs for this
realm, and S4U2Self and S4U2Proxy operations will be disabled.
The default is false, which will permit the KDC to issue PACs.
New in release 1.20.
**encrypted_challenge_indicator**
(String.) Specifies the authentication indicator value that the KDC
asserts into tickets obtained using FAST encrypted challenge
pre-authentication. New in 1.16.
**host_based_services**
(Whitespace- or comma-separated list.) Lists services which will
get host-based referral processing even if the server principal is
not marked as host-based by the client.
**iprop_enable**
(Boolean value.) Specifies whether incremental database
propagation is enabled. The default value is false.
**iprop_ulogsize**
(Integer.) Specifies the maximum number of log entries to be
retained for incremental propagation. The default value is 1000.
Prior to release 1.11, the maximum value was 2500. New in release
1.19.
**iprop_master_ulogsize**
The name for **iprop_ulogsize** prior to release 1.19. Its value is
used as a fallback if **iprop_ulogsize** is not specified.
**iprop_replica_poll**
(Delta time string.) Specifies how often the replica KDC polls
for new updates from the primary. The default value is ``2m``
(that is, two minutes). New in release 1.17.
**iprop_slave_poll**
(Delta time string.) The name for **iprop_replica_poll** prior to
release 1.17. Its value is used as a fallback if
**iprop_replica_poll** is not specified.
**iprop_listen**
(Whitespace- or comma-separated list.) Specifies the iprop RPC
listening addresses and/or ports for the :ref:`kadmind(8)` daemon.
Each entry may be an interface address, a port number, or an
address and port number separated by a colon. If the address
contains colons, enclose it in square brackets. If no address is
specified, the wildcard address is used. If kadmind fails to bind
to any of the specified addresses, it will fail to start. The
default (when **iprop_enable** is true) is to bind to the wildcard
address at the port specified in **iprop_port**. New in release
1.15.
**iprop_port**
(Port number.) Specifies the port number to be used for
incremental propagation. When **iprop_enable** is true, this
relation is required in the replica KDC configuration file, and
this relation or **iprop_listen** is required in the primary
configuration file, as there is no default port number. Port
numbers specified in **iprop_listen** entries will override this
port number for the :ref:`kadmind(8)` daemon.
**iprop_resync_timeout**
(Delta time string.) Specifies the amount of time to wait for a
full propagation to complete. This is optional in configuration
files, and is used by replica KDCs only. The default value is 5
minutes (``5m``). New in release 1.11.
**iprop_logfile**
(File name.) Specifies where the update log file for the realm
database is to be stored. The default is to use the
**database_name** entry from the realms section of the krb5 config
file, with ``.ulog`` appended. (NOTE: If **database_name** isn't
specified in the realms section, perhaps because the LDAP database
back end is being used, or the file name is specified in the
[dbmodules] section, then the hard-coded default for
**database_name** is used. Determination of the **iprop_logfile**
default value will not use values from the [dbmodules] section.)
**kadmind_listen**
(Whitespace- or comma-separated list.) Specifies the kadmin RPC
listening addresses and/or ports for the :ref:`kadmind(8)` daemon.
Each entry may be an interface address, a port number, or an
address and port number separated by a colon. If the address
contains colons, enclose it in square brackets. If no address is
specified, the wildcard address is used. If kadmind fails to bind
to any of the specified addresses, it will fail to start. The
default is to bind to the wildcard address at the port specified
in **kadmind_port**, or the standard kadmin port (749). New in
release 1.15.
**kadmind_port**
(Port number.) Specifies the port on which the :ref:`kadmind(8)`
daemon is to listen for this realm. Port numbers specified in
**kadmind_listen** entries will override this port number. The
assigned port for kadmind is 749, which is used by default.
**key_stash_file**
(String.) Specifies the location where the master key has been
stored (via kdb5_util stash). The default is |kdcdir|\
``/.k5.REALM``, where *REALM* is the Kerberos realm.
**kdc_listen**
(Whitespace- or comma-separated list.) Specifies the UDP
listening addresses and/or ports for the :ref:`krb5kdc(8)` daemon.
Each entry may be an interface address, a port number, or an
address and port number separated by a colon. If the address
contains colons, enclose it in square brackets. If no address is
specified, the wildcard address is used. If no port is specified,
the standard port (88) is used. If the KDC daemon fails to bind
to any of the specified addresses, it will fail to start. The
default is to bind to the wildcard address on the standard port.
New in release 1.15.
**kdc_ports**
(Whitespace- or comma-separated list, deprecated.) Prior to
release 1.15, this relation lists the ports for the
:ref:`krb5kdc(8)` daemon to listen on for UDP requests. In
release 1.15 and later, it has the same meaning as **kdc_listen**
if that relation is not defined.
**kdc_tcp_listen**
(Whitespace- or comma-separated list.) Specifies the TCP
listening addresses and/or ports for the :ref:`krb5kdc(8)` daemon.
Each entry may be an interface address, a port number, or an
address and port number separated by a colon. If the address
contains colons, enclose it in square brackets. If no address is
specified, the wildcard address is used. If no port is specified,
the standard port (88) is used. To disable listening on TCP, set
this relation to the empty string with ``kdc_tcp_listen = ""``.
If the KDC daemon fails to bind to any of the specified addresses,
it will fail to start. The default is to bind to the wildcard
address on the standard port. New in release 1.15.
**kdc_tcp_ports**
(Whitespace- or comma-separated list, deprecated.) Prior to
release 1.15, this relation lists the ports for the
:ref:`krb5kdc(8)` daemon to listen on for UDP requests. In
release 1.15 and later, it has the same meaning as
**kdc_tcp_listen** if that relation is not defined.
**kpasswd_listen**
(Comma-separated list.) Specifies the kpasswd listening addresses
and/or ports for the :ref:`kadmind(8)` daemon. Each entry may be
an interface address, a port number, or an address and port number
separated by a colon. If the address contains colons, enclose it
in square brackets. If no address is specified, the wildcard
address is used. If kadmind fails to bind to any of the specified
addresses, it will fail to start. The default is to bind to the
wildcard address at the port specified in **kpasswd_port**, or the
standard kpasswd port (464). New in release 1.15.
**kpasswd_port**
(Port number.) Specifies the port on which the :ref:`kadmind(8)`
daemon is to listen for password change requests for this realm.
Port numbers specified in **kpasswd_listen** entries will override
this port number. The assigned port for password change requests
is 464, which is used by default.
**master_key_name**
(String.) Specifies the name of the principal associated with the
master key. The default is ``K/M``.
**master_key_type**
(Key type string.) Specifies the master key's key type. The
default value for this is |defmkey|. For a list of all possible
values, see :ref:`Encryption_types`.
**max_life**
(:ref:`duration` string.) Specifies the maximum time period for
which a ticket may be valid in this realm. The default value is
24 hours.
**max_renewable_life**
(:ref:`duration` string.) Specifies the maximum time period
during which a valid ticket may be renewed in this realm.
The default value is 0.
**no_host_referral**
(Whitespace- or comma-separated list.) Lists services to block
from getting host-based referral processing, even if the client
marks the server principal as host-based or the service is also
listed in **host_based_services**. ``no_host_referral = *`` will
disable referral processing altogether.
**reject_bad_transit**
(Boolean value.) If set to true, the KDC will check the list of
transited realms for cross-realm tickets against the transit path
computed from the realm names and the capaths section of its
:ref:`krb5.conf(5)` file; if the path in the ticket to be issued
contains any realms not in the computed path, the ticket will not
be issued, and an error will be returned to the client instead.
If this value is set to false, such tickets will be issued
anyways, and it will be left up to the application server to
validate the realm transit path.
If the disable-transited-check flag is set in the incoming
request, this check is not performed at all. Having the
**reject_bad_transit** option will cause such ticket requests to
be rejected always.
This transit path checking and config file option currently apply
only to TGS requests.
The default value is true.
**restrict_anonymous_to_tgt**
(Boolean value.) If set to true, the KDC will reject ticket
requests from anonymous principals to service principals other
than the realm's ticket-granting service. This option allows
anonymous PKINIT to be enabled for use as FAST armor tickets
without allowing anonymous authentication to services. The
default value is false. New in release 1.9.
**spake_preauth_indicator**
(String.) Specifies an authentication indicator value that the
KDC asserts into tickets obtained using SPAKE pre-authentication.
The default is not to add any indicators. This option may be
specified multiple times. New in release 1.17.
**supported_enctypes**
(List of *key*:*salt* strings.) Specifies the default key/salt
combinations of principals for this realm. Any principals created
through :ref:`kadmin(1)` will have keys of these types. The
default value for this tag is |defkeysalts|. For lists of
possible values, see :ref:`Keysalt_lists`.
.. _dbdefaults:
[dbdefaults]
~~~~~~~~~~~~
The [dbdefaults] section specifies default values for some database
parameters, to be used if the [dbmodules] subsection does not contain
a relation for the tag. See the :ref:`dbmodules` section for the
definitions of these relations.
* **ldap_kerberos_container_dn**
* **ldap_kdc_dn**
* **ldap_kdc_sasl_authcid**
* **ldap_kdc_sasl_authzid**
* **ldap_kdc_sasl_mech**
* **ldap_kdc_sasl_realm**
* **ldap_kadmind_dn**
* **ldap_kadmind_sasl_authcid**
* **ldap_kadmind_sasl_authzid**
* **ldap_kadmind_sasl_mech**
* **ldap_kadmind_sasl_realm**
* **ldap_service_password_file**
* **ldap_conns_per_server**
.. _dbmodules:
[dbmodules]
~~~~~~~~~~~
The [dbmodules] section contains parameters used by the KDC database
library and database modules. Each tag in the [dbmodules] section is
the name of a Kerberos realm or a section name specified by a realm's
**database_module** parameter. The following example shows how to
define one database parameter for the ATHENA.MIT.EDU realm::
[dbmodules]
ATHENA.MIT.EDU = {
disable_last_success = true
}
The following tags may be specified in a [dbmodules] subsection:
**database_name**
This DB2-specific tag indicates the location of the database in
the filesystem. The default is |kdcdir|\ ``/principal``.
**db_library**
This tag indicates the name of the loadable database module. The
value should be ``db2`` for the DB2 module, ``klmdb`` for the LMDB
module, or ``kldap`` for the LDAP module.
**disable_last_success**
If set to ``true``, suppresses KDC updates to the "Last successful
authentication" field of principal entries requiring
preauthentication. Setting this flag may improve performance.
(Principal entries which do not require preauthentication never
update the "Last successful authentication" field.). First
introduced in release 1.9.
**disable_lockout**
If set to ``true``, suppresses KDC updates to the "Last failed
authentication" and "Failed password attempts" fields of principal
entries requiring preauthentication. Setting this flag may
improve performance, but also disables account lockout. First
introduced in release 1.9.
**ldap_conns_per_server**
This LDAP-specific tag indicates the number of connections to be
maintained per LDAP server.
**ldap_kdc_dn** and **ldap_kadmind_dn**
These LDAP-specific tags indicate the default DN for binding to
the LDAP server. The :ref:`krb5kdc(8)` daemon uses
**ldap_kdc_dn**, while the :ref:`kadmind(8)` daemon and other
administrative programs use **ldap_kadmind_dn**. The kadmind DN
must have the rights to read and write the Kerberos data in the
LDAP database. The KDC DN must have the same rights, unless
**disable_lockout** and **disable_last_success** are true, in
which case it only needs to have rights to read the Kerberos data.
These tags are ignored if a SASL mechanism is set with
**ldap_kdc_sasl_mech** or **ldap_kadmind_sasl_mech**.
**ldap_kdc_sasl_mech** and **ldap_kadmind_sasl_mech**
These LDAP-specific tags specify the SASL mechanism (such as
``EXTERNAL``) to use when binding to the LDAP server. New in
release 1.13.
**ldap_kdc_sasl_authcid** and **ldap_kadmind_sasl_authcid**
These LDAP-specific tags specify the SASL authentication identity
to use when binding to the LDAP server. Not all SASL mechanisms
require an authentication identity. If the SASL mechanism
requires a secret (such as the password for ``DIGEST-MD5``), these
tags also determine the name within the
**ldap_service_password_file** where the secret is stashed. New
in release 1.13.
**ldap_kdc_sasl_authzid** and **ldap_kadmind_sasl_authzid**
These LDAP-specific tags specify the SASL authorization identity
to use when binding to the LDAP server. In most circumstances
they do not need to be specified. New in release 1.13.
**ldap_kdc_sasl_realm** and **ldap_kadmind_sasl_realm**
These LDAP-specific tags specify the SASL realm to use when
binding to the LDAP server. In most circumstances they do not
need to be set. New in release 1.13.
**ldap_kerberos_container_dn**
This LDAP-specific tag indicates the DN of the container object
where the realm objects will be located.
**ldap_servers**
This LDAP-specific tag indicates the list of LDAP servers that the
Kerberos servers can connect to. The list of LDAP servers is
whitespace-separated. The LDAP server is specified by a LDAP URI.
It is recommended to use ``ldapi:`` or ``ldaps:`` URLs to connect
to the LDAP server.
**ldap_service_password_file**
This LDAP-specific tag indicates the file containing the stashed
passwords (created by ``kdb5_ldap_util stashsrvpw``) for the
**ldap_kdc_dn** and **ldap_kadmind_dn** objects, or for the
**ldap_kdc_sasl_authcid** or **ldap_kadmind_sasl_authcid** names
for SASL authentication. This file must be kept secure.
**mapsize**
This LMDB-specific tag indicates the maximum size of the two
database environments in megabytes. The default value is 128.
Increase this value to address "Environment mapsize limit reached"
errors. New in release 1.17.
**max_readers**
This LMDB-specific tag indicates the maximum number of concurrent
reading processes for the databases. The default value is 128.
New in release 1.17.
**nosync**
This LMDB-specific tag can be set to improve the throughput of
kadmind and other administrative agents, at the expense of
durability (recent database changes may not survive a power outage
or other sudden reboot). It does not affect the throughput of the
KDC. The default value is false. New in release 1.17.
**unlockiter**
If set to ``true``, this DB2-specific tag causes iteration
operations to release the database lock while processing each
principal. Setting this flag to ``true`` can prevent extended
blocking of KDC or kadmin operations when dumps of large databases
are in progress. First introduced in release 1.13.
The following tag may be specified directly in the [dbmodules]
section to control where database modules are loaded from:
**db_module_dir**
This tag controls where the plugin system looks for database
modules. The value should be an absolute path.
.. _logging:
[logging]
~~~~~~~~~
The [logging] section indicates how :ref:`krb5kdc(8)` and
:ref:`kadmind(8)` perform logging. It may contain the following
relations:
**admin_server**
Specifies how :ref:`kadmind(8)` performs logging.
**kdc**
Specifies how :ref:`krb5kdc(8)` performs logging.
**default**
Specifies how either daemon performs logging in the absence of
relations specific to the daemon.
**debug**
(Boolean value.) Specifies whether debugging messages are
included in log outputs other than SYSLOG. Debugging messages are
always included in the system log output because syslog performs
its own priority filtering. The default value is false. New in
release 1.15.
Logging specifications may have the following forms:
**FILE=**\ *filename* or **FILE:**\ *filename*
This value causes the daemon's logging messages to go to the
*filename*. If the ``=`` form is used, the file is overwritten.
If the ``:`` form is used, the file is appended to.
**STDERR**
This value causes the daemon's logging messages to go to its
standard error stream.
**CONSOLE**
This value causes the daemon's logging messages to go to the
console, if the system supports it.
**DEVICE=**\ *<devicename>*
This causes the daemon's logging messages to go to the specified
device.
**SYSLOG**\ [\ **:**\ *severity*\ [\ **:**\ *facility*\ ]]
This causes the daemon's logging messages to go to the system log.
For backward compatibility, a severity argument may be specified,
and must be specified in order to specify a facility. This
argument will be ignored.
The facility argument specifies the facility under which the
messages are logged. This may be any of the following facilities
supported by the syslog(3) call minus the LOG\_ prefix: **KERN**,
**USER**, **MAIL**, **DAEMON**, **AUTH**, **LPR**, **NEWS**,
**UUCP**, **CRON**, and **LOCAL0** through **LOCAL7**. If no
facility is specified, the default is **AUTH**.
In the following example, the logging messages from the KDC will go to
the console and to the system log under the facility LOG_DAEMON, and
the logging messages from the administrative server will be appended
to the file ``/var/adm/kadmin.log`` and sent to the device
``/dev/tty04``. ::
[logging]
kdc = CONSOLE
kdc = SYSLOG:INFO:DAEMON
admin_server = FILE:/var/adm/kadmin.log
admin_server = DEVICE=/dev/tty04
If no logging specification is given, the default is to use syslog.
To disable logging entirely, specify ``default = DEVICE=/dev/null``.
.. _otp:
[otp]
~~~~~
Each subsection of [otp] is the name of an OTP token type. The tags
within the subsection define the configuration required to forward a
One Time Password request to a RADIUS server.
For each token type, the following tags may be specified:
**server**
This is the server to send the RADIUS request to. It can be a
hostname with optional port, an ip address with optional port, or
a Unix domain socket address. The default is
|kdcdir|\ ``/<name>.socket``.
**secret**
This tag indicates a filename (which may be relative to |kdcdir|)
containing the secret used to encrypt the RADIUS packets. The
secret should appear in the first line of the file by itself;
leading and trailing whitespace on the line will be removed. If
the value of **server** is a Unix domain socket address, this tag
is optional, and an empty secret will be used if it is not
specified. Otherwise, this tag is required.
**timeout**
An integer which specifies the time in seconds during which the
KDC should attempt to contact the RADIUS server. This tag is the
total time across all retries and should be less than the time
which an OTP value remains valid for. The default is 5 seconds.
**retries**
This tag specifies the number of retries to make to the RADIUS
server. The default is 3 retries (4 tries).
**strip_realm**
If this tag is ``true``, the principal without the realm will be
passed to the RADIUS server. Otherwise, the realm will be
included. The default value is ``true``.
**indicator**
This tag specifies an authentication indicator to be included in
the ticket if this token type is used to authenticate. This
option may be specified multiple times. (New in release 1.14.)
In the following example, requests are sent to a remote server via UDP::
[otp]
MyRemoteTokenType = {
server = radius.mydomain.com:1812
secret = SEmfiajf42$
timeout = 15
retries = 5
strip_realm = true
}
An implicit default token type named ``DEFAULT`` is defined for when
the per-principal configuration does not specify a token type. Its
configuration is shown below. You may override this token type to
something applicable for your situation::
[otp]
DEFAULT = {
strip_realm = false
}
PKINIT options
--------------
.. note::
The following are pkinit-specific options. These values may
be specified in [kdcdefaults] as global defaults, or within
a realm-specific subsection of [realms]. Also note that a
realm-specific value over-rides, does not add to, a generic
[kdcdefaults] specification. The search order is:
1. realm-specific subsection of [realms]::
[realms]
EXAMPLE.COM = {
pkinit_anchors = FILE:/usr/local/example.com.crt
}
2. generic value in the [kdcdefaults] section::
[kdcdefaults]
pkinit_anchors = DIR:/usr/local/generic_trusted_cas/
For information about the syntax of some of these options, see
:ref:`Specifying PKINIT identity information <pkinit_identity>` in
:ref:`krb5.conf(5)`.
**pkinit_anchors**
Specifies the location of trusted anchor (root) certificates which
the KDC trusts to sign client certificates. This option is
required if pkinit is to be supported by the KDC. This option may
be specified multiple times.
**pkinit_dh_min_bits**
Specifies the minimum number of bits the KDC is willing to accept
for a client's Diffie-Hellman key. The default is 2048.
**pkinit_allow_upn**
Specifies that the KDC is willing to accept client certificates
with the Microsoft UserPrincipalName (UPN) Subject Alternative
Name (SAN). This means the KDC accepts the binding of the UPN in
the certificate to the Kerberos principal name. The default value
is false.
Without this option, the KDC will only accept certificates with
the id-pkinit-san as defined in :rfc:`4556`. There is currently
no option to disable SAN checking in the KDC.
**pkinit_eku_checking**
This option specifies what Extended Key Usage (EKU) values the KDC
is willing to accept in client certificates. The values
recognized in the kdc.conf file are:
**kpClientAuth**
This is the default value and specifies that client
certificates must have the id-pkinit-KPClientAuth EKU as
defined in :rfc:`4556`.
**scLogin**
If scLogin is specified, client certificates with the
Microsoft Smart Card Login EKU (id-ms-kp-sc-logon) will be
accepted.
**none**
If none is specified, then client certificates will not be
checked to verify they have an acceptable EKU. The use of
this option is not recommended.
**pkinit_identity**
Specifies the location of the KDC's X.509 identity information.
This option is required if pkinit is to be supported by the KDC.
**pkinit_indicator**
Specifies an authentication indicator to include in the ticket if
pkinit is used to authenticate. This option may be specified
multiple times. (New in release 1.14.)
**pkinit_pool**
Specifies the location of intermediate certificates which may be
used by the KDC to complete the trust chain between a client's
certificate and a trusted anchor. This option may be specified
multiple times.
**pkinit_revoke**
Specifies the location of Certificate Revocation List (CRL)
information to be used by the KDC when verifying the validity of
client certificates. This option may be specified multiple times.
**pkinit_require_crl_checking**
The default certificate verification process will always check the
available revocation information to see if a certificate has been
revoked. If a match is found for the certificate in a CRL,
verification fails. If the certificate being verified is not
listed in a CRL, or there is no CRL present for its issuing CA,
and **pkinit_require_crl_checking** is false, then verification
succeeds.
However, if **pkinit_require_crl_checking** is true and there is
no CRL information available for the issuing CA, then verification
fails.
**pkinit_require_crl_checking** should be set to true if the
policy is such that up-to-date CRLs must be present for every CA.
**pkinit_require_freshness**
Specifies whether to require clients to include a freshness token
in PKINIT requests. The default value is false. (New in release
1.17.)
.. _Encryption_types:
Encryption types
----------------
Any tag in the configuration files which requires a list of encryption
types can be set to some combination of the following strings.
Encryption types marked as "weak" and "deprecated" are available for
compatibility but not recommended for use.
==================================================== =========================================================
des3-cbc-raw Triple DES cbc mode raw (weak)
des3-cbc-sha1 des3-hmac-sha1 des3-cbc-sha1-kd Triple DES cbc mode with HMAC/sha1 (deprecated)
aes256-cts-hmac-sha1-96 aes256-cts aes256-sha1 AES-256 CTS mode with 96-bit SHA-1 HMAC
aes128-cts-hmac-sha1-96 aes128-cts aes128-sha1 AES-128 CTS mode with 96-bit SHA-1 HMAC
aes256-cts-hmac-sha384-192 aes256-sha2 AES-256 CTS mode with 192-bit SHA-384 HMAC
aes128-cts-hmac-sha256-128 aes128-sha2 AES-128 CTS mode with 128-bit SHA-256 HMAC
arcfour-hmac rc4-hmac arcfour-hmac-md5 RC4 with HMAC/MD5 (deprecated)
arcfour-hmac-exp rc4-hmac-exp arcfour-hmac-md5-exp Exportable RC4 with HMAC/MD5 (weak)
camellia256-cts-cmac camellia256-cts Camellia-256 CTS mode with CMAC
camellia128-cts-cmac camellia128-cts Camellia-128 CTS mode with CMAC
des3 The triple DES family: des3-cbc-sha1
aes The AES family: aes256-cts-hmac-sha1-96, aes128-cts-hmac-sha1-96, aes256-cts-hmac-sha384-192, and aes128-cts-hmac-sha256-128
rc4 The RC4 family: arcfour-hmac
camellia The Camellia family: camellia256-cts-cmac and camellia128-cts-cmac
==================================================== =========================================================
The string **DEFAULT** can be used to refer to the default set of
types for the variable in question. Types or families can be removed
from the current list by prefixing them with a minus sign ("-").
Types or families can be prefixed with a plus sign ("+") for symmetry;
it has the same meaning as just listing the type or family. For
example, "``DEFAULT -rc4``" would be the default set of encryption
types with RC4 types removed, and "``des3 DEFAULT``" would be the
default set of encryption types with triple DES types moved to the
front.
While **aes128-cts** and **aes256-cts** are supported for all Kerberos
operations, they are not supported by very old versions of our GSSAPI
implementation (krb5-1.3.1 and earlier). Services running versions of
krb5 without AES support must not be given keys of these encryption
types in the KDC database.
The **aes128-sha2** and **aes256-sha2** encryption types are new in
release 1.15. Services running versions of krb5 without support for
these newer encryption types must not be given keys of these
encryption types in the KDC database.
.. _Keysalt_lists:
Keysalt lists
-------------
Kerberos keys for users are usually derived from passwords. Kerberos
commands and configuration parameters that affect generation of keys
take lists of enctype-salttype ("keysalt") pairs, known as *keysalt
lists*. Each keysalt pair is an enctype name followed by a salttype
name, in the format *enc*:*salt*. Individual keysalt list members are
separated by comma (",") characters or space characters. For example::
kadmin -e aes256-cts:normal,aes128-cts:normal
would start up kadmin so that by default it would generate
password-derived keys for the **aes256-cts** and **aes128-cts**
encryption types, using a **normal** salt.
To ensure that people who happen to pick the same password do not have
the same key, Kerberos 5 incorporates more information into the key
using something called a salt. The supported salt types are as
follows:
================= ============================================
normal default for Kerberos Version 5
norealm same as the default, without using realm information
onlyrealm uses only realm information as the salt
special generate a random salt
================= ============================================
Sample kdc.conf File
--------------------
Here's an example of a kdc.conf file::
[kdcdefaults]
kdc_listen = 88
kdc_tcp_listen = 88
[realms]
ATHENA.MIT.EDU = {
kadmind_port = 749
max_life = 12h 0m 0s
max_renewable_life = 7d 0h 0m 0s
master_key_type = aes256-cts-hmac-sha1-96
supported_enctypes = aes256-cts-hmac-sha1-96:normal aes128-cts-hmac-sha1-96:normal
database_module = openldap_ldapconf
}
[logging]
kdc = FILE:/usr/local/var/krb5kdc/kdc.log
admin_server = FILE:/usr/local/var/krb5kdc/kadmin.log
[dbdefaults]
ldap_kerberos_container_dn = cn=krbcontainer,dc=mit,dc=edu
[dbmodules]
openldap_ldapconf = {
db_library = kldap
disable_last_success = true
ldap_kdc_dn = "cn=krbadmin,dc=mit,dc=edu"
# this object needs to have read rights on
# the realm container and principal subtrees
ldap_kadmind_dn = "cn=krbadmin,dc=mit,dc=edu"
# this object needs to have read and write rights on
# the realm container and principal subtrees
ldap_service_password_file = /etc/kerberos/service.keyfile
ldap_servers = ldaps://kerberos.mit.edu
ldap_conns_per_server = 5
}
FILES
------
|kdcdir|\ ``/kdc.conf``
SEE ALSO
---------
:ref:`krb5.conf(5)`, :ref:`krb5kdc(8)`, :ref:`kadm5.acl(5)`

File diff suppressed because it is too large Load diff

View file

@ -0,0 +1,132 @@
.. _conf_ldap:
Configuring Kerberos with OpenLDAP back-end
===========================================
1. Make sure the LDAP server is using local authentication
(``ldapi://``) or TLS (``ldaps``). See
https://www.openldap.org/doc/admin/tls.html for instructions on
configuring TLS support in OpenLDAP.
2. Add the Kerberos schema file to the LDAP Server using the OpenLDAP
LDIF file from the krb5 source directory
(``src/plugins/kdb/ldap/libkdb_ldap/kerberos.openldap.ldif``).
The following example uses local authentication::
ldapadd -Y EXTERNAL -H ldapi:/// -f /path/to/kerberos.openldap.ldif
3. Choose DNs for the :ref:`krb5kdc(8)` and :ref:`kadmind(8)` servers
to bind to the LDAP server, and create them if necessary. Specify
these DNs with the **ldap_kdc_dn** and **ldap_kadmind_dn**
directives in :ref:`kdc.conf(5)`. The kadmind DN will also be
used for administrative commands such as :ref:`kdb5_util(8)`.
Alternatively, you may configure krb5kdc and kadmind to use SASL
authentication to access the LDAP server; see the :ref:`dbmodules`
relations **ldap_kdc_sasl_mech** and similar.
4. Specify a location for the LDAP service password file by setting
**ldap_service_password_file**. Use ``kdb5_ldap_util stashsrvpw``
to stash passwords for the KDC and kadmind DNs chosen above. For
example::
kdb5_ldap_util stashsrvpw -f /path/to/service.keyfile cn=krbadmin,dc=example,dc=com
Skip this step if you are using SASL authentication and the
mechanism does not require a password.
5. Choose a DN for the global Kerberos container entry (but do not
create the entry at this time). Specify this DN with the
**ldap_kerberos_container_dn** directive in :ref:`kdc.conf(5)`.
Realm container entries will be created underneath this DN.
Principal entries may exist either underneath the realm container
(the default) or in separate trees referenced from the realm
container.
6. Configure the LDAP server ACLs to enable the KDC and kadmin server
DNs to read and write the Kerberos data. If
**disable_last_success** and **disable_lockout** are both set to
true in the :ref:`dbmodules` subsection for the realm, then the
KDC DN only requires read access to the Kerberos data.
Sample access control information::
access to dn.base=""
by * read
access to dn.base="cn=Subschema"
by * read
# Provide access to the realm container.
access to dn.subtree= "cn=EXAMPLE.COM,cn=krbcontainer,dc=example,dc=com"
by dn.exact="cn=kdc-service,dc=example,dc=com" write
by dn.exact="cn=adm-service,dc=example,dc=com" write
by * none
# Provide access to principals, if not underneath the realm container.
access to dn.subtree= "ou=users,dc=example,dc=com"
by dn.exact="cn=kdc-service,dc=example,dc=com" write
by dn.exact="cn=adm-service,dc=example,dc=com" write
by * none
access to *
by * read
If the locations of the container and principals or the DNs of the
service objects for a realm are changed then this information
should be updated.
7. In :ref:`kdc.conf(5)`, make sure the following relations are set
in the :ref:`dbmodules` subsection for the realm::
db_library (set to ``kldap``)
ldap_kerberos_container_dn
ldap_kdc_dn
ldap_kadmind_dn
ldap_service_password_file
ldap_servers
8. Create the realm using :ref:`kdb5_ldap_util(8)`:
kdb5_ldap_util create -subtrees ou=users,dc=example,dc=com -s
Use the **-subtrees** option if the principals are to exist in a
separate subtree from the realm container. Before executing the
command, make sure that the subtree mentioned above
``(ou=users,dc=example,dc=com)`` exists. If the principals will
exist underneath the realm container, omit the **-subtrees** option
and do not worry about creating the principal subtree.
For more information, refer to the section :ref:`ops_on_ldap`.
The realm object is created under the
**ldap_kerberos_container_dn** specified in the configuration
file. This operation will also create the Kerberos container, if
not present already. This container can be used to store
information related to multiple realms.
9. Add an ``eq`` index for ``krbPrincipalName`` to speed up principal
lookup operations. See
https://www.openldap.org/doc/admin/tuning.html#Indexes for
details.
With the LDAP back end it is possible to provide aliases for principal
entries. Currently we provide no administrative utilities for
creating aliases, so it must be done by direct manipulation of the
LDAP entries.
An entry with aliases contains multiple values of the
*krbPrincipalName* attribute. Since LDAP attribute values are not
ordered, it is necessary to specify which principal name is canonical,
by using the *krbCanonicalName* attribute. Therefore, to create
aliases for an entry, first set the *krbCanonicalName* attribute of
the entry to the canonical principal name (which should be identical
to the pre-existing *krbPrincipalName* value), and then add additional
*krbPrincipalName* attributes for the aliases.
Principal aliases are only returned by the KDC when the client
requests canonicalization. Canonicalization is normally requested for
service principals; for client principals, an explicit flag is often
required (e.g., ``kinit -C``) and canonicalization is only performed
for initial ticket requests.

View file

@ -0,0 +1,587 @@
Database administration
=======================
A Kerberos database contains all of a realm's Kerberos principals,
their passwords, and other administrative information about each
principal. For the most part, you will use the :ref:`kdb5_util(8)`
program to manipulate the Kerberos database as a whole, and the
:ref:`kadmin(1)` program to make changes to the entries in the
database. (One notable exception is that users will use the
:ref:`kpasswd(1)` program to change their own passwords.) The kadmin
program has its own command-line interface, to which you type the
database administrating commands.
:ref:`kdb5_util(8)` provides a means to create, delete, load, or dump
a Kerberos database. It also contains commands to roll over the
database master key, and to stash a copy of the key so that the
:ref:`kadmind(8)` and :ref:`krb5kdc(8)` daemons can use the database
without manual input.
:ref:`kadmin(1)` provides for the maintenance of Kerberos principals,
password policies, and service key tables (keytabs). Normally it
operates as a network client using Kerberos authentication to
communicate with :ref:`kadmind(8)`, but there is also a variant, named
kadmin.local, which directly accesses the Kerberos database on the
local filesystem (or through LDAP). kadmin.local is necessary to set
up enough of the database to be able to use the remote version.
kadmin can authenticate to the admin server using the service
principal ``kadmin/admin`` or ``kadmin/HOST`` (where *HOST* is the
hostname of the admin server). If the credentials cache contains a
ticket for either service principal and the **-c** ccache option is
specified, that ticket is used to authenticate to KADM5. Otherwise,
the **-p** and **-k** options are used to specify the client Kerberos
principal name used to authenticate. Once kadmin has determined the
principal name, it requests a ``kadmin/admin`` Kerberos service ticket
from the KDC, and uses that service ticket to authenticate to KADM5.
See :ref:`kadmin(1)` for the available kadmin and kadmin.local
commands and options.
.. _principals:
Principals
----------
Each entry in the Kerberos database contains a Kerberos principal and
the attributes and policies associated with that principal.
To add a principal to the database, use the :ref:`kadmin(1)`
**add_principal** command. User principals should usually be created
with the ``+requires_preauth -allow_svr`` options to help mitigate
dictionary attacks (see :ref:`dictionary`)::
kadmin: addprinc +requires_preauth -allow_svr alice
Enter password for principal "alice@KRBTEST.COM":
Re-enter password for principal "alice@KRBTEST.COM":
User principals which will authenticate with :ref:`pkinit` should
instead by created with the ``-nokey`` option:
kadmin: addprinc -nokey alice
Service principals can be created with the ``-nokey`` option;
long-term keys will be added when a keytab is generated::
kadmin: addprinc -nokey host/foo.mit.edu
kadmin: ktadd -k foo.keytab host/foo.mit.edu
Entry for principal host/foo.mit.edu with kvno 1, encryption type aes256-cts-hmac-sha1-96 added to keytab WRFILE:foo.keytab.
Entry for principal host/foo.mit.edu with kvno 1, encryption type aes128-cts-hmac-sha1-96 added to keytab WRFILE:foo.keytab.
To modify attributes of an existing principal, use the kadmin
**modify_principal** command::
kadmin: modprinc -expire tomorrow alice
Principal "alice@KRBTEST.COM" modified.
To delete a principal, use the kadmin **delete_principal** command::
kadmin: delprinc alice
Are you sure you want to delete the principal "alice@KRBTEST.COM"? (yes/no): yes
Principal "alice@KRBTEST.COM" deleted.
Make sure that you have removed this principal from all ACLs before reusing.
To change a principal's password, use the kadmin **change_password**
command. Password changes made through kadmin are subject to the same
password policies as would apply to password changes made through
:ref:`kpasswd(1)`.
To view the attributes of a principal, use the kadmin`
**get_principal** command.
To generate a listing of principals, use the kadmin
**list_principals** command.
.. _policies:
Policies
--------
A policy is a set of rules governing passwords. Policies can dictate
minimum and maximum password lifetimes, minimum number of characters
and character classes a password must contain, and the number of old
passwords kept in the database.
To add a new policy, use the :ref:`kadmin(1)` **add_policy** command::
kadmin: addpol -maxlife "1 year" -history 3 stduser
To modify attributes of a principal, use the kadmin **modify_policy**
command. To delete a policy, use the kadmin **delete_policy**
command.
To associate a policy with a principal, use the kadmin
**modify_principal** command with the **-policy** option:
kadmin: modprinc -policy stduser alice
Principal "alice@KRBTEST.COM" modified.
A principal entry may be associated with a nonexistent policy, either
because the policy did not exist at the time of associated or was
deleted afterwards. kadmin will warn when associated a principal with
a nonexistent policy, and will annotate the policy name with "[does
not exist]" in the **get_principal** output.
.. _updating_history_key:
Updating the history key
~~~~~~~~~~~~~~~~~~~~~~~~
If a policy specifies a number of old keys kept of two or more, the
stored old keys are encrypted in a history key, which is found in the
key data of the ``kadmin/history`` principal.
Currently there is no support for proper rollover of the history key,
but you can change the history key (for example, to use a better
encryption type) at the cost of invalidating currently stored old
keys. To change the history key, run::
kadmin: change_password -randkey kadmin/history
This command will fail if you specify the **-keepold** flag. Only one
new history key will be created, even if you specify multiple key/salt
combinations.
In the future, we plan to migrate towards encrypting old keys in the
master key instead of the history key, and implementing proper
rollover support for stored old keys.
.. _privileges:
Privileges
----------
Administrative privileges for the Kerberos database are stored in the
file :ref:`kadm5.acl(5)`.
.. note::
A common use of an admin instance is so you can grant
separate permissions (such as administrator access to the
Kerberos database) to a separate Kerberos principal. For
example, the user ``joeadmin`` might have a principal for
his administrative use, called ``joeadmin/admin``. This
way, ``joeadmin`` would obtain ``joeadmin/admin`` tickets
only when he actually needs to use those permissions.
.. _db_operations:
Operations on the Kerberos database
-----------------------------------
The :ref:`kdb5_util(8)` command is the primary tool for administrating
the Kerberos database when using the DB2 or LMDB modules (see
:ref:`dbtypes`). Creating a database is described in
:ref:`create_db`.
To create a stash file using the master password (because the database
was not created with one using the ``create -s`` flag, or after
restoring from a backup which did not contain the stash file), use the
kdb5_util **stash** command::
$ kdb5_util stash
kdb5_util: Cannot find/read stored master key while reading master key
kdb5_util: Warning: proceeding without master key
Enter KDC database master key: <= Type the KDC database master password.
To destroy a database, use the kdb5_util destroy command::
$ kdb5_util destroy
Deleting KDC database stored in '/var/krb5kdc/principal', are you sure?
(type 'yes' to confirm)? yes
OK, deleting database '/var/krb5kdc/principal'...
** Database '/var/krb5kdc/principal' destroyed.
.. _restore_from_dump:
Dumping and loading a Kerberos database
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
To dump a Kerberos database into a text file for backup or transfer
purposes, use the :ref:`kdb5_util(8)` **dump** command on one of the
KDCs::
$ kdb5_util dump dumpfile
$ kbd5_util dump -verbose dumpfile
kadmin/admin@ATHENA.MIT.EDU
krbtgt/ATHENA.MIT.EDU@ATHENA.MIT.EDU
kadmin/history@ATHENA.MIT.EDU
K/M@ATHENA.MIT.EDU
kadmin/changepw@ATHENA.MIT.EDU
You may specify which principals to dump, using full principal names
including realm::
$ kdb5_util dump -verbose someprincs K/M@ATHENA.MIT.EDU kadmin/admin@ATHENA.MIT.EDU
kadmin/admin@ATHENA.MIT.EDU
K/M@ATHENA.MIT.EDU
To restore a Kerberos database dump from a file, use the
:ref:`kdb5_util(8)` **load** command::
$ kdb5_util load dumpfile
To update an existing database with a partial dump file containing
only some principals, use the ``-update`` flag::
$ kdb5_util load -update someprincs
.. note::
If the database file exists, and the *-update* flag was not
given, *kdb5_util* will overwrite the existing database.
.. _updating_master_key:
Updating the master key
~~~~~~~~~~~~~~~~~~~~~~~
Starting with release 1.7, :ref:`kdb5_util(8)` allows the master key
to be changed using a rollover process, with minimal loss of
availability. To roll over the master key, follow these steps:
#. On the primary KDC, run ``kdb5_util list_mkeys`` to view the
current master key version number (KVNO). If you have never rolled
over the master key before, this will likely be version 1::
$ kdb5_util list_mkeys
Master keys for Principal: K/M@KRBTEST.COM
KVNO: 1, Enctype: aes256-cts-hmac-sha384-192, Active on: Thu Jan 01 00:00:00 UTC 1970 *
#. On the primary KDC, run ``kdb5_util use_mkey 1`` to ensure that a
master key activation list is present in the database. This step
is unnecessary in release 1.11.4 or later, or if the database was
initially created with release 1.7 or later.
#. On the primary KDC, run ``kdb5_util add_mkey -s`` to create a new
master key and write it to the stash file. Enter a secure password
when prompted. If this is the first time you are changing the
master key, the new key will have version 2. The new master key
will not be used until you make it active.
#. Propagate the database to all replica KDCs, either manually or by
waiting until the next scheduled propagation. If you do not have
any replica KDCs, you can skip this and the next step.
#. On each replica KDC, run ``kdb5_util list_mkeys`` to verify that
the new master key is present, and then ``kdb5_util stash`` to
write the new master key to the replica KDC's stash file.
#. On the primary KDC, run ``kdb5_util use_mkey 2`` to begin using the
new master key. Replace ``2`` with the version of the new master
key, as appropriate. You can optionally specify a date for the new
master key to become active; by default, it will become active
immediately. Prior to release 1.12, :ref:`kadmind(8)` must be
restarted for this change to take full effect.
#. On the primary KDC, run ``kdb5_util update_princ_encryption``.
This command will iterate over the database and re-encrypt all keys
in the new master key. If the database is large and uses DB2, the
primary KDC will become unavailable while this command runs, but
clients should fail over to replica KDCs (if any are present)
during this time period. In release 1.13 and later, you can
instead run ``kdb5_util -x unlockiter update_princ_encryption`` to
use unlocked iteration; this variant will take longer, but will
keep the database available to the KDC and kadmind while it runs.
#. Wait until the above changes have propagated to all replica KDCs
and until all running KDC and kadmind processes have serviced
requests using updated principal entries.
#. On the primary KDC, run ``kdb5_util purge_mkeys`` to clean up the
old master key.
.. _ops_on_ldap:
Operations on the LDAP database
-------------------------------
The :ref:`kdb5_ldap_util(8)` command is the primary tool for
administrating the Kerberos database when using the LDAP module.
Creating an LDAP Kerberos database is describe in :ref:`conf_ldap`.
To view a list of realms in the LDAP database, use the kdb5_ldap_util
**list** command::
$ kdb5_ldap_util list
KRBTEST.COM
To modify the attributes of a realm, use the kdb5_ldap_util **modify**
command. For example, to change the default realm's maximum ticket
life::
$ kdb5_ldap_util modify -maxtktlife "10 hours"
To display the attributes of a realm, use the kdb5_ldap_util **view**
command::
$ kdb5_ldap_util view
Realm Name: KRBTEST.COM
Maximum Ticket Life: 0 days 00:10:00
To remove a realm from the LDAP database, destroying its contents, use
the kdb5_ldap_util **destroy** command::
$ kdb5_ldap_util destroy
Deleting KDC database of 'KRBTEST.COM', are you sure?
(type 'yes' to confirm)? yes
OK, deleting database of 'KRBTEST.COM'...
** Database of 'KRBTEST.COM' destroyed.
Ticket Policy operations
~~~~~~~~~~~~~~~~~~~~~~~~
Unlike the DB2 and LMDB modules, the LDAP module supports ticket
policy objects, which can be associated with principals to restrict
maximum ticket lifetimes and set mandatory principal flags. Ticket
policy objects are distinct from the password policies described
earlier on this page, and are chiefly managed through kdb5_ldap_util
rather than kadmin. To create a new ticket policy, use the
kdb5_ldap_util **create_policy** command::
$ kdb5_ldap_util create_policy -maxrenewlife "2 days" users
To associate a ticket policy with a principal, use the
:ref:`kadmin(1)` **modify_principal** (or **add_principal**) command
with the **-x tktpolicy=**\ *policy* option::
$ kadmin.local modprinc -x tktpolicy=users alice
To remove a ticket policy reference from a principal, use the same
command with an empty *policy*::
$ kadmin.local modprinc -x tktpolicy= alice
To list the existing ticket policy objects, use the kdb5_ldap_util
**list_policy** command::
$ kdb5_ldap_util list_policy
users
To modify the attributes of a ticket policy object, use the
kdb5_ldap_util **modify_policy** command::
$ kdb5_ldap_util modify_policy -allow_svr +requires_preauth users
To view the attributes of a ticket policy object, use the
kdb5_ldap_util **view_policy** command::
$ kdb5_ldap_util view_policy users
Ticket policy: users
Maximum renewable life: 2 days 00:00:00
Ticket flags: REQUIRES_PRE_AUTH DISALLOW_SVR
To destroy an ticket policy object, use the kdb5_ldap_util
**destroy_policy** command::
$ kdb5_ldap_util destroy_policy users
This will delete the policy object 'users', are you sure?
(type 'yes' to confirm)? yes
** policy object 'users' deleted.
.. _xrealm_authn:
Cross-realm authentication
--------------------------
In order for a KDC in one realm to authenticate Kerberos users in a
different realm, it must share a key with the KDC in the other realm.
In both databases, there must be krbtgt service principals for both realms.
For example, if you need to do cross-realm authentication between the realms
``ATHENA.MIT.EDU`` and ``EXAMPLE.COM``, you would need to add the
principals ``krbtgt/EXAMPLE.COM@ATHENA.MIT.EDU`` and
``krbtgt/ATHENA.MIT.EDU@EXAMPLE.COM`` to both databases.
These principals must all have the same passwords, key version
numbers, and encryption types; this may require explicitly setting
the key version number with the **-kvno** option.
In the ATHENA.MIT.EDU and EXAMPLE.COM cross-realm case, the administrators
would run the following commands on the KDCs in both realms::
shell%: kadmin.local -e "aes256-cts:normal"
kadmin: addprinc -requires_preauth krbtgt/ATHENA.MIT.EDU@EXAMPLE.COM
Enter password for principal krbtgt/ATHENA.MIT.EDU@EXAMPLE.COM:
Re-enter password for principal krbtgt/ATHENA.MIT.EDU@EXAMPLE.COM:
kadmin: addprinc -requires_preauth krbtgt/EXAMPLE.COM@ATHENA.MIT.EDU
Enter password for principal krbtgt/EXAMPLE.COM@ATHENA.MIT.EDU:
Enter password for principal krbtgt/EXAMPLE.COM@ATHENA.MIT.EDU:
kadmin:
.. note::
Even if most principals in a realm are generally created
with the **requires_preauth** flag enabled, this flag is not
desirable on cross-realm authentication keys because doing
so makes it impossible to disable preauthentication on a
service-by-service basis. Disabling it as in the example
above is recommended.
.. note::
It is very important that these principals have good
passwords. MIT recommends that TGT principal passwords be
at least 26 characters of random ASCII text.
.. _changing_krbtgt_key:
Changing the krbtgt key
-----------------------
A Kerberos Ticket Granting Ticket (TGT) is a service ticket for the
principal ``krbtgt/REALM``. The key for this principal is created
when the Kerberos database is initialized and need not be changed.
However, it will only have the encryption types supported by the KDC
at the time of the initial database creation. To allow use of newer
encryption types for the TGT, this key has to be changed.
Changing this key using the normal :ref:`kadmin(1)`
**change_password** command would invalidate any previously issued
TGTs. Therefore, when changing this key, normally one should use the
**-keepold** flag to change_password to retain the previous key in the
database as well as the new key. For example::
kadmin: change_password -randkey -keepold krbtgt/ATHENA.MIT.EDU@ATHENA.MIT.EDU
.. warning::
After issuing this command, the old key is still valid
and is still vulnerable to (for instance) brute force
attacks. To completely retire an old key or encryption
type, run the kadmin **purgekeys** command to delete keys
with older kvnos, ideally first making sure that all
tickets issued with the old keys have expired.
Only the first krbtgt key of the newest key version is used to encrypt
ticket-granting tickets. However, the set of encryption types present
in the krbtgt keys is used by default to determine the session key
types supported by the krbtgt service (see
:ref:`session_key_selection`). Because non-MIT Kerberos clients
sometimes send a limited set of encryption types when making AS
requests, it can be important for the krbtgt service to support
multiple encryption types. This can be accomplished by giving the
krbtgt principal multiple keys, which is usually as simple as not
specifying any **-e** option when changing the krbtgt key, or by
setting the **session_enctypes** string attribute on the krbtgt
principal (see :ref:`set_string`).
Due to a bug in releases 1.8 through 1.13, renewed and forwarded
tickets may not work if the original ticket was obtained prior to a
krbtgt key change and the modified ticket is obtained afterwards.
Upgrading the KDC to release 1.14 or later will correct this bug.
.. _incr_db_prop:
Incremental database propagation
--------------------------------
Overview
~~~~~~~~
At some very large sites, dumping and transmitting the database can
take more time than is desirable for changes to propagate from the
primary KDC to the replica KDCs. The incremental propagation support
added in the 1.7 release is intended to address this.
With incremental propagation enabled, all programs on the primary KDC
that change the database also write information about the changes to
an "update log" file, maintained as a circular buffer of a certain
size. A process on each replica KDC connects to a service on the
primary KDC (currently implemented in the :ref:`kadmind(8)` server) and
periodically requests the changes that have been made since the last
check. By default, this check is done every two minutes.
Incremental propagation uses the following entries in the per-realm
data in the KDC config file (See :ref:`kdc.conf(5)`):
====================== =============== ===========================================
iprop_enable *boolean* If *true*, then incremental propagation is enabled, and (as noted below) normal kprop propagation is disabled. The default is *false*.
iprop_master_ulogsize *integer* Indicates the number of entries that should be retained in the update log. The default is 1000; the maximum number is 2500.
iprop_replica_poll *time interval* Indicates how often the replica should poll the primary KDC for changes to the database. The default is two minutes.
iprop_port *integer* Specifies the port number to be used for incremental propagation. This is required in both primary and replica configuration files.
iprop_resync_timeout *integer* Specifies the number of seconds to wait for a full propagation to complete. This is optional on replica configurations. Defaults to 300 seconds (5 minutes).
iprop_logfile *file name* Specifies where the update log file for the realm database is to be stored. The default is to use the *database_name* entry from the realms section of the config file :ref:`kdc.conf(5)`, with *.ulog* appended. (NOTE: If database_name isn't specified in the realms section, perhaps because the LDAP database back end is being used, or the file name is specified in the *dbmodules* section, then the hard-coded default for *database_name* is used. Determination of the *iprop_logfile* default value will not use values from the *dbmodules* section.)
====================== =============== ===========================================
Both primary and replica sides must have a principal named
``kiprop/hostname`` (where *hostname* is the lowercase,
fully-qualified, canonical name for the host) registered in the
Kerberos database, and have keys for that principal stored in the
default keytab file (|keytab|). The ``kiprop/hostname`` principal may
have been created automatically for the primary KDC, but it must
always be created for replica KDCs.
On the primary KDC side, the ``kiprop/hostname`` principal must be
listed in the kadmind ACL file :ref:`kadm5.acl(5)`, and given the
**p** privilege (see :ref:`privileges`).
On the replica KDC side, :ref:`kpropd(8)` should be run. When
incremental propagation is enabled, it will connect to the kadmind on
the primary KDC and start requesting updates.
The normal kprop mechanism is disabled by the incremental propagation
support. However, if the replica has been unable to fetch changes
from the primary KDC for too long (network problems, perhaps), the log
on the primary may wrap around and overwrite some of the updates that
the replica has not yet retrieved. In this case, the replica will
instruct the primary KDC to dump the current database out to a file
and invoke a one-time kprop propagation, with special options to also
convey the point in the update log at which the replica should resume
fetching incremental updates. Thus, all the keytab and ACL setup
previously described for kprop propagation is still needed.
If an environment has a large number of replicas, it may be desirable
to arrange them in a hierarchy instead of having the primary serve
updates to every replica. To do this, run ``kadmind -proponly`` on
each intermediate replica, and ``kpropd -A upstreamhostname`` on
downstream replicas to direct each one to the appropriate upstream
replica.
There are several known restrictions in the current implementation:
- The incremental update protocol does not transport changes to policy
objects. Any policy changes on the primary will result in full
resyncs to all replicas.
- The replica's KDB module must support locking; it cannot be using the
LDAP KDB module.
- The primary and replica must be able to initiate TCP connections in
both directions, without an intervening NAT.
Sun/MIT incremental propagation differences
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Sun donated the original code for supporting incremental database
propagation to MIT. Some changes have been made in the MIT source
tree that will be visible to administrators. (These notes are based
on Sun's patches. Changes to Sun's implementation since then may not
be reflected here.)
The Sun config file support looks for ``sunw_dbprop_enable``,
``sunw_dbprop_master_ulogsize``, and ``sunw_dbprop_slave_poll``.
The incremental propagation service is implemented as an ONC RPC
service. In the Sun implementation, the service is registered with
rpcbind (also known as portmapper) and the client looks up the port
number to contact. In the MIT implementation, where interaction with
some modern versions of rpcbind doesn't always work well, the port
number must be specified in the config file on both the primary and
replica sides.
The Sun implementation hard-codes pathnames in ``/var/krb5`` for the
update log and the per-replica kprop dump files. In the MIT
implementation, the pathname for the update log is specified in the
config file, and the per-replica dump files are stored in
|kdcdir|\ ``/replica_datatrans_hostname``.

View file

@ -0,0 +1,149 @@
.. _dbtypes:
Database types
==============
A Kerberos database can be implemented with one of three built-in
database providers, called KDB modules. Software which incorporates
the MIT krb5 KDC may also provide its own KDB module. The following
subsections describe the three built-in KDB modules and the
configuration specific to them.
The database type can be configured with the **db_library** variable
in the :ref:`dbmodules` subsection for the realm. For example::
[dbmodules]
ATHENA.MIT.EDU = {
db_library = db2
}
If the ``ATHENA.MIT.EDU`` realm subsection contains a
**database_module** setting, then the subsection within
``[dbmodules]`` should use that name instead of ``ATHENA.MIT.EDU``.
To transition from one database type to another, stop the
:ref:`kadmind(8)` service, use ``kdb5_util dump`` to create a dump
file, change the **db_library** value and set any appropriate
configuration for the new database type, and use ``kdb5_util load`` to
create and populate the new database. If the new database type is
LDAP, create the new database using ``kdb5_ldap_util`` and populate it
from the dump file using ``kdb5_util load -update``. Then restart the
:ref:`krb5kdc(8)` and :ref:`kadmind(8)` services.
Berkeley database module (db2)
------------------------------
The default KDB module is ``db2``, which uses a version of the
Berkeley DB library. It creates four files based on the database
pathname. If the pathname ends with ``principal`` then the four files
are:
* ``principal``, containing principal entry data
* ``principal.ok``, a lock file for the principal database
* ``principal.kadm5``, containing policy object data
* ``principal.kadm5.lock``, a lock file for the policy database
For large databases, the :ref:`kdb5_util(8)` **dump** command (perhaps
invoked by :ref:`kprop(8)` or by :ref:`kadmind(8)` for incremental
propagation) may cause :ref:`krb5kdc(8)` to stop for a noticeable
period of time while it iterates over the database. This delay can be
avoided by disabling account lockout features so that the KDC does not
perform database writes (see :ref:`disable_lockout`). Alternatively,
a slower form of iteration can be enabled by setting the
**unlockiter** variable to ``true``. For example::
[dbmodules]
ATHENA.MIT.EDU = {
db_library = db2
unlockiter = true
}
In rare cases, a power failure or other unclean system shutdown may
cause inconsistencies in the internal pointers within a database file,
such that ``kdb5_util dump`` cannot retrieve all principal entries in
the database. In this situation, it may be possible to retrieve all
of the principal data by running ``kdb5_util dump -recurse`` to
iterate over the database using the tree pointers instead of the
iteration pointers. Running ``kdb5_util dump -rev`` to iterate over
the database backwards may also retrieve some of the data which is not
retrieved by a normal dump operation.
Lightning Memory-Mapped Database module (klmdb)
-----------------------------------------------
The klmdb module was added in release 1.17. It uses the LMDB library,
and may offer better performance and reliability than the db2 module.
It creates four files based on the database pathname. If the pathname
ends with ``principal``, then the four files are:
* ``principal.mdb``, containing policy object data and most principal
entry data
* ``principal.mdb-lock``, a lock file for the primary database
* ``principal.lockout.mdb``, containing the account lockout attributes
(last successful authentication time, last failed authentication
time, and number of failed attempts) for each principal entry
* ``principal.lockout.mdb-lock``, a lock file for the lockout database
Separating out the lockout attributes ensures that the KDC will never
block on an administrative operation such as a database dump or load.
It also allows the KDC to operate without write access to the primary
database. If both account lockout features are disabled (see
:ref:`disable_lockout`), the lockout database files will be created
but will not subsequently be opened, and the account lockout
attributes will always have zero values.
Because LMDB creates a memory map to the database files, it requires a
configured memory map size which also determines the maximum size of
the database. This size is applied equally to the two databases, so
twice the configured size will be consumed in the process address
space; this is primarily a limitation on 32-bit platforms. The
default value of 128 megabytes should be sufficient for several
hundred thousand principal entries. If the limit is reached, kadmin
operations will fail and the error message "Environment mapsize limit
reached" will appear in the kadmind log file. In this case, the
**mapsize** variable can be used to increase the map size. The
following example sets the map size to 512 megabytes::
[dbmodules]
ATHENA.MIT.EDU = {
db_library = klmdb
mapsize = 512
}
LMDB has a configurable maximum number of readers. The default value
of 128 should be sufficient for most deployments. If you are going to
use a large number of KDC worker processes, it may be necessary to set
the **max_readers** variable to a larger number.
By default, LMDB synchronizes database files to disk after each write
transaction to ensure durability in the case of an unclean system
shutdown. The klmdb module always turns synchronization off for the
lockout database to ensure reasonable KDC performance, but leaves it
on for the primary database. If high throughput for administrative
operations (including password changes) is required, the **nosync**
variable can be set to "true" to disable synchronization for the
primary database.
The klmdb module does not support explicit locking with the
:ref:`kadmin(1)` **lock** command.
LDAP module (kldap)
-------------------
The kldap module stores principal and policy data using an LDAP
server. To use it you must configure an LDAP server to use the
Kerberos schema. See :ref:`conf_ldap` for details.
Because :ref:`krb5kdc(8)` is single-threaded, latency in LDAP database
accesses may limit KDC operation throughput. If the LDAP server is
located on the same server host as the KDC and accessed through an
``ldapi://`` URL, latency should be minimal. If this is not possible,
consider starting multiple KDC worker processes with the
:ref:`krb5kdc(8)` **-w** option to enable concurrent processing of KDC
requests.
The kldap module does not support explicit locking with the
:ref:`kadmin(1)` **lock** command.

View file

@ -0,0 +1,88 @@
.. _dictionary:
Addressing dictionary attack risks
==================================
Kerberos initial authentication is normally secured using the client
principal's long-term key, which for users is generally derived from a
password. Using a pasword-derived long-term key carries the risk of a
dictionary attack, where an attacker tries a sequence of possible
passwords, possibly requiring much less effort than would be required
to try all possible values of the key. Even if :ref:`password policy
objects <policies>` are used to force users not to pick trivial
passwords, dictionary attacks can sometimes be successful against a
significant fraction of the users in a realm. Dictionary attacks are
not a concern for principals using random keys.
A dictionary attack may be online or offline. An online dictionary
attack is performed by trying each password in a separate request to
the KDC, and is therefore visible to the KDC and also limited in speed
by the KDC's processing power and the network capacity between the
client and the KDC. Online dictionary attacks can be mitigated using
:ref:`account lockout <lockout>`. This measure is not totally
satisfactory, as it makes it easy for an attacker to deny access to a
client principal.
An offline dictionary attack is performed by obtaining a ciphertext
generated using the password-derived key, and trying each password
against the ciphertext. This category of attack is invisible to the
KDC and can be performed much faster than an online attack. The
attack will generally take much longer with more recent encryption
types (particularly the ones based on AES), because those encryption
types use a much more expensive string-to-key function. However, the
best defense is to deny the attacker access to a useful ciphertext.
The required defensive measures depend on the attacker's level of
network access.
An off-path attacker has no access to packets sent between legitimate
users and the KDC. An off-path attacker could gain access to an
attackable ciphertext either by making an AS request for a client
principal which does not have the **+requires_preauth** flag, or by
making a TGS request (after authenticating as a different user) for a
server principal which does not have the **-allow_svr** flag. To
address off-path attackers, a KDC administrator should set those flags
on principals with password-derived keys::
kadmin: add_principal +requires_preauth -allow_svr princname
An attacker with passive network access (one who can monitor packets
sent between legitimate users and the KDC, but cannot change them or
insert their own packets) can gain access to an attackable ciphertext
by observing an authentication by a user using the most common form of
preauthentication, encrypted timestamp. Any of the following methods
can prevent dictionary attacks by attackers with passive network
access:
* Enabling :ref:`SPAKE preauthentication <spake>` (added in release
1.17) on the KDC, and ensuring that all clients are able to support
it.
* Using an :ref:`HTTPS proxy <https>` for communication with the KDC,
if the attacker cannot monitor communication between the proxy
server and the KDC.
* Using FAST, protecting the initial authentication with either a
random key (such as a host key) or with :ref:`anonymous PKINIT
<anonymous_pkinit>`.
An attacker with active network access (one who can inject or modify
packets sent between legitimate users and the KDC) can try to fool the
client software into sending an attackable ciphertext using an
encryption type and salt string of the attacker's choosing. Any of the
following methods can prevent dictionary attacks by active attackers:
* Enabling SPAKE preauthentication and setting the
**disable_encrypted_timestamp** variable to ``true`` in the
:ref:`realms` subsection of the client configuration.
* Using an HTTPS proxy as described above, configured in the client's
krb5.conf realm configuration. If :ref:`KDC discovery
<kdc_discovery>` is used to locate a proxy server, an active
attacker may be able to use DNS spoofing to cause the client to use
a different HTTPS server or to not use HTTPS.
* Using FAST as described above.
If :ref:`PKINIT <pkinit>` or :ref:`OTP <otp_preauth>` are used for
initial authentication, the principal's long-term keys are not used
and dictionary attacks are usually not a concern.

View file

@ -0,0 +1,222 @@
.. _enctypes:
Encryption types
================
Kerberos can use a variety of cipher algorithms to protect data. A
Kerberos **encryption type** (also known as an **enctype**) is a
specific combination of a cipher algorithm with an integrity algorithm
to provide both confidentiality and integrity to data.
Enctypes in requests
--------------------
Clients make two types of requests (KDC-REQ) to the KDC: AS-REQs and
TGS-REQs. The client uses the AS-REQ to obtain initial tickets
(typically a Ticket-Granting Ticket (TGT)), and uses the TGS-REQ to
obtain service tickets.
The KDC uses three different keys when issuing a ticket to a client:
* The long-term key of the service: the KDC uses this to encrypt the
actual service ticket. The KDC only uses the first long-term key in
the most recent kvno for this purpose.
* The session key: the KDC randomly chooses this key and places one
copy inside the ticket and the other copy inside the encrypted part
of the reply.
* The reply-encrypting key: the KDC uses this to encrypt the reply it
sends to the client. For AS replies, this is a long-term key of the
client principal. For TGS replies, this is either the session key of the
authenticating ticket, or a subsession key.
Each of these keys is of a specific enctype.
Each request type allows the client to submit a list of enctypes that
it is willing to accept. For the AS-REQ, this list affects both the
session key selection and the reply-encrypting key selection. For the
TGS-REQ, this list only affects the session key selection.
.. _session_key_selection:
Session key selection
---------------------
The KDC chooses the session key enctype by taking the intersection of
its **permitted_enctypes** list, the list of long-term keys for the
most recent kvno of the service, and the client's requested list of
enctypes. Starting in krb5-1.21, all services are assumed to support
aes256-cts-hmac-sha1-96; also, des3-cbc-sha1 and arcfour-hmac session
keys will not be issued by default.
Starting in krb5-1.11, it is possible to set a string attribute on a
service principal to control what session key enctypes the KDC may
issue for service tickets for that principal, overriding the service's
long-term keys and the assumption of aes256-cts-hmac-sha1-96 support.
See :ref:`set_string` in :ref:`kadmin(1)` for details.
Choosing enctypes for a service
-------------------------------
Generally, a service should have a key of the strongest
enctype that both it and the KDC support. If the KDC is running a
release earlier than krb5-1.11, it is also useful to generate an
additional key for each enctype that the service can support. The KDC
will only use the first key in the list of long-term keys for encrypting
the service ticket, but the additional long-term keys indicate the
other enctypes that the service supports.
As noted above, starting with release krb5-1.11, there are additional
configuration settings that control session key enctype selection
independently of the set of long-term keys that the KDC has stored for
a service principal.
Configuration variables
-----------------------
The following ``[libdefaults]`` settings in :ref:`krb5.conf(5)` will
affect how enctypes are chosen.
**allow_weak_crypto**
defaults to *false* starting with krb5-1.8. When *false*, removes
weak enctypes from **permitted_enctypes**,
**default_tkt_enctypes**, and **default_tgs_enctypes**. Do not
set this to *true* unless the use of weak enctypes is an
acceptable risk for your environment and the weak enctypes are
required for backward compatibility.
**allow_des3**
was added in release 1.21 and defaults to *false*. Unless this
flag is set to *true*, the KDC will not issue tickets with
des3-cbc-sha1 session keys. In a future release, this flag will
control whether des3-cbc-sha1 is permitted in similar fashion to
weak enctypes.
**allow_rc4**
was added in release 1.21 and defaults to *false*. Unless this
flag is set to *true*, the KDC will not issue tickets with
arcfour-hmac session keys. In a future release, this flag will
control whether arcfour-hmac is permitted in similar fashion to
weak enctypes.
**permitted_enctypes**
controls the set of enctypes that a service will permit for
session keys and for ticket and authenticator encryption. The KDC
and other programs that access the Kerberos database will ignore
keys of non-permitted enctypes. Starting in release 1.18, this
setting also acts as the default for **default_tkt_enctypes** and
**default_tgs_enctypes**.
**default_tkt_enctypes**
controls the default set of enctypes that the Kerberos client
library requests when making an AS-REQ. Do not set this unless
required for specific backward compatibility purposes; stale
values of this setting can prevent clients from taking advantage
of new stronger enctypes when the libraries are upgraded.
**default_tgs_enctypes**
controls the default set of enctypes that the Kerberos client
library requests when making a TGS-REQ. Do not set this unless
required for specific backward compatibility purposes; stale
values of this setting can prevent clients from taking advantage
of new stronger enctypes when the libraries are upgraded.
The following per-realm setting in :ref:`kdc.conf(5)` affects the
generation of long-term keys.
**supported_enctypes**
controls the default set of enctype-salttype pairs that :ref:`kadmind(8)`
will use for generating long-term keys, either randomly or from
passwords
Enctype compatibility
---------------------
See :ref:`Encryption_types` for additional information about enctypes.
========================== ========== ======== =======
enctype weak? krb5 Windows
========================== ========== ======== =======
des-cbc-crc weak <1.18 >=2000
des-cbc-md4 weak <1.18 ?
des-cbc-md5 weak <1.18 >=2000
des3-cbc-sha1 deprecated >=1.1 none
arcfour-hmac deprecated >=1.3 >=2000
arcfour-hmac-exp weak >=1.3 >=2000
aes128-cts-hmac-sha1-96 >=1.3 >=Vista
aes256-cts-hmac-sha1-96 >=1.3 >=Vista
aes128-cts-hmac-sha256-128 >=1.15 none
aes256-cts-hmac-sha384-192 >=1.15 none
camellia128-cts-cmac >=1.9 none
camellia256-cts-cmac >=1.9 none
========================== ========== ======== =======
krb5 releases 1.18 and later do not support single-DES. krb5 releases
1.8 and later disable the single-DES enctypes by default. Microsoft
Windows releases Windows 7 and later disable single-DES enctypes by
default.
krb5 releases 1.17 and later flag deprecated encryption types
(including ``des3-cbc-sha1`` and ``arcfour-hmac``) in KDC logs and
kadmin output. krb5 release 1.19 issues a warning during initial
authentication if ``des3-cbc-sha1`` is used. Future releases will
disable ``des3-cbc-sha1`` by default and eventually remove support for
it.
Migrating away from older encryption types
------------------------------------------
Administrator intervention may be required to migrate a realm away
from legacy encryption types, especially if the realm was created
using krb5 release 1.2 or earlier. This migration should be performed
before upgrading to krb5 versions which disable or remove support for
legacy encryption types.
If there is a **supported_enctypes** setting in :ref:`kdc.conf(5)` on
the KDC, make sure that it does not include weak or deprecated
encryption types. This will ensure that newly created keys do not use
those encryption types by default.
Check the ``krbtgt/REALM`` principal using the :ref:`kadmin(1)`
**getprinc** command. If it lists a weak or deprecated encryption
type as the first key, it must be migrated using the procedure in
:ref:`changing_krbtgt_key`.
Check the ``kadmin/history`` principal, which should have only one key
entry. If it uses a weak or deprecated encryption type, it should be
upgraded following the notes in :ref:`updating_history_key`.
Check the other kadmin principals: kadmin/changepw, kadmin/admin, and
any kadmin/hostname principals that may exist. These principals can
be upgraded with **change_password -randkey** in kadmin.
Check the ``K/M`` entry. If it uses a weak or deprecated encryption
type, it should be upgraded following the procedure in
:ref:`updating_master_key`.
User and service principals using legacy encryption types can be
enumerated with the :ref:`kdb5_util(8)` **tabdump keyinfo** command.
Service principals can be migrated with a keytab rotation on the
service host, which can be accomplished using the :ref:`k5srvutil(1)`
**change** and **delold** commands. Allow enough time for existing
tickets to expire between the change and delold operations.
User principals with password-based keys can be migrated with a
password change. The realm administrator can set a password
expiration date using the :ref:`kadmin(1)` **modify_principal
-pwexpire** command to force a password change.
If a legacy encryption type has not yet been disabled by default in
the version of krb5 running on the KDC, it can be disabled
administratively with the **permitted_enctypes** variable. For
example, setting **permitted_enctypes** to ``DEFAULT -des3 -rc4`` will
cause any database keys of the triple-DES and RC4 encryption types to
be ignored.

View file

@ -0,0 +1,4 @@
Environment variables
=====================
This content has moved to :ref:`kerberos(7)`.

View file

@ -0,0 +1,235 @@
Host configuration
==================
All hosts running Kerberos software, whether they are clients,
application servers, or KDCs, can be configured using
:ref:`krb5.conf(5)`. Here we describe some of the behavior changes
you might want to make.
Default realm
-------------
In the :ref:`libdefaults` section, the **default_realm** realm
relation sets the default Kerberos realm. For example::
[libdefaults]
default_realm = ATHENA.MIT.EDU
The default realm affects Kerberos behavior in the following ways:
* When a principal name is parsed from text, the default realm is used
if no ``@REALM`` component is specified.
* The default realm affects login authorization as described below.
* For programs which operate on a Kerberos database, the default realm
is used to determine which database to operate on, unless the **-r**
parameter is given to specify a realm.
* A server program may use the default realm when looking up its key
in a :ref:`keytab file <keytab_file>`, if its realm is not
determined by :ref:`domain_realm` configuration or by the server
program itself.
* If :ref:`kinit(1)` is passed the **-n** flag, it requests anonymous
tickets from the default realm.
In some situations, these uses of the default realm might conflict.
For example, it might be desirable for principal name parsing to use
one realm by default, but for login authorization to use a second
realm. In this situation, the first realm can be configured as the
default realm, and **auth_to_local** relations can be used as
described below to use the second realm for login authorization.
.. _login_authorization:
Login authorization
-------------------
If a host runs a Kerberos-enabled login service such as OpenSSH with
GSSAPIAuthentication enabled, login authorization rules determine
whether a Kerberos principal is allowed to access a local account.
By default, a Kerberos principal is allowed access to an account if
its realm matches the default realm and its name matches the account
name. (For historical reasons, access is also granted by default if
the name has two components and the second component matches the
default realm; for instance, ``alice/ATHENA.MIT.EDU@ATHENA.MIT.EDU``
is granted access to the ``alice`` account if ``ATHENA.MIT.EDU`` is
the default realm.)
The simplest way to control local access is using :ref:`.k5login(5)`
files. To use these, place a ``.k5login`` file in the home directory
of each account listing the principal names which should have login
access to that account. If it is not desirable to use ``.k5login``
files located in account home directories, the **k5login_directory**
relation in the :ref:`libdefaults` section can specify a directory
containing one file per account uname.
By default, if a ``.k5login`` file is present, it controls
authorization both positively and negatively--any principal name
contained in the file is granted access and any other principal name
is denied access, even if it would have had access if the ``.k5login``
file didn't exist. The **k5login_authoritative** relation in the
:ref:`libdefaults` section can be set to false to make ``.k5login``
files provide positive authorization only.
The **auth_to_local** relation in the :ref:`realms` section for the
default realm can specify pattern-matching rules to control login
authorization. For example, the following configuration allows access
to principals from a different realm than the default realm::
[realms]
DEFAULT.REALM = {
# Allow access to principals from OTHER.REALM.
#
# [1:$1@$0] matches single-component principal names and creates
# a selection string containing the principal name and realm.
#
# (.*@OTHER\.REALM) matches against the selection string, so that
# only principals in OTHER.REALM are matched.
#
# s/@OTHER\.REALM$// removes the realm name, leaving behind the
# principal name as the account name.
auth_to_local = RULE:[1:$1@$0](.*@OTHER\.REALM)s/@OTHER\.REALM$//
# Also allow principals from the default realm. Omit this line
# to only allow access to principals in OTHER.REALM.
auth_to_local = DEFAULT
}
The **auth_to_local_names** subsection of the :ref:`realms` section
for the default realm can specify explicit mappings from principal
names to local accounts. The key used in this subsection is the
principal name without realm, so it is only safe to use in a Kerberos
environment with a single realm or a tightly controlled set of realms.
An example use of **auth_to_local_names** might be::
[realms]
ATHENA.MIT.EDU = {
auth_to_local_names = {
# Careful, these match principals in any realm!
host/example.com = hostaccount
fred = localfred
}
}
Local authorization behavior can also be modified using plugin
modules; see :ref:`hostrealm_plugin` for details.
.. _plugin_config:
Plugin module configuration
---------------------------
Many aspects of Kerberos behavior, such as client preauthentication
and KDC service location, can be modified through the use of plugin
modules. For most of these behaviors, you can use the :ref:`plugins`
section of krb5.conf to register third-party modules, and to switch
off registered or built-in modules.
A plugin module takes the form of a Unix shared object
(``modname.so``) or Windows DLL (``modname.dll``). If you have
installed a third-party plugin module and want to register it, you do
so using the **module** relation in the appropriate subsection of the
[plugins] section. The value for **module** must give the module name
and the path to the module, separated by a colon. The module name
will often be the same as the shared object's name, but in unusual
cases (such as a shared object which implements multiple modules for
the same interface) it might not be. For example, to register a
client preauthentication module named ``mypreauth`` installed at
``/path/to/mypreauth.so``, you could write::
[plugins]
clpreauth = {
module = mypreauth:/path/to/mypreauth.so
}
Many of the pluggable behaviors in MIT krb5 contain built-in modules
which can be switched off. You can disable a built-in module (or one
you have registered) using the **disable** directive in the
appropriate subsection of the [plugins] section. For example, to
disable the use of .k5identity files to select credential caches, you
could write::
[plugins]
ccselect = {
disable = k5identity
}
If you want to disable multiple modules, specify the **disable**
directive multiple times, giving one module to disable each time.
Alternatively, you can explicitly specify which modules you want to be
enabled for that behavior using the **enable_only** directive. For
example, to make :ref:`kadmind(8)` check password quality using only a
module you have registered, and no other mechanism, you could write::
[plugins]
pwqual = {
module = mymodule:/path/to/mymodule.so
enable_only = mymodule
}
Again, if you want to specify multiple modules, specify the
**enable_only** directive multiple times, giving one module to enable
each time.
Some Kerberos interfaces use different mechanisms to register plugin
modules.
KDC location modules
~~~~~~~~~~~~~~~~~~~~
For historical reasons, modules to control how KDC servers are located
are registered simply by placing the shared object or DLL into the
"libkrb5" subdirectory of the krb5 plugin directory, which defaults to
|libdir|\ ``/krb5/plugins``. For example, Samba's winbind krb5
locator plugin would be registered by placing its shared object in
|libdir|\ ``/krb5/plugins/libkrb5/winbind_krb5_locator.so``.
.. _gssapi_plugin_config:
GSSAPI mechanism modules
~~~~~~~~~~~~~~~~~~~~~~~~
GSSAPI mechanism modules are registered using the file
|sysconfdir|\ ``/gss/mech`` or configuration files in the
|sysconfdir|\ ``/gss/mech.d`` directory with a ``.conf``
suffix. Each line in these files has the form::
name oid pathname [options] <type>
Only the name, oid, and pathname are required. *name* is the
mechanism name, which may be used for debugging or logging purposes.
*oid* is the object identifier of the GSSAPI mechanism to be
registered. *pathname* is a path to the module shared object or DLL.
*options* (if present) are options provided to the plugin module,
surrounded in square brackets. *type* (if present) can be used to
indicate a special type of module. Currently the only special module
type is "interposer", for a module designed to intercept calls to
other mechanisms.
If the environment variable **GSS_MECH_CONFIG** is set, its value is
used as the sole mechanism configuration filename.
.. _profile_plugin_config:
Configuration profile modules
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
A configuration profile module replaces the information source for
:ref:`krb5.conf(5)` itself. To use a profile module, begin krb5.conf
with the line::
module PATHNAME:STRING
where *PATHNAME* is a path to the module shared object or DLL, and
*STRING* is a string to provide to the module. The module will then
take over, and the rest of krb5.conf will be ignored.

View file

@ -0,0 +1,48 @@
.. _https:
HTTPS proxy configuration
=========================
In addition to being able to use UDP or TCP to communicate directly
with a KDC as is outlined in RFC4120, and with kpasswd services in a
similar fashion, the client libraries can attempt to use an HTTPS
proxy server to communicate with a KDC or kpasswd service, using the
protocol outlined in [MS-KKDCP].
Communicating with a KDC through an HTTPS proxy allows clients to
contact servers when network firewalls might otherwise prevent them
from doing so. The use of TLS also encrypts all traffic between the
clients and the KDC, preventing observers from conducting password
dictionary attacks or from observing the client and server principals
being authenticated, at additional computational cost to both clients
and servers.
An HTTPS proxy server is provided as a feature in some versions of
Microsoft Windows Server, and a WSGI implementation named `kdcproxy`
is available in the python package index.
Configuring the clients
-----------------------
To use an HTTPS proxy, a client host must trust the CA which issued
that proxy's SSL certificate. If that CA's certificate is not in the
system-wide default set of trusted certificates, configure the
following relation in the client host's :ref:`krb5.conf(5)` file in
the appropriate :ref:`realms` subsection::
http_anchors = FILE:/etc/krb5/cacert.pem
Adjust the pathname to match the path of the file which contains a
copy of the CA's certificate. The `http_anchors` option is documented
more fully in :ref:`krb5.conf(5)`.
Configure the client to access the KDC and kpasswd service by
specifying their locations in its :ref:`krb5.conf(5)` file in the form
of HTTPS URLs for the proxy server::
kdc = https://server.fqdn/KdcProxy
kpasswd_server = https://server.fqdn/KdcProxy
If the proxy and client are properly configured, client commands such
as ``kinit``, ``kvno``, and ``kpasswd`` should all function normally.

View file

@ -0,0 +1,34 @@
For administrators
==================
.. toctree::
:maxdepth: 1
install.rst
conf_files/index.rst
realm_config.rst
database.rst
dbtypes.rst
lockout.rst
conf_ldap.rst
appl_servers.rst
host_config.rst
backup_host.rst
pkinit.rst
otp.rst
spake.rst
dictionary.rst
princ_dns.rst
enctypes.rst
https.rst
auth_indicator.rst
.. toctree::
:maxdepth: 1
admin_commands/index.rst
../mitK5defaults.rst
env_variables.rst
troubleshoot.rst
advanced/index.rst
various_envs.rst

View file

@ -0,0 +1,21 @@
Installation guide
==================
Contents
--------
.. toctree::
:maxdepth: 2
install_kdc.rst
install_clients.rst
install_appl_srv.rst
Additional references
---------------------
#. Debian: `Setting up MIT Kerberos 5
<http://techpubs.spinlocksolutions.com/dklar/kerberos.html>`_
#. Solaris: `Configuring the Kerberos Service
<https://docs.oracle.com/cd/E19253-01/816-4557/6maosrjv2/index.html>`_

View file

@ -0,0 +1,78 @@
UNIX Application Servers
========================
An application server is a host that provides one or more services
over the network. Application servers can be "secure" or "insecure."
A "secure" host is set up to require authentication from every client
connecting to it. An "insecure" host will still provide Kerberos
authentication, but will also allow unauthenticated clients to
connect.
If you have Kerberos V5 installed on all of your client machines, MIT
recommends that you make your hosts secure, to take advantage of the
security that Kerberos authentication affords. However, if you have
some clients that do not have Kerberos V5 installed, you can run an
insecure server, and still take advantage of Kerberos V5's single
sign-on capability.
.. _keytab_file:
The keytab file
---------------
All Kerberos server machines need a keytab file to authenticate to the
KDC. By default on UNIX-like systems this file is named |keytab|.
The keytab file is an local copy of the host's key. The keytab file
is a potential point of entry for a break-in, and if compromised,
would allow unrestricted access to its host. The keytab file should
be readable only by root, and should exist only on the machine's local
disk. The file should not be part of any backup of the machine,
unless access to the backup data is secured as tightly as access to
the machine's root password.
In order to generate a keytab for a host, the host must have a
principal in the Kerberos database. The procedure for adding hosts to
the database is described fully in :ref:`principals`. (See
:ref:`replica_host_key` for a brief description.) The keytab is
generated by running :ref:`kadmin(1)` and issuing the :ref:`ktadd`
command.
For example, to generate a keytab file to allow the host
``trillium.mit.edu`` to authenticate for the services host, ftp, and
pop, the administrator ``joeadmin`` would issue the command (on
``trillium.mit.edu``)::
trillium% kadmin
Authenticating as principal root/admin@ATHENA.MIT.EDU with password.
Password for root/admin@ATHENA.MIT.EDU:
kadmin: ktadd host/trillium.mit.edu ftp/trillium.mit.edu pop/trillium.mit.edu
Entry for principal host/trillium.mit.edu@ATHENA.MIT.EDU with kvno 3, encryption type aes256-cts-hmac-sha384-192 added to keytab FILE:/etc/krb5.keytab.
kadmin: Entry for principal ftp/trillium.mit.edu@ATHENA.MIT.EDU with kvno 3, encryption type aes256-cts-hmac-sha384-192 added to keytab FILE:/etc/krb5.keytab.
kadmin: Entry for principal pop/trillium.mit.edu@ATHENA.MIT.EDU with kvno 3, encryption type aes256-cts-hmac-sha384-192 added to keytab FILE:/etc/krb5.keytab.
kadmin: quit
trillium%
If you generate the keytab file on another host, you need to get a
copy of the keytab file onto the destination host (``trillium``, in
the above example) without sending it unencrypted over the network.
Some advice about secure hosts
------------------------------
Kerberos V5 can protect your host from certain types of break-ins, but
it is possible to install Kerberos V5 and still leave your host
vulnerable to attack. Obviously an installation guide is not the
place to try to include an exhaustive list of countermeasures for
every possible attack, but it is worth noting some of the larger holes
and how to close them.
We recommend that backups of secure machines exclude the keytab file
(|keytab|). If this is not possible, the backups should at least be
done locally, rather than over a network, and the backup tapes should
be physically secured.
The keytab file and any programs run by root, including the Kerberos
V5 binaries, should be kept on local disk. The keytab file should be
readable only by root.

View file

@ -0,0 +1,58 @@
Installing and configuring UNIX client machines
===============================================
The Kerberized client programs include :ref:`kinit(1)`,
:ref:`klist(1)`, :ref:`kdestroy(1)`, and :ref:`kpasswd(1)`. All of
these programs are in the directory |bindir|.
You can often integrate Kerberos with the login system on client
machines, typically through the use of PAM. The details vary by
operating system, and should be covered in your operating system's
documentation. If you do this, you will need to make sure your users
know to use their Kerberos passwords when they log in.
You will also need to educate your users to use the ticket management
programs kinit, klist, and kdestroy. If you do not have Kerberos
password changing integrated into the native password program (again,
typically through PAM), you will need to educate users to use kpasswd
in place of its non-Kerberos counterparts passwd.
Client machine configuration files
----------------------------------
Each machine running Kerberos should have a :ref:`krb5.conf(5)` file.
At a minimum, it should define a **default_realm** setting in
:ref:`libdefaults`. If you are not using DNS SRV records
(:ref:`kdc_hostnames`) or URI records (:ref:`kdc_discovery`), it must
also contain a :ref:`realms` section containing information for your
realm's KDCs.
Consider setting **rdns** to false in order to reduce your dependence
on precisely correct DNS information for service hostnames. Turning
this flag off means that service hostnames will be canonicalized
through forward name resolution (which adds your domain name to
unqualified hostnames, and resolves CNAME records in DNS), but not
through reverse address lookup. The default value of this flag is
true for historical reasons only.
If you anticipate users frequently logging into remote hosts
(e.g., using ssh) using forwardable credentials, consider setting
**forwardable** to true so that users obtain forwardable tickets by
default. Otherwise users will need to use ``kinit -f`` to get
forwardable tickets.
Consider adjusting the **ticket_lifetime** setting to match the likely
length of sessions for your users. For instance, if most of your
users will be logging in for an eight-hour workday, you could set the
default to ten hours so that tickets obtained in the morning expire
shortly after the end of the workday. Users can still manually
request longer tickets when necessary, up to the maximum allowed by
each user's principal record on the KDC.
If a client host may access services in different realms, it may be
useful to define a :ref:`domain_realm` mapping so that clients know
which hosts belong to which realms. However, if your clients and KDC
are running release 1.7 or later, it is also reasonable to leave this
section out on client machines and just define it in the KDC's
krb5.conf.

View file

@ -0,0 +1,536 @@
Installing KDCs
===============
When setting up Kerberos in a production environment, it is best to
have multiple replica KDCs alongside with a primary KDC to ensure the
continued availability of the Kerberized services. Each KDC contains
a copy of the Kerberos database. The primary KDC contains the
writable copy of the realm database, which it replicates to the
replica KDCs at regular intervals. All database changes (such as
password changes) are made on the primary KDC. Replica KDCs provide
Kerberos ticket-granting services, but not database administration,
when the primary KDC is unavailable. MIT recommends that you install
all of your KDCs to be able to function as either the primary or one
of the replicas. This will enable you to easily switch your primary
KDC with one of the replicas if necessary (see
:ref:`switch_primary_replica`). This installation procedure is based
on that recommendation.
.. warning::
- The Kerberos system relies on the availability of correct time
information. Ensure that the primary and all replica KDCs have
properly synchronized clocks.
- It is best to install and run KDCs on secured and dedicated
hardware with limited access. If your KDC is also a file
server, FTP server, Web server, or even just a client machine,
someone who obtained root access through a security hole in any
of those areas could potentially gain access to the Kerberos
database.
Install and configure the primary KDC
-------------------------------------
Install Kerberos either from the OS-provided packages or from the
source (See :ref:`do_build`).
.. note::
For the purpose of this document we will use the following
names::
kerberos.mit.edu - primary KDC
kerberos-1.mit.edu - replica KDC
ATHENA.MIT.EDU - realm name
.k5.ATHENA.MIT.EDU - stash file
admin/admin - admin principal
See :ref:`mitK5defaults` for the default names and locations
of the relevant to this topic files. Adjust the names and
paths to your system environment.
Edit KDC configuration files
----------------------------
Modify the configuration files, :ref:`krb5.conf(5)` and
:ref:`kdc.conf(5)`, to reflect the correct information (such as
domain-realm mappings and Kerberos servers names) for your realm.
(See :ref:`mitK5defaults` for the recommended default locations for
these files).
Most of the tags in the configuration have default values that will
work well for most sites. There are some tags in the
:ref:`krb5.conf(5)` file whose values must be specified, and this
section will explain those.
If the locations for these configuration files differs from the
default ones, set **KRB5_CONFIG** and **KRB5_KDC_PROFILE** environment
variables to point to the krb5.conf and kdc.conf respectively. For
example::
export KRB5_CONFIG=/yourdir/krb5.conf
export KRB5_KDC_PROFILE=/yourdir/kdc.conf
krb5.conf
~~~~~~~~~
If you are not using DNS TXT records (see :ref:`mapping_hostnames`),
you must specify the **default_realm** in the :ref:`libdefaults`
section. If you are not using DNS URI or SRV records (see
:ref:`kdc_hostnames` and :ref:`kdc_discovery`), you must include the
**kdc** tag for each *realm* in the :ref:`realms` section. To
communicate with the kadmin server in each realm, the **admin_server**
tag must be set in the
:ref:`realms` section.
An example krb5.conf file::
[libdefaults]
default_realm = ATHENA.MIT.EDU
[realms]
ATHENA.MIT.EDU = {
kdc = kerberos.mit.edu
kdc = kerberos-1.mit.edu
admin_server = kerberos.mit.edu
}
kdc.conf
~~~~~~~~
The kdc.conf file can be used to control the listening ports of the
KDC and kadmind, as well as realm-specific defaults, the database type
and location, and logging.
An example kdc.conf file::
[kdcdefaults]
kdc_listen = 88
kdc_tcp_listen = 88
[realms]
ATHENA.MIT.EDU = {
kadmind_port = 749
max_life = 12h 0m 0s
max_renewable_life = 7d 0h 0m 0s
master_key_type = aes256-cts
supported_enctypes = aes256-cts:normal aes128-cts:normal
# If the default location does not suit your setup,
# explicitly configure the following values:
# database_name = /var/krb5kdc/principal
# key_stash_file = /var/krb5kdc/.k5.ATHENA.MIT.EDU
# acl_file = /var/krb5kdc/kadm5.acl
}
[logging]
# By default, the KDC and kadmind will log output using
# syslog. You can instead send log output to files like this:
kdc = FILE:/var/log/krb5kdc.log
admin_server = FILE:/var/log/kadmin.log
default = FILE:/var/log/krb5lib.log
Replace ``ATHENA.MIT.EDU`` and ``kerberos.mit.edu`` with the name of
your Kerberos realm and server respectively.
.. note::
You have to have write permission on the target directories
(these directories must exist) used by **database_name**,
**key_stash_file**, and **acl_file**.
.. _create_db:
Create the KDC database
-----------------------
You will use the :ref:`kdb5_util(8)` command on the primary KDC to
create the Kerberos database and the optional :ref:`stash_definition`.
.. note::
If you choose not to install a stash file, the KDC will
prompt you for the master key each time it starts up. This
means that the KDC will not be able to start automatically,
such as after a system reboot.
:ref:`kdb5_util(8)` will prompt you for the master password for the
Kerberos database. This password can be any string. A good password
is one you can remember, but that no one else can guess. Examples of
bad passwords are words that can be found in a dictionary, any common
or popular name, especially a famous person (or cartoon character),
your username in any form (e.g., forward, backward, repeated twice,
etc.), and any of the sample passwords that appear in this manual.
One example of a password which might be good if it did not appear in
this manual is "MITiys4K5!", which represents the sentence "MIT is
your source for Kerberos 5!" (It's the first letter of each word,
substituting the numeral "4" for the word "for", and includes the
punctuation mark at the end.)
The following is an example of how to create a Kerberos database and
stash file on the primary KDC, using the :ref:`kdb5_util(8)` command.
Replace ``ATHENA.MIT.EDU`` with the name of your Kerberos realm::
shell% kdb5_util create -r ATHENA.MIT.EDU -s
Initializing database '/usr/local/var/krb5kdc/principal' for realm 'ATHENA.MIT.EDU',
master key name 'K/M@ATHENA.MIT.EDU'
You will be prompted for the database Master Password.
It is important that you NOT FORGET this password.
Enter KDC database master key: <= Type the master password.
Re-enter KDC database master key to verify: <= Type it again.
shell%
This will create five files in |kdcdir| (or at the locations specified
in :ref:`kdc.conf(5)`):
* two Kerberos database files, ``principal``, and ``principal.ok``
* the Kerberos administrative database file, ``principal.kadm5``
* the administrative database lock file, ``principal.kadm5.lock``
* the stash file, in this example ``.k5.ATHENA.MIT.EDU``. If you do
not want a stash file, run the above command without the **-s**
option.
For more information on administrating Kerberos database see
:ref:`db_operations`.
.. _admin_acl:
Add administrators to the ACL file
----------------------------------
Next, you need create an Access Control List (ACL) file and put the
Kerberos principal of at least one of the administrators into it.
This file is used by the :ref:`kadmind(8)` daemon to control which
principals may view and make privileged modifications to the Kerberos
database files. The ACL filename is determined by the **acl_file**
variable in :ref:`kdc.conf(5)`; the default is |kdcdir|\
``/kadm5.acl``.
For more information on Kerberos ACL file see :ref:`kadm5.acl(5)`.
.. _addadmin_kdb:
Add administrators to the Kerberos database
-------------------------------------------
Next you need to add administrative principals (i.e., principals who
are allowed to administer Kerberos database) to the Kerberos database.
You *must* add at least one principal now to allow communication
between the Kerberos administration daemon kadmind and the kadmin
program over the network for further administration. To do this, use
the kadmin.local utility on the primary KDC. kadmin.local is designed
to be run on the primary KDC host without using Kerberos
authentication to an admin server; instead, it must have read and
write access to the Kerberos database on the local filesystem.
The administrative principals you create should be the ones you added
to the ACL file (see :ref:`admin_acl`).
In the following example, the administrative principal ``admin/admin``
is created::
shell% kadmin.local
kadmin.local: addprinc admin/admin@ATHENA.MIT.EDU
No policy specified for "admin/admin@ATHENA.MIT.EDU";
assigning "default".
Enter password for principal admin/admin@ATHENA.MIT.EDU: <= Enter a password.
Re-enter password for principal admin/admin@ATHENA.MIT.EDU: <= Type it again.
Principal "admin/admin@ATHENA.MIT.EDU" created.
kadmin.local:
.. _start_kdc_daemons:
Start the Kerberos daemons on the primary KDC
---------------------------------------------
At this point, you are ready to start the Kerberos KDC
(:ref:`krb5kdc(8)`) and administrative daemons on the primary KDC. To
do so, type::
shell% krb5kdc
shell% kadmind
Each server daemon will fork and run in the background.
.. note::
Assuming you want these daemons to start up automatically at
boot time, you can add them to the KDC's ``/etc/rc`` or
``/etc/inittab`` file. You need to have a
:ref:`stash_definition` in order to do this.
You can verify that they started properly by checking for their
startup messages in the logging locations you defined in
:ref:`krb5.conf(5)` (see :ref:`logging`). For example::
shell% tail /var/log/krb5kdc.log
Dec 02 12:35:47 beeblebrox krb5kdc[3187](info): commencing operation
shell% tail /var/log/kadmin.log
Dec 02 12:35:52 beeblebrox kadmind[3189](info): starting
Any errors the daemons encounter while starting will also be listed in
the logging output.
As an additional verification, check if :ref:`kinit(1)` succeeds
against the principals that you have created on the previous step
(:ref:`addadmin_kdb`). Run::
shell% kinit admin/admin@ATHENA.MIT.EDU
Install the replica KDCs
------------------------
You are now ready to start configuring the replica KDCs.
.. note::
Assuming you are setting the KDCs up so that you can easily
switch the primary KDC with one of the replicas, you should
perform each of these steps on the primary KDC as well as
the replica KDCs, unless these instructions specify
otherwise.
.. _replica_host_key:
Create host keytabs for replica KDCs
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Each KDC needs a ``host`` key in the Kerberos database. These keys
are used for mutual authentication when propagating the database dump
file from the primary KDC to the secondary KDC servers.
On the primary KDC, connect to administrative interface and create the
host principal for each of the KDCs' ``host`` services. For example,
if the primary KDC were called ``kerberos.mit.edu``, and you had a
replica KDC named ``kerberos-1.mit.edu``, you would type the
following::
shell% kadmin
kadmin: addprinc -randkey host/kerberos.mit.edu
No policy specified for "host/kerberos.mit.edu@ATHENA.MIT.EDU"; assigning "default"
Principal "host/kerberos.mit.edu@ATHENA.MIT.EDU" created.
kadmin: addprinc -randkey host/kerberos-1.mit.edu
No policy specified for "host/kerberos-1.mit.edu@ATHENA.MIT.EDU"; assigning "default"
Principal "host/kerberos-1.mit.edu@ATHENA.MIT.EDU" created.
It is not strictly necessary to have the primary KDC server in the
Kerberos database, but it can be handy if you want to be able to swap
the primary KDC with one of the replicas.
Next, extract ``host`` random keys for all participating KDCs and
store them in each host's default keytab file. Ideally, you should
extract each keytab locally on its own KDC. If this is not feasible,
you should use an encrypted session to send them across the network.
To extract a keytab directly on a replica KDC called
``kerberos-1.mit.edu``, you would execute the following command::
kadmin: ktadd host/kerberos-1.mit.edu
Entry for principal host/kerberos-1.mit.edu with kvno 2, encryption
type aes256-cts-hmac-sha1-96 added to keytab FILE:/etc/krb5.keytab.
Entry for principal host/kerberos-1.mit.edu with kvno 2, encryption
type aes128-cts-hmac-sha1-96 added to keytab FILE:/etc/krb5.keytab.
Entry for principal host/kerberos-1.mit.edu with kvno 2, encryption
type aes256-cts-hmac-sha384-192 added to keytab FILE:/etc/krb5.keytab.
Entry for principal host/kerberos-1.mit.edu with kvno 2, encryption
type arcfour-hmac added to keytab FILE:/etc/krb5.keytab.
If you are instead extracting a keytab for the replica KDC called
``kerberos-1.mit.edu`` on the primary KDC, you should use a dedicated
temporary keytab file for that machine's keytab::
kadmin: ktadd -k /tmp/kerberos-1.keytab host/kerberos-1.mit.edu
Entry for principal host/kerberos-1.mit.edu with kvno 2, encryption
type aes256-cts-hmac-sha1-96 added to keytab FILE:/etc/krb5.keytab.
Entry for principal host/kerberos-1.mit.edu with kvno 2, encryption
type aes128-cts-hmac-sha1-96 added to keytab FILE:/etc/krb5.keytab.
The file ``/tmp/kerberos-1.keytab`` can then be installed as
``/etc/krb5.keytab`` on the host ``kerberos-1.mit.edu``.
Configure replica KDCs
~~~~~~~~~~~~~~~~~~~~~~
Database propagation copies the contents of the primary's database,
but does not propagate configuration files, stash files, or the kadm5
ACL file. The following files must be copied by hand to each replica
(see :ref:`mitK5defaults` for the default locations for these files):
* krb5.conf
* kdc.conf
* kadm5.acl
* master key stash file
Move the copied files into their appropriate directories, exactly as
on the primary KDC. kadm5.acl is only needed to allow a replica to
swap with the primary KDC.
The database is propagated from the primary KDC to the replica KDCs
via the :ref:`kpropd(8)` daemon. You must explicitly specify the
principals which are allowed to provide Kerberos dump updates on the
replica machine with a new database. Create a file named kpropd.acl
in the KDC state directory containing the ``host`` principals for each
of the KDCs::
host/kerberos.mit.edu@ATHENA.MIT.EDU
host/kerberos-1.mit.edu@ATHENA.MIT.EDU
.. note::
If you expect that the primary and replica KDCs will be
switched at some point of time, list the host principals
from all participating KDC servers in kpropd.acl files on
all of the KDCs. Otherwise, you only need to list the
primary KDC's host principal in the kpropd.acl files of the
replica KDCs.
Then, add the following line to ``/etc/inetd.conf`` on each KDC
(adjust the path to kpropd)::
krb5_prop stream tcp nowait root /usr/local/sbin/kpropd kpropd
You also need to add the following line to ``/etc/services`` on each
KDC, if it is not already present (assuming that the default port is
used)::
krb5_prop 754/tcp # Kerberos replica propagation
Restart inetd daemon.
Alternatively, start :ref:`kpropd(8)` as a stand-alone daemon. This is
required when incremental propagation is enabled.
Now that the replica KDC is able to accept database propagation,
youll need to propagate the database from the primary server.
NOTE: Do not start the replica KDC yet; you still do not have a copy
of the primary's database.
.. _kprop_to_replicas:
Propagate the database to each replica KDC
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
First, create a dump file of the database on the primary KDC, as
follows::
shell% kdb5_util dump /usr/local/var/krb5kdc/replica_datatrans
Then, manually propagate the database to each replica KDC, as in the
following example::
shell% kprop -f /usr/local/var/krb5kdc/replica_datatrans kerberos-1.mit.edu
Database propagation to kerberos-1.mit.edu: SUCCEEDED
You will need a script to dump and propagate the database. The
following is an example of a Bourne shell script that will do this.
.. note::
Remember that you need to replace ``/usr/local/var/krb5kdc``
with the name of the KDC state directory.
::
#!/bin/sh
kdclist = "kerberos-1.mit.edu kerberos-2.mit.edu"
kdb5_util dump /usr/local/var/krb5kdc/replica_datatrans
for kdc in $kdclist
do
kprop -f /usr/local/var/krb5kdc/replica_datatrans $kdc
done
You will need to set up a cron job to run this script at the intervals
you decided on earlier (see :ref:`db_prop`).
Now that the replica KDC has a copy of the Kerberos database, you can
start the krb5kdc daemon::
shell% krb5kdc
As with the primary KDC, you will probably want to add this command to
the KDCs' ``/etc/rc`` or ``/etc/inittab`` files, so they will start
the krb5kdc daemon automatically at boot time.
Propagation failed?
###################
You may encounter the following error messages. For a more detailed
discussion on possible causes and solutions click on the error link
to be redirected to :ref:`troubleshoot` section.
.. include:: ./troubleshoot.rst
:start-after: _prop_failed_start:
:end-before: _prop_failed_end:
Add Kerberos principals to the database
---------------------------------------
Once your KDCs are set up and running, you are ready to use
:ref:`kadmin(1)` to load principals for your users, hosts, and other
services into the Kerberos database. This procedure is described
fully in :ref:`principals`.
You may occasionally want to use one of your replica KDCs as the
primary. This might happen if you are upgrading the primary KDC, or
if your primary KDC has a disk crash. See the following section for
the instructions.
.. _switch_primary_replica:
Switching primary and replica KDCs
----------------------------------
You may occasionally want to use one of your replica KDCs as the
primary. This might happen if you are upgrading the primary KDC, or
if your primary KDC has a disk crash.
Assuming you have configured all of your KDCs to be able to function
as either the primary KDC or a replica KDC (as this document
recommends), all you need to do to make the changeover is:
If the primary KDC is still running, do the following on the *old*
primary KDC:
#. Kill the kadmind process.
#. Disable the cron job that propagates the database.
#. Run your database propagation script manually, to ensure that the
replicas all have the latest copy of the database (see
:ref:`kprop_to_replicas`).
On the *new* primary KDC:
#. Start the :ref:`kadmind(8)` daemon (see :ref:`start_kdc_daemons`).
#. Set up the cron job to propagate the database (see
:ref:`kprop_to_replicas`).
#. Switch the CNAMEs of the old and new primary KDCs. If you can't do
this, you'll need to change the :ref:`krb5.conf(5)` file on every
client machine in your Kerberos realm.
Incremental database propagation
--------------------------------
If you expect your Kerberos database to become large, you may wish to
set up incremental propagation to replica KDCs. See
:ref:`incr_db_prop` for details.

View file

@ -0,0 +1,154 @@
.. _lockout:
Account lockout
===============
As of release 1.8, the KDC can be configured to lock out principals
after a number of failed authentication attempts within a period of
time. Account lockout can make it more difficult to attack a
principal's password by brute force, but also makes it easy for an
attacker to deny access to a principal.
Configuring account lockout
---------------------------
Account lockout only works for principals with the
**+requires_preauth** flag set. Without this flag, the KDC cannot
know whether or not a client successfully decrypted the ticket it
issued. It is also important to set the **-allow_svr** flag on a
principal to protect its password from an off-line dictionary attack
through a TGS request. You can set these flags on a principal with
:ref:`kadmin(1)` as follows::
kadmin: modprinc +requires_preauth -allow_svr PRINCNAME
Account lockout parameters are configured via :ref:`policy objects
<policies>`. There may be an existing policy associated with user
principals (such as the "default" policy), or you may need to create a
new one and associate it with each user principal.
The policy parameters related to account lockout are:
* :ref:`maxfailure <policy_maxfailure>`: the number of failed attempts
before the principal is locked out
* :ref:`failurecountinterval <policy_failurecountinterval>`: the
allowable interval between failed attempts
* :ref:`lockoutduration <policy_lockoutduration>`: the amount of time
a principal is locked out for
Here is an example of setting these parameters on a new policy and
associating it with a principal::
kadmin: addpol -maxfailure 10 -failurecountinterval 180
-lockoutduration 60 lockout_policy
kadmin: modprinc -policy lockout_policy PRINCNAME
Testing account lockout
-----------------------
To test that account lockout is working, try authenticating as the
principal (hopefully not one that might be in use) multiple times with
the wrong password. For instance, if **maxfailure** is set to 2, you
might see::
$ kinit user
Password for user@KRBTEST.COM:
kinit: Password incorrect while getting initial credentials
$ kinit user
Password for user@KRBTEST.COM:
kinit: Password incorrect while getting initial credentials
$ kinit user
kinit: Client's credentials have been revoked while getting initial credentials
Account lockout principal state
-------------------------------
A principal entry keeps three pieces of state related to account
lockout:
* The time of last successful authentication
* The time of last failed authentication
* A counter of failed attempts
The time of last successful authentication is not actually needed for
the account lockout system to function, but may be of administrative
interest. These fields can be observed with the **getprinc** kadmin
command. For example::
kadmin: getprinc user
Principal: user@KRBTEST.COM
...
Last successful authentication: [never]
Last failed authentication: Mon Dec 03 12:30:33 EST 2012
Failed password attempts: 2
...
A principal which has been locked out can be administratively unlocked
with the **-unlock** option to the **modprinc** kadmin command::
kadmin: modprinc -unlock PRINCNAME
This command will reset the number of failed attempts to 0.
KDC replication and account lockout
-----------------------------------
The account lockout state of a principal is not replicated by either
traditional :ref:`kprop(8)` or incremental propagation. Because of
this, the number of attempts an attacker can make within a time period
is multiplied by the number of KDCs. For instance, if the
**maxfailure** parameter on a policy is 10 and there are four KDCs in
the environment (a primary and three replicas), an attacker could make
as many as 40 attempts before the principal is locked out on all four
KDCs.
An administrative unlock is propagated from the primary to the replica
KDCs during the next propagation. Propagation of an administrative
unlock will cause the counter of failed attempts on each replica to
reset to 1 on the next failure.
If a KDC environment uses a replication strategy other than kprop or
incremental propagation, such as the LDAP KDB module with multi-master
LDAP replication, then account lockout state may be replicated between
KDCs and the concerns of this section may not apply.
.. _disable_lockout:
KDC performance and account lockout
-----------------------------------
In order to fully track account lockout state, the KDC must write to
the the database on each successful and failed authentication.
Writing to the database is generally more expensive than reading from
it, so these writes may have a significant impact on KDC performance.
As of release 1.9, it is possible to turn off account lockout state
tracking in order to improve performance, by setting the
**disable_last_success** and **disable_lockout** variables in the
database module subsection of :ref:`kdc.conf(5)`. For example::
[dbmodules]
DB = {
disable_last_success = true
disable_lockout = true
}
Of the two variables, setting **disable_last_success** will usually
have the largest positive impact on performance, and will still allow
account lockout policies to operate. However, it will make it
impossible to observe the last successful authentication time with
kadmin.
KDC setup and account lockout
-----------------------------
To update the account lockout state on principals, the KDC must be
able to write to the principal database. For the DB2 module, no
special setup is required. For the LDAP module, the KDC DN must be
granted write access to the principal objects. If the KDC DN has only
read access, account lockout will not function.

View file

@ -0,0 +1,100 @@
.. _otp_preauth:
OTP Preauthentication
=====================
OTP is a preauthentication mechanism for Kerberos 5 which uses One
Time Passwords (OTP) to authenticate the client to the KDC. The OTP
is passed to the KDC over an encrypted FAST channel in clear-text.
The KDC uses the password along with per-user configuration to proxy
the request to a third-party RADIUS system. This enables
out-of-the-box compatibility with a large number of already widely
deployed proprietary systems.
Additionally, our implementation of the OTP system allows for the
passing of RADIUS requests over a UNIX domain stream socket. This
permits the use of a local companion daemon which can handle the
details of authentication.
Defining token types
--------------------
Token types are defined in either :ref:`krb5.conf(5)` or
:ref:`kdc.conf(5)` according to the following format::
[otp]
<name> = {
server = <host:port or filename> (default: see below)
secret = <filename>
timeout = <integer> (default: 5 [seconds])
retries = <integer> (default: 3)
strip_realm = <boolean> (default: true)
indicator = <string> (default: none)
}
If the server field begins with '/', it will be interpreted as a UNIX
socket. Otherwise, it is assumed to be in the format host:port. When
a UNIX domain socket is specified, the secret field is optional and an
empty secret is used by default. If the server field is not
specified, it defaults to |kdcrundir|\ ``/<name>.socket``.
When forwarding the request over RADIUS, by default the principal is
used in the User-Name attribute of the RADIUS packet. The strip_realm
parameter controls whether the principal is forwarded with or without
the realm portion.
If an indicator field is present, tickets issued using this token type
will be annotated with the specified authentication indicator (see
:ref:`auth_indicator`). This key may be specified multiple times to
add multiple indicators.
The default token type
----------------------
A default token type is used internally when no token type is specified for a
given user. It is defined as follows::
[otp]
DEFAULT = {
strip_realm = false
}
The administrator may override the internal ``DEFAULT`` token type
simply by defining a configuration with the same name.
Token instance configuration
----------------------------
To enable OTP for a client principal, the administrator must define
the **otp** string attribute for that principal. (See
:ref:`set_string`.) The **otp** user string is a JSON string of the
format:
.. code-block:: xml
[{
"type": <string>,
"username": <string>,
"indicators": [<string>, ...]
}, ...]
This is an array of token objects. Both fields of token objects are
optional. The **type** field names the token type of this token; if
not specified, it defaults to ``DEFAULT``. The **username** field
specifies the value to be sent in the User-Name RADIUS attribute. If
not specified, the principal name is sent, with or without realm as
defined in the token type. The **indicators** field specifies a list
of authentication indicators to annotate tickets with, overriding any
indicators specified in the token type.
For ease of configuration, an empty array (``[]``) is treated as
equivalent to one DEFAULT token (``[{}]``).
Other considerations
--------------------
#. FAST is required for OTP to work.

View file

@ -0,0 +1,354 @@
.. _pkinit:
PKINIT configuration
====================
PKINIT is a preauthentication mechanism for Kerberos 5 which uses
X.509 certificates to authenticate the KDC to clients and vice versa.
PKINIT can also be used to enable anonymity support, allowing clients
to communicate securely with the KDC or with application servers
without authenticating as a particular client principal.
Creating certificates
---------------------
PKINIT requires an X.509 certificate for the KDC and one for each
client principal which will authenticate using PKINIT. For anonymous
PKINIT, a KDC certificate is required, but client certificates are
not. A commercially issued server certificate can be used for the KDC
certificate, but generally cannot be used for client certificates.
The instruction in this section describe how to establish a
certificate authority and create standard PKINIT certificates. Skip
this section if you are using a commercially issued server certificate
as the KDC certificate for anonymous PKINIT, or if you are configuring
a client to use an Active Directory KDC.
Generating a certificate authority certificate
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
You can establish a new certificate authority (CA) for use with a
PKINIT deployment with the commands::
openssl genrsa -out cakey.pem 2048
openssl req -key cakey.pem -new -x509 -out cacert.pem -days 3650
The second command will ask for the values of several certificate
fields. These fields can be set to any values. You can adjust the
expiration time of the CA certificate by changing the number after
``-days``. Since the CA certificate must be deployed to client
machines each time it changes, it should normally have an expiration
time far in the future; however, expiration times after 2037 may cause
interoperability issues in rare circumstances.
The result of these commands will be two files, cakey.pem and
cacert.pem. cakey.pem will contain a 2048-bit RSA private key, which
must be carefully protected. cacert.pem will contain the CA
certificate, which must be placed in the filesystems of the KDC and
each client host. cakey.pem will be required to create KDC and client
certificates.
Generating a KDC certificate
~~~~~~~~~~~~~~~~~~~~~~~~~~~~
A KDC certificate for use with PKINIT is required to have some unusual
fields, which makes generating them with OpenSSL somewhat complicated.
First, you will need a file containing the following::
[kdc_cert]
basicConstraints=CA:FALSE
keyUsage=nonRepudiation,digitalSignature,keyEncipherment,keyAgreement
extendedKeyUsage=1.3.6.1.5.2.3.5
subjectKeyIdentifier=hash
authorityKeyIdentifier=keyid,issuer
issuerAltName=issuer:copy
subjectAltName=otherName:1.3.6.1.5.2.2;SEQUENCE:kdc_princ_name
[kdc_princ_name]
realm=EXP:0,GeneralString:${ENV::REALM}
principal_name=EXP:1,SEQUENCE:kdc_principal_seq
[kdc_principal_seq]
name_type=EXP:0,INTEGER:2
name_string=EXP:1,SEQUENCE:kdc_principals
[kdc_principals]
princ1=GeneralString:krbtgt
princ2=GeneralString:${ENV::REALM}
If the above contents are placed in extensions.kdc, you can generate
and sign a KDC certificate with the following commands::
openssl genrsa -out kdckey.pem 2048
openssl req -new -out kdc.req -key kdckey.pem
env REALM=YOUR_REALMNAME openssl x509 -req -in kdc.req \
-CAkey cakey.pem -CA cacert.pem -out kdc.pem -days 365 \
-extfile extensions.kdc -extensions kdc_cert -CAcreateserial
rm kdc.req
The second command will ask for the values of certificate fields,
which can be set to any values. In the third command, substitute your
KDC's realm name for YOUR_REALMNAME. You can adjust the certificate's
expiration date by changing the number after ``-days``. Remember to
create a new KDC certificate before the old one expires.
The result of this operation will be in two files, kdckey.pem and
kdc.pem. Both files must be placed in the KDC's filesystem.
kdckey.pem, which contains the KDC's private key, must be carefully
protected.
If you examine the KDC certificate with ``openssl x509 -in kdc.pem
-text -noout``, OpenSSL will not know how to display the KDC principal
name in the Subject Alternative Name extension, so it will appear as
``othername:<unsupported>``. This is normal and does not mean
anything is wrong with the KDC certificate.
Generating client certificates
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
PKINIT client certificates also must have some unusual certificate
fields. To generate a client certificate with OpenSSL for a
single-component principal name, you will need an extensions file
(different from the KDC extensions file above) containing::
[client_cert]
basicConstraints=CA:FALSE
keyUsage=digitalSignature,keyEncipherment,keyAgreement
extendedKeyUsage=1.3.6.1.5.2.3.4
subjectKeyIdentifier=hash
authorityKeyIdentifier=keyid,issuer
issuerAltName=issuer:copy
subjectAltName=otherName:1.3.6.1.5.2.2;SEQUENCE:princ_name
[princ_name]
realm=EXP:0,GeneralString:${ENV::REALM}
principal_name=EXP:1,SEQUENCE:principal_seq
[principal_seq]
name_type=EXP:0,INTEGER:1
name_string=EXP:1,SEQUENCE:principals
[principals]
princ1=GeneralString:${ENV::CLIENT}
If the above contents are placed in extensions.client, you can
generate and sign a client certificate with the following commands::
openssl genrsa -out clientkey.pem 2048
openssl req -new -key clientkey.pem -out client.req
env REALM=YOUR_REALMNAME CLIENT=YOUR_PRINCNAME openssl x509 \
-CAkey cakey.pem -CA cacert.pem -req -in client.req \
-extensions client_cert -extfile extensions.client \
-days 365 -out client.pem
rm client.req
Normally, the first two commands should be run on the client host, and
the resulting client.req file transferred to the certificate authority
host for the third command. As in the previous steps, the second
command will ask for the values of certificate fields, which can be
set to any values. In the third command, substitute your realm's name
for YOUR_REALMNAME and the client's principal name (without realm) for
YOUR_PRINCNAME. You can adjust the certificate's expiration date by
changing the number after ``-days``.
The result of this operation will be two files, clientkey.pem and
client.pem. Both files must be present on the client's host;
clientkey.pem, which contains the client's private key, must be
protected from access by others.
As in the KDC certificate, OpenSSL will display the client principal
name as ``othername:<unsupported>`` in the Subject Alternative Name
extension of a PKINIT client certificate.
If the client principal name contains more than one component
(e.g. ``host/example.com@REALM``), the ``[principals]`` section of
``extensions.client`` must be altered to contain multiple entries.
(Simply setting ``CLIENT`` to ``host/example.com`` would generate a
certificate for ``host\/example.com@REALM`` which would not match the
multi-component principal name.) For a two-component principal, the
section should read::
[principals]
princ1=GeneralString:${ENV::CLIENT1}
princ2=GeneralString:${ENV::CLIENT2}
The environment variables ``CLIENT1`` and ``CLIENT2`` must then be set
to the first and second components when running ``openssl x509``.
Configuring the KDC
-------------------
The KDC must have filesystem access to the KDC certificate (kdc.pem)
and the KDC private key (kdckey.pem). Configure the following
relation in the KDC's :ref:`kdc.conf(5)` file, either in the
:ref:`kdcdefaults` section or in a :ref:`kdc_realms` subsection (with
appropriate pathnames)::
pkinit_identity = FILE:/var/lib/krb5kdc/kdc.pem,/var/lib/krb5kdc/kdckey.pem
If any clients will authenticate using regular (as opposed to
anonymous) PKINIT, the KDC must also have filesystem access to the CA
certificate (cacert.pem), and the following configuration (with the
appropriate pathname)::
pkinit_anchors = FILE:/var/lib/krb5kdc/cacert.pem
Because of the larger size of requests and responses using PKINIT, you
may also need to allow TCP access to the KDC::
kdc_tcp_listen = 88
Restart the :ref:`krb5kdc(8)` daemon to pick up the configuration
changes.
The principal entry for each PKINIT-using client must be configured to
require preauthentication. Ensure this with the command::
kadmin -q 'modprinc +requires_preauth YOUR_PRINCNAME'
Starting with release 1.12, it is possible to remove the long-term
keys of a principal entry, which can save some space in the database
and help to clarify some PKINIT-related error conditions by not asking
for a password::
kadmin -q 'purgekeys -all YOUR_PRINCNAME'
These principal options can also be specified at principal creation
time as follows::
kadmin -q 'add_principal +requires_preauth -nokey YOUR_PRINCNAME'
By default, the KDC requires PKINIT client certificates to have the
standard Extended Key Usage and Subject Alternative Name attributes
for PKINIT. Starting in release 1.16, it is possible to authorize
client certificates based on the subject or other criteria instead of
the standard PKINIT Subject Alternative Name, by setting the
**pkinit_cert_match** string attribute on each client principal entry.
For example::
kadmin set_string user@REALM pkinit_cert_match "<SUBJECT>CN=user@REALM$"
The **pkinit_cert_match** string attribute follows the syntax used by
the :ref:`krb5.conf(5)` **pkinit_cert_match** relation. To allow the
use of non-PKINIT client certificates, it will also be necessary to
disable key usage checking using the **pkinit_eku_checking** relation;
for example::
[kdcdefaults]
pkinit_eku_checking = none
Configuring the clients
-----------------------
Client hosts must be configured to trust the issuing authority for the
KDC certificate. For a newly established certificate authority, the
client host must have filesystem access to the CA certificate
(cacert.pem) and the following relation in :ref:`krb5.conf(5)` in the
appropriate :ref:`realms` subsection (with appropriate pathnames)::
pkinit_anchors = FILE:/etc/krb5/cacert.pem
If the KDC certificate is a commercially issued server certificate,
the issuing certificate is most likely included in a system directory.
You can specify it by filename as above, or specify the whole
directory like so::
pkinit_anchors = DIR:/etc/ssl/certs
A commercially issued server certificate will usually not have the
standard PKINIT principal name or Extended Key Usage extensions, so
the following additional configuration is required::
pkinit_eku_checking = kpServerAuth
pkinit_kdc_hostname = hostname.of.kdc.certificate
Multiple **pkinit_kdc_hostname** relations can be configured to
recognize multiple KDC certificates. If the KDC is an Active
Directory domain controller, setting **pkinit_kdc_hostname** is
necessary, but it should not be necessary to set
**pkinit_eku_checking**.
To perform regular (as opposed to anonymous) PKINIT authentication, a
client host must have filesystem access to a client certificate
(client.pem), and the corresponding private key (clientkey.pem).
Configure the following relations in the client host's
:ref:`krb5.conf(5)` file in the appropriate :ref:`realms` subsection
(with appropriate pathnames)::
pkinit_identities = FILE:/etc/krb5/client.pem,/etc/krb5/clientkey.pem
If the KDC and client are properly configured, it should now be
possible to run ``kinit username`` without entering a password.
.. _anonymous_pkinit:
Anonymous PKINIT
----------------
Anonymity support in Kerberos allows a client to obtain a ticket
without authenticating as any particular principal. Such a ticket can
be used as a FAST armor ticket, or to securely communicate with an
application server anonymously.
To configure anonymity support, you must generate or otherwise procure
a KDC certificate and configure the KDC host, but you do not need to
generate any client certificates. On the KDC, you must set the
**pkinit_identity** variable to provide the KDC certificate, but do
not need to set the **pkinit_anchors** variable or store the issuing
certificate if you won't have any client certificates to verify. On
client hosts, you must set the **pkinit_anchors** variable (and
possibly **pkinit_kdc_hostname** and **pkinit_eku_checking**) in order
to trust the issuing authority for the KDC certificate, but do not
need to set the **pkinit_identities** variable.
Anonymity support is not enabled by default. To enable it, you must
create the principal ``WELLKNOWN/ANONYMOUS`` using the command::
kadmin -q 'addprinc -randkey WELLKNOWN/ANONYMOUS'
Some Kerberos deployments include application servers which lack
proper access control, and grant some level of access to any user who
can authenticate. In such an environment, enabling anonymity support
on the KDC would present a security issue. If you need to enable
anonymity support for TGTs (for use as FAST armor tickets) without
enabling anonymous authentication to application servers, you can set
the variable **restrict_anonymous_to_tgt** to ``true`` in the
appropriate :ref:`kdc_realms` subsection of the KDC's
:ref:`kdc.conf(5)` file.
To obtain anonymous credentials on a client, run ``kinit -n``, or
``kinit -n @REALMNAME`` to specify a realm. The resulting tickets
will have the client name ``WELLKNOWN/ANONYMOUS@WELLKNOWN:ANONYMOUS``.
Freshness tokens
----------------
Freshness tokens can ensure that the client has recently had access to
its certificate private key. If freshness tokens are not required by
the KDC, a client program with temporary possession of the private key
can compose requests for future timestamps and use them later.
In release 1.17 and later, freshness tokens are supported by the
client and are sent by the KDC when the client indicates support for
them. Because not all clients support freshness tokens yet, they are
not required by default. To check if freshness tokens are supported
by a realm's clients, look in the KDC logs for the lines::
PKINIT: freshness token received from <client principal>
PKINIT: no freshness token received from <client principal>
To require freshness tokens for all clients in a realm (except for
clients authenticating anonymously), set the
**pkinit_require_freshness** variable to ``true`` in the appropriate
:ref:`kdc_realms` subsection of the KDC's :ref:`kdc.conf(5)` file. To
test that this option is in effect, run ``kinit -X disable_freshness``
and verify that authentication is unsuccessful.

View file

@ -0,0 +1,126 @@
Principal names and DNS
=======================
Kerberos clients can do DNS lookups to canonicalize service principal
names. This can cause difficulties when setting up Kerberos
application servers, especially when the client's name for the service
is different from what the service thinks its name is.
Service principal names
-----------------------
A frequently used kind of principal name is the host-based service
principal name. This kind of principal name has two components: a
service name and a hostname. For example, ``imap/imap.example.com``
is the principal name of the "imap" service on the host
"imap.example.com". Other possible service names for the first
component include "host" (remote login services such as ssh), "HTTP",
and "nfs" (Network File System).
Service administrators often publish well-known hostname aliases that
they would prefer users to use instead of the canonical name of the
service host. This gives service administrators more flexibility in
deploying services. For example, a shell login server might be named
"long-vanity-hostname.example.com", but users will naturally prefer to
type something like "login.example.com". Hostname aliases also allow
for administrators to set up load balancing for some sorts of services
based on rotating ``CNAME`` records in DNS.
Service principal canonicalization
----------------------------------
In the MIT krb5 client library, canonicalization of host-based service
principals is controlled by the **dns_canonicalize_hostname**,
**rnds**, and **qualify_shortname** variables in :ref:`libdefaults`.
If **dns_canonicalize_hostname** is set to ``true`` (the default
value), the client performs forward resolution by looking up the IPv4
and/or IPv6 addresses of the hostname using ``getaddrinfo()``. This
process will typically add a domain suffix to the hostname if needed,
and follow CNAME records in the DNS. If **rdns** is also set to
``true`` (the default), the client will then perform a reverse lookup
of the first returned Internet address using ``getnameinfo()``,
finding the name associated with the PTR record.
If **dns_canonicalize_hostname** is set to ``false``, the hostname is
not canonicalized using DNS. If the hostname has only one component
(i.e. it contains no "." characters), the host's primary DNS search
domain will be appended, if there is one. The **qualify_shortname**
variable can be used to override or disable this suffix.
If **dns_canonicalize_hostname** is set to ``fallback`` (added in
release 1.18), the hostname is initially treated according to the
rules for ``dns_canonicalize_hostname=false``. If a ticket request
fails because the service principal is unknown, the hostname will be
canonicalized according to the rules for
``dns_canonicalize_hostname=true`` and the request will be retried.
In all cases, the hostname is converted to lowercase, and any trailing
dot is removed.
Reverse DNS mismatches
----------------------
Sometimes, an enterprise will have control over its forward DNS but
not its reverse DNS. The reverse DNS is sometimes under the control
of the Internet service provider of the enterprise, and the enterprise
may not have much influence in setting up reverse DNS records for its
address space. If there are difficulties with getting forward and
reverse DNS to match, it is best to set ``rdns = false`` on client
machines.
Overriding application behavior
-------------------------------
Applications can choose to use a default hostname component in their
service principal name when accepting authentication, which avoids
some sorts of hostname mismatches. Because not all relevant
applications do this yet, using the :ref:`krb5.conf(5)` setting::
[libdefaults]
ignore_acceptor_hostname = true
will allow the Kerberos library to override the application's choice
of service principal hostname and will allow a server program to
accept incoming authentications using any key in its keytab that
matches the service name and realm name (if given). This setting
defaults to "false" and is available in releases krb5-1.10 and later.
Provisioning keytabs
--------------------
One service principal entry that should be in the keytab is a
principal whose hostname component is the canonical hostname that
``getaddrinfo()`` reports for all known aliases for the host. If the
reverse DNS information does not match this canonical hostname, an
additional service principal entry should be in the keytab for this
different hostname.
Specific application advice
---------------------------
Secure shell (ssh)
~~~~~~~~~~~~~~~~~~
Setting ``GSSAPIStrictAcceptorCheck = no`` in the configuration file
of modern versions of the openssh daemon will allow the daemon to try
any key in its keytab when accepting a connection, rather than looking
for the keytab entry that matches the host's own idea of its name
(typically the name that ``gethostname()`` returns). This requires
krb5-1.10 or later.
OpenLDAP (ldapsearch, etc.)
~~~~~~~~~~~~~~~~~~~~~~~~~~~
OpenLDAP's SASL implementation performs reverse DNS lookup in order to
canonicalize service principal names, even if **rdns** is set to
``false`` in the Kerberos configuration. To disable this behavior,
add ``SASL_NOCANON on`` to ``ldap.conf``, or set the
``LDAPSASL_NOCANON`` environment variable.

View file

@ -0,0 +1,268 @@
Realm configuration decisions
=============================
Before installing Kerberos V5, it is necessary to consider the
following issues:
* The name of your Kerberos realm (or the name of each realm, if you
need more than one).
* How you will assign your hostnames to Kerberos realms.
* Which ports your KDC and and kadmind services will use, if they will
not be using the default ports.
* How many replica KDCs you need and where they should be located.
* The hostnames of your primary and replica KDCs.
* How frequently you will propagate the database from the primary KDC
to the replica KDCs.
Realm name
----------
Although your Kerberos realm can be any ASCII string, convention is to
make it the same as your domain name, in upper-case letters.
For example, hosts in the domain ``example.com`` would be in the
Kerberos realm::
EXAMPLE.COM
If you need multiple Kerberos realms, MIT recommends that you use
descriptive names which end with your domain name, such as::
BOSTON.EXAMPLE.COM
HOUSTON.EXAMPLE.COM
.. _mapping_hostnames:
Mapping hostnames onto Kerberos realms
--------------------------------------
Mapping hostnames onto Kerberos realms is done in one of three ways.
The first mechanism works through a set of rules in the
:ref:`domain_realm` section of :ref:`krb5.conf(5)`. You can specify
mappings for an entire domain or on a per-hostname basis. Typically
you would do this by specifying the mappings for a given domain or
subdomain and listing the exceptions.
The second mechanism is to use KDC host-based service referrals. With
this method, the KDC's krb5.conf has a full [domain_realm] mapping for
hosts, but the clients do not, or have mappings for only a subset of
the hosts they might contact. When a client needs to contact a server
host for which it has no mapping, it will ask the client realm's KDC
for the service ticket, and will receive a referral to the appropriate
service realm.
To use referrals, clients must be running MIT krb5 1.6 or later, and
the KDC must be running MIT krb5 1.7 or later. The
**host_based_services** and **no_host_referral** variables in the
:ref:`kdc_realms` section of :ref:`kdc.conf(5)` can be used to
fine-tune referral behavior on the KDC.
It is also possible for clients to use DNS TXT records, if
**dns_lookup_realm** is enabled in :ref:`krb5.conf(5)`. Such lookups
are disabled by default because DNS is an insecure protocol and security
holes could result if DNS records are spoofed. If enabled, the client
will try to look up a TXT record formed by prepending the prefix
``_kerberos`` to the hostname in question. If that record is not
found, the client will attempt a lookup by prepending ``_kerberos`` to the
host's domain name, then its parent domain, up to the top-level domain.
For the hostname ``boston.engineering.example.com``, the names looked up
would be::
_kerberos.boston.engineering.example.com
_kerberos.engineering.example.com
_kerberos.example.com
_kerberos.com
The value of the first TXT record found is taken as the realm name.
Even if you do not choose to use this mechanism within your site,
you may wish to set it up anyway, for use when interacting with other sites.
Ports for the KDC and admin services
------------------------------------
The default ports used by Kerberos are port 88 for the KDC and port
749 for the admin server. You can, however, choose to run on other
ports, as long as they are specified in each host's
:ref:`krb5.conf(5)` files or in DNS SRV records, and the
:ref:`kdc.conf(5)` file on each KDC. For a more thorough treatment of
port numbers used by the Kerberos V5 programs, refer to the
:ref:`conf_firewall`.
Replica KDCs
------------
Replica KDCs provide an additional source of Kerberos ticket-granting
services in the event of inaccessibility of the primary KDC. The
number of replica KDCs you need and the decision of where to place them,
both physically and logically, depends on the specifics of your
network.
Kerberos authentication requires that each client be able to contact a
KDC. Therefore, you need to anticipate any likely reason a KDC might
be unavailable and have a replica KDC to take up the slack.
Some considerations include:
* Have at least one replica KDC as a backup, for when the primary KDC
is down, is being upgraded, or is otherwise unavailable.
* If your network is split such that a network outage is likely to
cause a network partition (some segment or segments of the network
to become cut off or isolated from other segments), have a replica
KDC accessible to each segment.
* If possible, have at least one replica KDC in a different building
from the primary, in case of power outages, fires, or other
localized disasters.
.. _kdc_hostnames:
Hostnames for KDCs
------------------
MIT recommends that your KDCs have a predefined set of CNAME records
(DNS hostname aliases), such as ``kerberos`` for the primary KDC and
``kerberos-1``, ``kerberos-2``, ... for the replica KDCs. This way,
if you need to swap a machine, you only need to change a DNS entry,
rather than having to change hostnames.
As of MIT krb5 1.4, clients can locate a realm's KDCs through DNS
using SRV records (:rfc:`2782`), assuming the Kerberos realm name is
also a DNS domain name. These records indicate the hostname and port
number to contact for that service, optionally with weighting and
prioritization. The domain name used in the SRV record name is the
realm name. Several different Kerberos-related service names are
used:
_kerberos._udp
This is for contacting any KDC by UDP. This entry will be used
the most often. Normally you should list port 88 on each of your
KDCs.
_kerberos._tcp
This is for contacting any KDC by TCP. Normally you should use
port 88. This entry should be omitted if the KDC does not listen
on TCP ports, as was the default prior to release 1.13.
_kerberos-master._udp
This entry should refer to those KDCs, if any, that will
immediately see password changes to the Kerberos database. If a
user is logging in and the password appears to be incorrect, the
client will retry with the primary KDC before failing with an
"incorrect password" error given.
If you have only one KDC, or for whatever reason there is no
accessible KDC that would get database changes faster than the
others, you do not need to define this entry.
_kerberos-adm._tcp
This should list port 749 on your primary KDC. Support for it is
not complete at this time, but it will eventually be used by the
:ref:`kadmin(1)` program and related utilities. For now, you will
also need the **admin_server** variable in :ref:`krb5.conf(5)`.
_kerberos-master._tcp
The corresponding TCP port for _kerberos-master._udp, assuming the
primary KDC listens on a TCP port.
_kpasswd._udp
This entry should list port 464 on your primary KDC. It is used
when a user changes her password. If this entry is not defined
but a _kerberos-adm._tcp entry is defined, the client will use the
_kerberos-adm._tcp entry with the port number changed to 464.
_kpasswd._tcp
The corresponding TCP port for _kpasswd._udp.
The DNS SRV specification requires that the hostnames listed be the
canonical names, not aliases. So, for example, you might include the
following records in your (BIND-style) zone file::
$ORIGIN foobar.com.
_kerberos TXT "FOOBAR.COM"
kerberos CNAME daisy
kerberos-1 CNAME use-the-force-luke
kerberos-2 CNAME bunny-rabbit
_kerberos._udp SRV 0 0 88 daisy
SRV 0 0 88 use-the-force-luke
SRV 0 0 88 bunny-rabbit
_kerberos-master._udp SRV 0 0 88 daisy
_kerberos-adm._tcp SRV 0 0 749 daisy
_kpasswd._udp SRV 0 0 464 daisy
Clients can also be configured with the explicit location of services
using the **kdc**, **master_kdc**, **admin_server**, and
**kpasswd_server** variables in the :ref:`realms` section of
:ref:`krb5.conf(5)`. Even if some clients will be configured with
explicit server locations, providing SRV records will still benefit
unconfigured clients, and be useful for other sites.
.. _kdc_discovery:
KDC Discovery
-------------
As of MIT krb5 1.15, clients can also locate KDCs in DNS through URI
records (:rfc:`7553`). Limitations with the SRV record format may
result in extra DNS queries in situations where a client must failover
to other transport types, or find a primary server. The URI record
can convey more information about a realm's KDCs with a single query.
The client performs a query for the following URI records:
* ``_kerberos.REALM`` for finding KDCs.
* ``_kerberos-adm.REALM`` for finding kadmin services.
* ``_kpasswd.REALM`` for finding password services.
The URI record includes a priority, weight, and a URI string that
consists of case-insensitive colon separated fields, in the form
``scheme:[flags]:transport:residual``.
* *scheme* defines the registered URI type. It should always be
``krb5srv``.
* *flags* contains zero or more flag characters. Currently the only
valid flag is ``m``, which indicates that the record is for a
primary server.
* *transport* defines the transport type of the residual URL or
address. Accepted values are ``tcp``, ``udp``, or ``kkdcp`` for the
MS-KKDCP type.
* *residual* contains the hostname, IP address, or URL to be
contacted using the specified transport, with an optional port
extension. The MS-KKDCP transport type uses a HTTPS URL, and can
include a port and/or path extension.
An example of URI records in a zone file::
_kerberos.EXAMPLE.COM URI 10 1 krb5srv:m:tcp:kdc1.example.com
URI 20 1 krb5srv:m:udp:kdc2.example.com:89
URI 40 1 krb5srv::udp:10.10.0.23
URI 30 1 krb5srv::kkdcp:https://proxy:89/auth
URI lookups are enabled by default, and can be disabled by setting
**dns_uri_lookup** in the :ref:`libdefaults` section of
:ref:`krb5.conf(5)` to False. When enabled, URI lookups take
precedence over SRV lookups, falling back to SRV lookups if no URI
records are found.
.. _db_prop:
Database propagation
--------------------
The Kerberos database resides on the primary KDC, and must be
propagated regularly (usually by a cron job) to the replica KDCs. In
deciding how frequently the propagation should happen, you will need
to balance the amount of time the propagation takes against the
maximum reasonable amount of time a user should have to wait for a
password change to take effect.
If the propagation time is longer than this maximum reasonable time
(e.g., you have a particularly large database, you have a lot of
replicas, or you experience frequent network delays), you may wish to
cut down on your propagation delay by performing the propagation in
parallel. To do this, have the primary KDC propagate the database to
one set of replicas, and then have each of these replicas propagate
the database to additional replicas.
See also :ref:`incr_db_prop`

View file

@ -0,0 +1,56 @@
.. _spake:
SPAKE Preauthentication
=======================
SPAKE preauthentication (added in release 1.17) uses public key
cryptography techniques to protect against :ref:`password dictionary
attacks <dictionary>`. Unlike :ref:`PKINIT <pkinit>`, it does not
require any additional infrastructure such as certificates; it simply
needs to be turned on. Using SPAKE preauthentication may modestly
increase the CPU and network load on the KDC.
SPAKE preauthentication can use one of four elliptic curve groups for
its password-authenticated key exchange. The recommended group is
``edwards25519``; three NIST curves (``P-256``, ``P-384``, and
``P-521``) are also supported.
By default, SPAKE with the ``edwards25519`` group is enabled on
clients, but the KDC does not offer SPAKE by default. To turn it on,
set the **spake_preauth_groups** variable in :ref:`libdefaults` to a
list of allowed groups. This variable affects both the client and the
KDC. Simply setting it to ``edwards25519`` is recommended::
[libdefaults]
spake_preauth_groups = edwards25519
Set the **+requires_preauth** and **-allow_svr** flags on client
principal entries, as you would for any preauthentication mechanism::
kadmin: modprinc +requires_preauth -allow_svr PRINCNAME
Clients which do not implement SPAKE preauthentication will fall back
to encrypted timestamp.
An active attacker can force a fallback to encrypted timestamp by
modifying the initial KDC response, defeating the protection against
dictionary attacks. To prevent this fallback on clients which do
implement SPAKE preauthentication, set the
**disable_encrypted_timestamp** variable to ``true`` in the
:ref:`realms` subsection for realms whose KDCs offer SPAKE
preauthentication.
By default, SPAKE preauthentication requires an extra network round
trip to the KDC during initial authentication. If most of the clients
in a realm support SPAKE, this extra round trip can be eliminated
using an optimistic challenge, by setting the
**spake_preauth_kdc_challenge** variable in :ref:`kdcdefaults` to a
single group name::
[kdcdefaults]
spake_preauth_kdc_challenge = edwards25519
Using optimistic challenge will cause the KDC to do extra work for
initial authentication requests that do not result in SPAKE
preauthentication, but will save work when SPAKE preauthentication is
used.

View file

@ -0,0 +1,135 @@
.. _troubleshoot:
Troubleshooting
===============
.. _trace_logging:
Trace logging
-------------
Most programs using MIT krb5 1.9 or later can be made to provide
information about internal krb5 library operations using trace
logging. To enable this, set the **KRB5_TRACE** environment variable
to a filename before running the program. On many operating systems,
the filename ``/dev/stdout`` can be used to send trace logging output
to standard output.
Some programs do not honor **KRB5_TRACE**, either because they use
secure library contexts (this generally applies to setuid programs and
parts of the login system) or because they take direct control of the
trace logging system using the API.
Here is a short example showing trace logging output for an invocation
of the :ref:`kvno(1)` command::
shell% env KRB5_TRACE=/dev/stdout kvno krbtgt/KRBTEST.COM
[9138] 1332348778.823276: Getting credentials user@KRBTEST.COM ->
krbtgt/KRBTEST.COM@KRBTEST.COM using ccache
FILE:/me/krb5/build/testdir/ccache
[9138] 1332348778.823381: Retrieving user@KRBTEST.COM ->
krbtgt/KRBTEST.COM@KRBTEST.COM from
FILE:/me/krb5/build/testdir/ccache with result: 0/Unknown code 0
krbtgt/KRBTEST.COM@KRBTEST.COM: kvno = 1
List of errors
--------------
Frequently seen errors
~~~~~~~~~~~~~~~~~~~~~~
#. :ref:`init_creds_ETYPE_NOSUPP`
#. :ref:`cert_chain_ETYPE_NOSUPP`
#. :ref:`err_cert_chain_cert_expired`
Errors seen by admins
~~~~~~~~~~~~~~~~~~~~~
.. _prop_failed_start:
#. :ref:`kprop_no_route`
#. :ref:`kprop_con_refused`
#. :ref:`kprop_sendauth_exchange`
.. _prop_failed_end:
-----
.. _init_creds_etype_nosupp:
KDC has no support for encryption type while getting initial credentials
........................................................................
.. _cert_chain_etype_nosupp:
credential verification failed: KDC has no support for encryption type
......................................................................
This most commonly happens when trying to use a principal with only
DES keys, in a release (MIT krb5 1.7 or later) which disables DES by
default. DES encryption is considered weak due to its inadequate key
size. If you cannot migrate away from its use, you can re-enable DES
by adding ``allow_weak_crypto = true`` to the :ref:`libdefaults`
section of :ref:`krb5.conf(5)`.
.. _err_cert_chain_cert_expired:
Cannot create cert chain: certificate has expired
.................................................
This error message indicates that PKINIT authentication failed because
the client certificate, KDC certificate, or one of the certificates in
the signing chain above them has expired.
If the KDC certificate has expired, this message appears in the KDC
log file, and the client will receive a "Preauthentication failed"
error. (Prior to release 1.11, the KDC log file message erroneously
appears as "Out of memory". Prior to release 1.12, the client will
receive a "Generic error".)
If the client or a signing certificate has expired, this message may
appear in trace_logging_ output from :ref:`kinit(1)` or, starting in
release 1.12, as an error message from kinit or another program which
gets initial tickets. The error message is more likely to appear
properly on the client if the principal entry has no long-term keys.
.. _kprop_no_route:
kprop: No route to host while connecting to server
..................................................
Make sure that the hostname of the replica KDC (as given to kprop) is
correct, and that any firewalls between the primary and the replica
allow a connection on port 754.
.. _kprop_con_refused:
kprop: Connection refused while connecting to server
....................................................
If the replica KDC is intended to run kpropd out of inetd, make sure
that inetd is configured to accept krb5_prop connections. inetd may
need to be restarted or sent a SIGHUP to recognize the new
configuration. If the replica is intended to run kpropd in standalone
mode, make sure that it is running.
.. _kprop_sendauth_exchange:
kprop: Server rejected authentication (during sendauth exchange) while authenticating to server
...............................................................................................
Make sure that:
#. The time is synchronized between the primary and replica KDCs.
#. The master stash file was copied from the primary to the expected
location on the replica.
#. The replica has a keytab file in the default location containing a
``host`` principal for the replica's hostname.

View file

@ -0,0 +1,27 @@
Various links
=============
Whitepapers
-----------
#. https://kerberos.org/software/whitepapers.html
Tutorials
---------
#. Fulvio Ricciardi <https://www.kerberos.org/software/tutorial.html>_
Troubleshooting
---------------
#. https://wiki.ncsa.illinois.edu/display/ITS/Windows+Kerberos+Troubleshooting
#. https://www.shrubbery.net/solaris9ab/SUNWaadm/SYSADV6/p27.html
#. https://docs.oracle.com/cd/E19253-01/816-4557/trouble-1/index.html
#. https://docs.microsoft.com/en-us/previous-versions/tn-archive/bb463167(v=technet.10)#EBAA
#. https://bugs.launchpad.net/ubuntu/+source/libpam-heimdal/+bug/86528

View file

@ -0,0 +1,727 @@
Developing with GSSAPI
======================
The GSSAPI (Generic Security Services API) allows applications to
communicate securely using Kerberos 5 or other security mechanisms.
We recommend using the GSSAPI (or a higher-level framework which
encompasses GSSAPI, such as SASL) for secure network communication
over using the libkrb5 API directly.
GSSAPIv2 is specified in :rfc:`2743` and :rfc:`2744`. Also see
:rfc:`7546` for a description of how to use the GSSAPI in a client or
server program.
This documentation will describe how various ways of using the
GSSAPI will behave with the krb5 mechanism as implemented in MIT krb5,
as well as krb5-specific extensions to the GSSAPI.
Name types
----------
A GSSAPI application can name a local or remote entity by calling
gss_import_name_, specifying a name type and a value. The following
name types are supported by the krb5 mechanism:
* **GSS_C_NT_HOSTBASED_SERVICE**: The value should be a string of the
form ``service`` or ``service@hostname``. This is the most common
way to name target services when initiating a security context, and
is the most likely name type to work across multiple mechanisms.
* **GSS_KRB5_NT_PRINCIPAL_NAME**: The value should be a principal name
string. This name type only works with the krb5 mechanism, and is
defined in the ``<gssapi/gssapi_krb5.h>`` header.
* **GSS_C_NT_USER_NAME** or **GSS_C_NULL_OID**: The value is treated
as an unparsed principal name string, as above. These name types
may work with mechanisms other than krb5, but will have different
interpretations in those mechanisms. **GSS_C_NT_USER_NAME** is
intended to be used with a local username, which will parse into a
single-component principal in the default realm.
* **GSS_C_NT_ANONYMOUS**: The value is ignored. The anonymous
principal is used, allowing a client to authenticate to a server
without asserting a particular identity (which may or may not be
allowed by a particular server or Kerberos realm).
* **GSS_C_NT_MACHINE_UID_NAME**: The value is uid_t object. On
Unix-like systems, the username of the uid is looked up in the
system user database and the resulting username is parsed as a
principal name.
* **GSS_C_NT_STRING_UID_NAME**: As above, but the value is a decimal
string representation of the uid.
* **GSS_C_NT_EXPORT_NAME**: The value must be the result of a
gss_export_name_ call.
* **GSS_KRB5_NT_ENTERPRISE_NAME**: The value should be a krb5
enterprise name string (see :rfc:`6806` section 5), in the form
``user@suffix``. This name type is used to convey alias names, and
is defined in the ``<gssapi/gssapi_krb5.h>`` header. (New in
release 1.17.)
* **GSS_KRB5_NT_X509_CERT**: The value should be an X.509 certificate
encoded according to :rfc:`5280`. This name form can be used for
the desired_name parameter of gss_acquire_cred_impersonate_name(),
to identify the S4U2Self user by certificate. (New in release
1.19.)
Initiator credentials
---------------------
A GSSAPI client application uses gss_init_sec_context_ to establish a
security context. The *initiator_cred_handle* parameter determines
what tickets are used to establish the connection. An application can
either pass **GSS_C_NO_CREDENTIAL** to use the default client
credential, or it can use gss_acquire_cred_ beforehand to acquire an
initiator credential. The call to gss_acquire_cred_ may include a
*desired_name* parameter, or it may pass **GSS_C_NO_NAME** if it does
not have a specific name preference.
If the desired name for a krb5 initiator credential is a host-based
name, it is converted to a principal name of the form
``service/hostname`` in the local realm, where *hostname* is the local
hostname if not specified. The hostname will be canonicalized using
forward name resolution, and possibly also using reverse name
resolution depending on the value of the **rdns** variable in
:ref:`libdefaults`.
If a desired name is specified in the call to gss_acquire_cred_, the
krb5 mechanism will attempt to find existing tickets for that client
principal name in the default credential cache or collection. If the
default cache type does not support a collection, and the default
cache contains credentials for a different principal than the desired
name, a **GSS_S_CRED_UNAVAIL** error will be returned with a minor
code indicating a mismatch.
If no existing tickets are available for the desired name, but the
name has an entry in the default client :ref:`keytab_definition`, the
krb5 mechanism will acquire initial tickets for the name using the
default client keytab.
If no desired name is specified, credential acquisition will be
deferred until the credential is used in a call to
gss_init_sec_context_ or gss_inquire_cred_. If the call is to
gss_init_sec_context_, the target name will be used to choose a client
principal name using the credential cache selection facility. (This
facility might, for instance, try to choose existing tickets for a
client principal in the same realm as the target service). If there
are no existing tickets for the chosen principal, but it is present in
the default client keytab, the krb5 mechanism will acquire initial
tickets using the keytab.
If the target name cannot be used to select a client principal
(because the credentials are used in a call to gss_inquire_cred_), or
if the credential cache selection facility cannot choose a principal
for it, the default credential cache will be selected if it exists and
contains tickets.
If the default credential cache does not exist, but the default client
keytab does, the krb5 mechanism will try to acquire initial tickets
for the first principal in the default client keytab.
If the krb5 mechanism acquires initial tickets using the default
client keytab, the resulting tickets will be stored in the default
cache or collection, and will be refreshed by future calls to
gss_acquire_cred_ as they approach their expire time.
Acceptor names
--------------
A GSSAPI server application uses gss_accept_sec_context_ to establish
a security context based on tokens provided by the client. The
*acceptor_cred_handle* parameter determines what
:ref:`keytab_definition` entries may be authenticated to by the
client, if the krb5 mechanism is used.
The simplest choice is to pass **GSS_C_NO_CREDENTIAL** as the acceptor
credential. In this case, clients may authenticate to any service
principal in the default keytab (typically |keytab|, or the value of
the **KRB5_KTNAME** environment variable). This is the recommended
approach if the server application has no specific requirements to the
contrary.
A server may acquire an acceptor credential with gss_acquire_cred_ and
a *cred_usage* of **GSS_C_ACCEPT** or **GSS_C_BOTH**. If the
*desired_name* parameter is **GSS_C_NO_NAME**, then clients will be
allowed to authenticate to any service principal in the default
keytab, just as if no acceptor credential was supplied.
If a server wishes to specify a *desired_name* to gss_acquire_cred_,
the most common choice is a host-based name. If the host-based
*desired_name* contains just a *service*, then clients will be allowed
to authenticate to any host-based service principal (that is, a
principal of the form ``service/hostname@REALM``) for the named
service, regardless of hostname or realm, as long as it is present in
the default keytab. If the input name contains both a *service* and a
*hostname*, clients will be allowed to authenticate to any host-based
principal for the named service and hostname, regardless of realm.
.. note::
If a *hostname* is specified, it will be canonicalized
using forward name resolution, and possibly also using
reverse name resolution depending on the value of the
**rdns** variable in :ref:`libdefaults`.
.. note::
If the **ignore_acceptor_hostname** variable in
:ref:`libdefaults` is enabled, then *hostname* will be
ignored even if one is specified in the input name.
.. note::
In MIT krb5 versions prior to 1.10, and in Heimdal's
implementation of the krb5 mechanism, an input name with
just a *service* is treated like an input name of
``service@localhostname``, where *localhostname* is the
string returned by gethostname().
If the *desired_name* is a krb5 principal name or a local system name
type which is mapped to a krb5 principal name, clients will only be
allowed to authenticate to that principal in the default keytab.
Name Attributes
---------------
In release 1.8 or later, the gss_inquire_name_ and
gss_get_name_attribute_ functions, specified in :rfc:`6680`, can be
used to retrieve name attributes from the *src_name* returned by
gss_accept_sec_context_. The following attributes are defined when
the krb5 mechanism is used:
.. _gssapi_authind_attr:
* "auth-indicators" attribute:
This attribute will be included in the gss_inquire_name_ output if the
ticket contains :ref:`authentication indicators <auth_indicator>`.
One indicator is returned per invocation of gss_get_name_attribute_,
so multiple invocations may be necessary to retrieve all of the
indicators from the ticket. (New in release 1.15.)
Credential store extensions
---------------------------
Beginning with release 1.11, the following GSSAPI extensions declared
in ``<gssapi/gssapi_ext.h>`` can be used to specify how credentials
are acquired or stored::
struct gss_key_value_element_struct {
const char *key;
const char *value;
};
typedef struct gss_key_value_element_struct gss_key_value_element_desc;
struct gss_key_value_set_struct {
OM_uint32 count;
gss_key_value_element_desc *elements;
};
typedef const struct gss_key_value_set_struct gss_key_value_set_desc;
typedef const gss_key_value_set_desc *gss_const_key_value_set_t;
OM_uint32 gss_acquire_cred_from(OM_uint32 *minor_status,
const gss_name_t desired_name,
OM_uint32 time_req,
const gss_OID_set desired_mechs,
gss_cred_usage_t cred_usage,
gss_const_key_value_set_t cred_store,
gss_cred_id_t *output_cred_handle,
gss_OID_set *actual_mechs,
OM_uint32 *time_rec);
OM_uint32 gss_store_cred_into(OM_uint32 *minor_status,
gss_cred_id_t input_cred_handle,
gss_cred_usage_t cred_usage,
const gss_OID desired_mech,
OM_uint32 overwrite_cred,
OM_uint32 default_cred,
gss_const_key_value_set_t cred_store,
gss_OID_set *elements_stored,
gss_cred_usage_t *cred_usage_stored);
The additional *cred_store* parameter allows the caller to specify
information about how the credentials should be obtained and stored.
The following options are supported by the krb5 mechanism:
* **ccache**: For acquiring initiator credentials, the name of the
:ref:`credential cache <ccache_definition>` to which the handle will
refer. For storing credentials, the name of the cache or collection
where the credentials will be stored (see below).
* **client_keytab**: For acquiring initiator credentials, the name of
the :ref:`keytab <keytab_definition>` which will be used, if
necessary, to refresh the credentials in the cache.
* **keytab**: For acquiring acceptor credentials, the name of the
:ref:`keytab <keytab_definition>` to which the handle will refer.
In release 1.19 and later, this option also determines the keytab to
be used for verification when initiator credentials are acquired
using a password and verified.
* **password**: For acquiring initiator credentials, this option
instructs the mechanism to acquire fresh credentials into a unique
memory credential cache. This option may not be used with the
**ccache** or **client_keytab** options, and a *desired_name* must
be specified. (New in release 1.19.)
* **rcache**: For acquiring acceptor credentials, the name of the
:ref:`replay cache <rcache_definition>` to be used when processing
the initiator tokens. (New in release 1.13.)
* **verify**: For acquiring initiator credentials, this option
instructs the mechanism to verify the credentials by obtaining a
ticket to a service with a known key. The service key is obtained
from the keytab specified with the **keytab** option or the default
keytab. The value may be the name of a principal in the keytab, or
the empty string. If the empty string is given, any ``host``
service principal in the keytab may be used. (New in release 1.19.)
In release 1.20 or later, if a collection name is specified for
**cache** in a call to gss_store_cred_into(), an existing cache for
the client principal within the collection will be selected, or a new
cache will be created within the collection. If *overwrite_cred* is
false and the selected credential cache already exists, a
**GSS_S_DUPLICATE_ELEMENT** error will be returned. If *default_cred*
is true, the primary cache of the collection will be switched to the
selected cache.
Importing and exporting credentials
-----------------------------------
The following GSSAPI extensions can be used to import and export
credentials (declared in ``<gssapi/gssapi_ext.h>``)::
OM_uint32 gss_export_cred(OM_uint32 *minor_status,
gss_cred_id_t cred_handle,
gss_buffer_t token);
OM_uint32 gss_import_cred(OM_uint32 *minor_status,
gss_buffer_t token,
gss_cred_id_t *cred_handle);
The first function serializes a GSSAPI credential handle into a
buffer; the second unseralizes a buffer into a GSSAPI credential
handle. Serializing a credential does not destroy it. If any of the
mechanisms used in *cred_handle* do not support serialization,
gss_export_cred will return **GSS_S_UNAVAILABLE**. As with other
GSSAPI serialization functions, these extensions are only intended to
work with a matching implementation on the other side; they do not
serialize credentials in a standardized format.
A serialized credential may contain secret information such as ticket
session keys. The serialization format does not protect this
information from eavesdropping or tampering. The calling application
must take care to protect the serialized credential when communicating
it over an insecure channel or to an untrusted party.
A krb5 GSSAPI credential may contain references to a credential cache,
a client keytab, an acceptor keytab, and a replay cache. These
resources are normally serialized as references to their external
locations (such as the filename of the credential cache). Because of
this, a serialized krb5 credential can only be imported by a process
with similar privileges to the exporter. A serialized credential
should not be trusted if it originates from a source with lower
privileges than the importer, as it may contain references to external
credential cache, keytab, or replay cache resources not accessible to
the originator.
An exception to the above rule applies when a krb5 GSSAPI credential
refers to a memory credential cache, as is normally the case for
delegated credentials received by gss_accept_sec_context_. In this
case, the contents of the credential cache are serialized, so that the
resulting token may be imported even if the original memory credential
cache no longer exists.
Constrained delegation (S4U)
----------------------------
The Microsoft S4U2Self and S4U2Proxy Kerberos protocol extensions
allow an intermediate service to acquire credentials from a client to
a target service without requiring the client to delegate a
ticket-granting ticket, if the KDC is configured to allow it.
To perform a constrained delegation operation, the intermediate
service must submit to the KDC an "evidence ticket" from the client to
the intermediate service. An evidence ticket can be acquired when the
client authenticates to the intermediate service with Kerberos, or
with an S4U2Self request if the KDC allows it. The MIT krb5 GSSAPI
library represents an evidence ticket using a "proxy credential",
which is a special kind of gss_cred_id_t object whose underlying
credential cache contains the evidence ticket and a krbtgt ticket for
the intermediate service.
To acquire a proxy credential during client authentication, the
service should first create an acceptor credential using the
**GSS_C_BOTH** usage. The application should then pass this
credential as the *acceptor_cred_handle* to gss_accept_sec_context_,
and also pass a *delegated_cred_handle* output parameter to receive a
proxy credential containing the evidence ticket. The output value of
*delegated_cred_handle* may be a delegated ticket-granting ticket if
the client sent one, or a proxy credential if not. If the library can
determine that the client's ticket is not a valid evidence ticket, it
will place **GSS_C_NO_CREDENTIAL** in *delegated_cred_handle*.
To acquire a proxy credential using an S4U2Self request, the service
can use the following GSSAPI extension::
OM_uint32 gss_acquire_cred_impersonate_name(OM_uint32 *minor_status,
gss_cred_id_t icred,
gss_name_t desired_name,
OM_uint32 time_req,
gss_OID_set desired_mechs,
gss_cred_usage_t cred_usage,
gss_cred_id_t *output_cred,
gss_OID_set *actual_mechs,
OM_uint32 *time_rec);
The parameters to this function are similar to those of
gss_acquire_cred_, except that *icred* is used to make an S4U2Self
request to the KDC for a ticket from *desired_name* to the
intermediate service. Both *icred* and *desired_name* are required
for this function; passing **GSS_C_NO_CREDENTIAL** or
**GSS_C_NO_NAME** will cause the call to fail. *icred* must contain a
krbtgt ticket for the intermediate service. The result of this
operation is a proxy credential. (Prior to release 1.18, the result
of this operation may be a regular credential for *desired_name*, if
the KDC issues a non-forwardable ticket.)
Once the intermediate service has a proxy credential, it can simply
pass it to gss_init_sec_context_ as the *initiator_cred_handle*
parameter, and the desired service as the *target_name* parameter.
The GSSAPI library will present the krbtgt ticket and evidence ticket
in the proxy credential to the KDC in an S4U2Proxy request; if the
intermediate service has the appropriate permissions, the KDC will
issue a ticket from the client to the target service. The GSSAPI
library will then use this ticket to authenticate to the target
service.
If an application needs to find out whether a credential it holds is a
proxy credential and the name of the intermediate service, it can
query the credential with the **GSS_KRB5_GET_CRED_IMPERSONATOR** OID
(new in release 1.16, declared in ``<gssapi/gssapi_krb5.h>``) using
the gss_inquire_cred_by_oid extension (declared in
``<gssapi/gssapi_ext.h>``)::
OM_uint32 gss_inquire_cred_by_oid(OM_uint32 *minor_status,
const gss_cred_id_t cred_handle,
gss_OID desired_object,
gss_buffer_set_t *data_set);
If the call succeeds and *cred_handle* is a proxy credential,
*data_set* will be set to a single-element buffer set containing the
unparsed principal name of the intermediate service. If *cred_handle*
is not a proxy credential, *data_set* will be set to an empty buffer
set. If the library does not support the query,
gss_inquire_cred_by_oid will return **GSS_S_UNAVAILABLE**.
AEAD message wrapping
---------------------
The following GSSAPI extensions (declared in
``<gssapi/gssapi_ext.h>``) can be used to wrap and unwrap messages
with additional "associated data" which is integrity-checked but is
not included in the output buffer::
OM_uint32 gss_wrap_aead(OM_uint32 *minor_status,
gss_ctx_id_t context_handle,
int conf_req_flag, gss_qop_t qop_req,
gss_buffer_t input_assoc_buffer,
gss_buffer_t input_payload_buffer,
int *conf_state,
gss_buffer_t output_message_buffer);
OM_uint32 gss_unwrap_aead(OM_uint32 *minor_status,
gss_ctx_id_t context_handle,
gss_buffer_t input_message_buffer,
gss_buffer_t input_assoc_buffer,
gss_buffer_t output_payload_buffer,
int *conf_state,
gss_qop_t *qop_state);
Wrap tokens created with gss_wrap_aead will successfully unwrap only
if the same *input_assoc_buffer* contents are presented to
gss_unwrap_aead.
IOV message wrapping
--------------------
The following extensions (declared in ``<gssapi/gssapi_ext.h>``) can
be used for in-place encryption, fine-grained control over wrap token
layout, and for constructing wrap tokens compatible with Microsoft DCE
RPC::
typedef struct gss_iov_buffer_desc_struct {
OM_uint32 type;
gss_buffer_desc buffer;
} gss_iov_buffer_desc, *gss_iov_buffer_t;
OM_uint32 gss_wrap_iov(OM_uint32 *minor_status,
gss_ctx_id_t context_handle,
int conf_req_flag, gss_qop_t qop_req,
int *conf_state,
gss_iov_buffer_desc *iov, int iov_count);
OM_uint32 gss_unwrap_iov(OM_uint32 *minor_status,
gss_ctx_id_t context_handle,
int *conf_state, gss_qop_t *qop_state,
gss_iov_buffer_desc *iov, int iov_count);
OM_uint32 gss_wrap_iov_length(OM_uint32 *minor_status,
gss_ctx_id_t context_handle,
int conf_req_flag,
gss_qop_t qop_req, int *conf_state,
gss_iov_buffer_desc *iov,
int iov_count);
OM_uint32 gss_release_iov_buffer(OM_uint32 *minor_status,
gss_iov_buffer_desc *iov,
int iov_count);
The caller of gss_wrap_iov provides an array of gss_iov_buffer_desc
structures, each containing a type and a gss_buffer_desc structure.
Valid types include:
* **GSS_C_BUFFER_TYPE_DATA**: A data buffer to be included in the
token, and to be encrypted or decrypted in-place if the token is
confidentiality-protected.
* **GSS_C_BUFFER_TYPE_HEADER**: The GSSAPI wrap token header and
underlying cryptographic header.
* **GSS_C_BUFFER_TYPE_TRAILER**: The cryptographic trailer, if one is
required.
* **GSS_C_BUFFER_TYPE_PADDING**: Padding to be combined with the data
during encryption and decryption. (The implementation may choose to
place padding in the trailer buffer, in which case it will set the
padding buffer length to 0.)
* **GSS_C_BUFFER_TYPE_STREAM**: For unwrapping only, a buffer
containing a complete wrap token in standard format to be unwrapped.
* **GSS_C_BUFFER_TYPE_SIGN_ONLY**: A buffer to be included in the
token's integrity protection checksum, but not to be encrypted or
included in the token itself.
For gss_wrap_iov, the IOV list should contain one HEADER buffer,
followed by zero or more SIGN_ONLY buffers, followed by one or more
DATA buffers, followed by a TRAILER buffer. The memory pointed to by
the buffers is not required to be contiguous or in any particular
order. If *conf_req_flag* is true, DATA buffers will be encrypted
in-place, while SIGN_ONLY buffers will not be modified.
The type of an output buffer may be combined with
**GSS_C_BUFFER_FLAG_ALLOCATE** to request that gss_wrap_iov allocate
the buffer contents. If gss_wrap_iov allocates a buffer, it sets the
**GSS_C_BUFFER_FLAG_ALLOCATED** flag on the buffer type.
gss_release_iov_buffer can be used to release all allocated buffers
within an iov list and unset their allocated flags. Here is an
example of how gss_wrap_iov can be used with allocation requested
(*ctx* is assumed to be a previously established gss_ctx_id_t)::
OM_uint32 major, minor;
gss_iov_buffer_desc iov[4];
char str[] = "message";
iov[0].type = GSS_IOV_BUFFER_TYPE_HEADER | GSS_IOV_BUFFER_FLAG_ALLOCATE;
iov[1].type = GSS_IOV_BUFFER_TYPE_DATA;
iov[1].buffer.value = str;
iov[1].buffer.length = strlen(str);
iov[2].type = GSS_IOV_BUFFER_TYPE_PADDING | GSS_IOV_BUFFER_FLAG_ALLOCATE;
iov[3].type = GSS_IOV_BUFFER_TYPE_TRAILER | GSS_IOV_BUFFER_FLAG_ALLOCATE;
major = gss_wrap_iov(&minor, ctx, 1, GSS_C_QOP_DEFAULT, NULL,
iov, 4);
if (GSS_ERROR(major))
handle_error(major, minor);
/* Transmit or otherwise use resulting buffers. */
(void)gss_release_iov_buffer(&minor, iov, 4);
If the caller does not choose to request buffer allocation by
gss_wrap_iov, it should first call gss_wrap_iov_length to query the
lengths of the HEADER, PADDING, and TRAILER buffers. DATA buffers
must be provided in the iov list so that padding length can be
computed correctly, but the output buffers need not be initialized.
Here is an example of using gss_wrap_iov_length and gss_wrap_iov::
OM_uint32 major, minor;
gss_iov_buffer_desc iov[4];
char str[1024] = "message", *ptr;
iov[0].type = GSS_IOV_BUFFER_TYPE_HEADER;
iov[1].type = GSS_IOV_BUFFER_TYPE_DATA;
iov[1].buffer.value = str;
iov[1].buffer.length = strlen(str);
iov[2].type = GSS_IOV_BUFFER_TYPE_PADDING;
iov[3].type = GSS_IOV_BUFFER_TYPE_TRAILER;
major = gss_wrap_iov_length(&minor, ctx, 1, GSS_C_QOP_DEFAULT,
NULL, iov, 4);
if (GSS_ERROR(major))
handle_error(major, minor);
if (strlen(str) + iov[0].buffer.length + iov[2].buffer.length +
iov[3].buffer.length > sizeof(str))
handle_out_of_space_error();
ptr = str + strlen(str);
iov[0].buffer.value = ptr;
ptr += iov[0].buffer.length;
iov[2].buffer.value = ptr;
ptr += iov[2].buffer.length;
iov[3].buffer.value = ptr;
major = gss_wrap_iov(&minor, ctx, 1, GSS_C_QOP_DEFAULT, NULL,
iov, 4);
if (GSS_ERROR(major))
handle_error(major, minor);
If the context was established using the **GSS_C_DCE_STYLE** flag
(described in :rfc:`4757`), wrap tokens compatible with Microsoft DCE
RPC can be constructed. In this case, the IOV list must include a
SIGN_ONLY buffer, a DATA buffer, a second SIGN_ONLY buffer, and a
HEADER buffer in that order (the order of the buffer contents remains
arbitrary). The application must pad the DATA buffer to a multiple of
16 bytes as no padding or trailer buffer is used.
gss_unwrap_iov may be called with an IOV list just like one which
would be provided to gss_wrap_iov. DATA buffers will be decrypted
in-place if they were encrypted, and SIGN_ONLY buffers will not be
modified.
Alternatively, gss_unwrap_iov may be called with a single STREAM
buffer, zero or more SIGN_ONLY buffers, and a single DATA buffer. The
STREAM buffer is interpreted as a complete wrap token. The STREAM
buffer will be modified in-place to decrypt its contents. The DATA
buffer will be initialized to point to the decrypted data within the
STREAM buffer, unless it has the **GSS_C_BUFFER_FLAG_ALLOCATE** flag
set, in which case it will be initialized with a copy of the decrypted
data. Here is an example (*token* and *token_len* are assumed to be a
pre-existing pointer and length for a modifiable region of data)::
OM_uint32 major, minor;
gss_iov_buffer_desc iov[2];
iov[0].type = GSS_IOV_BUFFER_TYPE_STREAM;
iov[0].buffer.value = token;
iov[0].buffer.length = token_len;
iov[1].type = GSS_IOV_BUFFER_TYPE_DATA;
major = gss_unwrap_iov(&minor, ctx, NULL, NULL, iov, 2);
if (GSS_ERROR(major))
handle_error(major, minor);
/* Decrypted data is in iov[1].buffer, pointing to a subregion of
* token. */
.. _gssapi_mic_token:
IOV MIC tokens
--------------
The following extensions (declared in ``<gssapi/gssapi_ext.h>``) can
be used in release 1.12 or later to construct and verify MIC tokens
using an IOV list::
OM_uint32 gss_get_mic_iov(OM_uint32 *minor_status,
gss_ctx_id_t context_handle,
gss_qop_t qop_req,
gss_iov_buffer_desc *iov,
int iov_count);
OM_uint32 gss_get_mic_iov_length(OM_uint32 *minor_status,
gss_ctx_id_t context_handle,
gss_qop_t qop_req,
gss_iov_buffer_desc *iov,
iov_count);
OM_uint32 gss_verify_mic_iov(OM_uint32 *minor_status,
gss_ctx_id_t context_handle,
gss_qop_t *qop_state,
gss_iov_buffer_desc *iov,
int iov_count);
The caller of gss_get_mic_iov provides an array of gss_iov_buffer_desc
structures, each containing a type and a gss_buffer_desc structure.
Valid types include:
* **GSS_C_BUFFER_TYPE_DATA** and **GSS_C_BUFFER_TYPE_SIGN_ONLY**: The
corresponding buffer for each of these types will be signed for the
MIC token, in the order provided.
* **GSS_C_BUFFER_TYPE_MIC_TOKEN**: The GSSAPI MIC token.
The type of the MIC_TOKEN buffer may be combined with
**GSS_C_BUFFER_FLAG_ALLOCATE** to request that gss_get_mic_iov
allocate the buffer contents. If gss_get_mic_iov allocates the
buffer, it sets the **GSS_C_BUFFER_FLAG_ALLOCATED** flag on the buffer
type. gss_release_iov_buffer can be used to release all allocated
buffers within an iov list and unset their allocated flags. Here is
an example of how gss_get_mic_iov can be used with allocation
requested (*ctx* is assumed to be a previously established
gss_ctx_id_t)::
OM_uint32 major, minor;
gss_iov_buffer_desc iov[3];
iov[0].type = GSS_IOV_BUFFER_TYPE_DATA;
iov[0].buffer.value = "sign1";
iov[0].buffer.length = 5;
iov[1].type = GSS_IOV_BUFFER_TYPE_SIGN_ONLY;
iov[1].buffer.value = "sign2";
iov[1].buffer.length = 5;
iov[2].type = GSS_IOV_BUFFER_TYPE_MIC_TOKEN | GSS_IOV_BUFFER_FLAG_ALLOCATE;
major = gss_get_mic_iov(&minor, ctx, GSS_C_QOP_DEFAULT, iov, 3);
if (GSS_ERROR(major))
handle_error(major, minor);
/* Transmit or otherwise use iov[2].buffer. */
(void)gss_release_iov_buffer(&minor, iov, 3);
If the caller does not choose to request buffer allocation by
gss_get_mic_iov, it should first call gss_get_mic_iov_length to query
the length of the MIC_TOKEN buffer. Here is an example of using
gss_get_mic_iov_length and gss_get_mic_iov::
OM_uint32 major, minor;
gss_iov_buffer_desc iov[2];
char data[1024];
iov[0].type = GSS_IOV_BUFFER_TYPE_MIC_TOKEN;
iov[1].type = GSS_IOV_BUFFER_TYPE_DATA;
iov[1].buffer.value = "message";
iov[1].buffer.length = 7;
major = gss_get_mic_iov_length(&minor, ctx, GSS_C_QOP_DEFAULT, iov, 2);
if (GSS_ERROR(major))
handle_error(major, minor);
if (iov[0].buffer.length > sizeof(data))
handle_out_of_space_error();
iov[0].buffer.value = data;
major = gss_get_mic_iov(&minor, ctx, GSS_C_QOP_DEFAULT, iov, 2);
if (GSS_ERROR(major))
handle_error(major, minor);
.. _gss_accept_sec_context: https://tools.ietf.org/html/rfc2744.html#section-5.1
.. _gss_acquire_cred: https://tools.ietf.org/html/rfc2744.html#section-5.2
.. _gss_export_name: https://tools.ietf.org/html/rfc2744.html#section-5.13
.. _gss_get_name_attribute: https://tools.ietf.org/html/6680.html#section-7.5
.. _gss_import_name: https://tools.ietf.org/html/rfc2744.html#section-5.16
.. _gss_init_sec_context: https://tools.ietf.org/html/rfc2744.html#section-5.19
.. _gss_inquire_name: https://tools.ietf.org/html/rfc6680.txt#section-7.4
.. _gss_inquire_cred: https://tools.ietf.org/html/rfc2744.html#section-5.21

View file

@ -0,0 +1,28 @@
Differences between Heimdal and MIT Kerberos API
================================================
.. tabularcolumns:: |l|l|
.. table::
======================================== =================================================
:c:func:`krb5_auth_con_getaddrs()` H5l: If either of the pointers to local_addr
and remote_addr is not NULL, it is freed
first and then reallocated before being
populated with the content of corresponding
address from authentication context.
:c:func:`krb5_auth_con_setaddrs()` H5l: If either address is NULL, the previous
address remains in place
:c:func:`krb5_auth_con_setports()` H5l: Not implemented as of version 1.3.3
:c:func:`krb5_auth_con_setrecvsubkey()` H5l: If either port is NULL, the previous
port remains in place
:c:func:`krb5_auth_con_setsendsubkey()` H5l: Not implemented as of version 1.3.3
:c:func:`krb5_cc_set_config()` MIT: Before version 1.10 it was assumed that
the last argument *data* is ALWAYS non-zero.
:c:func:`krb5_cccol_last_change_time()` MIT: not implemented
:c:func:`krb5_set_default_realm()` H5l: Caches the computed default realm context
field. If the second argument is NULL,
it tries to retrieve it from libdefaults or DNS.
MIT: Computes the default realm each time
if it wasn't explicitly set in the context
======================================== =================================================

View file

@ -0,0 +1,16 @@
For application developers
==========================
.. toctree::
:maxdepth: 1
gssapi.rst
y2038.rst
h5l_mit_apidiff.rst
init_creds.rst
princ_handle.rst
.. toctree::
:maxdepth: 1
refs/index.rst

View file

@ -0,0 +1,304 @@
Initial credentials
===================
Software that performs tasks such as logging users into a computer
when they type their Kerberos password needs to get initial
credentials (usually ticket granting tickets) from Kerberos. Such
software shares some behavior with the :ref:`kinit(1)` program.
Whenever a program grants access to a resource (such as a local login
session on a desktop computer) based on a user successfully getting
initial Kerberos credentials, it must verify those credentials against
a secure shared secret (e.g., a host keytab) to ensure that the user
credentials actually originate from a legitimate KDC. Failure to
perform this verification is a critical vulnerability, because a
malicious user can execute the "Zanarotti attack": the user constructs
a fake response that appears to come from the legitimate KDC, but
whose contents come from an attacker-controlled KDC.
Some applications read a Kerberos password over the network (ideally
over a secure channel), which they then verify against the KDC. While
this technique may be the only practical way to integrate Kerberos
into some existing legacy systems, its use is contrary to the original
design goals of Kerberos.
The function :c:func:`krb5_get_init_creds_password` will get initial
credentials for a client using a password. An application that needs
to verify the credentials can call :c:func:`krb5_verify_init_creds`.
Here is an example of code to obtain and verify TGT credentials, given
strings *princname* and *password* for the client principal name and
password::
krb5_error_code ret;
krb5_creds creds;
krb5_principal client_princ = NULL;
memset(&creds, 0, sizeof(creds));
ret = krb5_parse_name(context, princname, &client_princ);
if (ret)
goto cleanup;
ret = krb5_get_init_creds_password(context, &creds, client_princ,
password, NULL, NULL, 0, NULL, NULL);
if (ret)
goto cleanup;
ret = krb5_verify_init_creds(context, &creds, NULL, NULL, NULL, NULL);
cleanup:
krb5_free_principal(context, client_princ);
krb5_free_cred_contents(context, &creds);
return ret;
Options for get_init_creds
--------------------------
The function :c:func:`krb5_get_init_creds_password` takes an options
parameter (which can be a null pointer). Use the function
:c:func:`krb5_get_init_creds_opt_alloc` to allocate an options
structure, and :c:func:`krb5_get_init_creds_opt_free` to free it. For
example::
krb5_error_code ret;
krb5_get_init_creds_opt *opt = NULL;
krb5_creds creds;
memset(&creds, 0, sizeof(creds));
ret = krb5_get_init_creds_opt_alloc(context, &opt);
if (ret)
goto cleanup;
krb5_get_init_creds_opt_set_tkt_life(opt, 24 * 60 * 60);
ret = krb5_get_init_creds_password(context, &creds, client_princ,
password, NULL, NULL, 0, NULL, opt);
if (ret)
goto cleanup;
cleanup:
krb5_get_init_creds_opt_free(context, opt);
krb5_free_cred_contents(context, &creds);
return ret;
Getting anonymous credentials
-----------------------------
As of release 1.8, it is possible to obtain fully anonymous or
partially anonymous (realm-exposed) credentials, if the KDC supports
it. The MIT KDC supports issuing fully anonymous credentials as of
release 1.8 if configured appropriately (see :ref:`anonymous_pkinit`),
but does not support issuing realm-exposed anonymous credentials at
this time.
To obtain fully anonymous credentials, call
:c:func:`krb5_get_init_creds_opt_set_anonymous` on the options
structure to set the anonymous flag, and specify a client principal
with the KDC's realm and a single empty data component (the principal
obtained by parsing ``@``\ *realmname*). Authentication will take
place using anonymous PKINIT; if successful, the client principal of
the resulting tickets will be
``WELLKNOWN/ANONYMOUS@WELLKNOWN:ANONYMOUS``. Here is an example::
krb5_get_init_creds_opt_set_anonymous(opt, 1);
ret = krb5_build_principal(context, &client_princ, strlen(myrealm),
myrealm, "", (char *)NULL);
if (ret)
goto cleanup;
ret = krb5_get_init_creds_password(context, &creds, client_princ,
password, NULL, NULL, 0, NULL, opt);
if (ret)
goto cleanup;
To obtain realm-exposed anonymous credentials, set the anonymous flag
on the options structure as above, but specify a normal client
principal in order to prove membership in the realm. Authentication
will take place as it normally does; if successful, the client
principal of the resulting tickets will be ``WELLKNOWN/ANONYMOUS@``\
*realmname*.
User interaction
----------------
Authenticating a user usually requires the entry of secret
information, such as a password. A password can be supplied directly
to :c:func:`krb5_get_init_creds_password` via the *password*
parameter, or the application can supply prompter and/or responder
callbacks instead. If callbacks are used, the user can also be
queried for other secret information such as a PIN, informed of
impending password expiration, or prompted to change a password which
has expired.
Prompter callback
~~~~~~~~~~~~~~~~~
A prompter callback can be specified via the *prompter* and *data*
parameters to :c:func:`krb5_get_init_creds_password`. The prompter
will be invoked each time the krb5 library has a question to ask or
information to present. When the prompter callback is invoked, the
*banner* argument (if not null) is intended to be displayed to the
user, and the questions to be answered are specified in the *prompts*
array. Each prompt contains a text question in the *prompt* field, a
*hidden* bit to indicate whether the answer should be hidden from
display, and a storage area for the answer in the *reply* field. The
callback should fill in each question's ``reply->data`` with the
answer, up to a maximum number of ``reply->length`` bytes, and then
reset ``reply->length`` to the length of the answer.
A prompter callback can call :c:func:`krb5_get_prompt_types` to get an
array of type constants corresponding to the prompts, to get
programmatic information about the semantic meaning of the questions.
:c:func:`krb5_get_prompt_types` may return a null pointer if no prompt
type information is available.
Text-based applications can use a built-in text prompter
implementation by supplying :c:func:`krb5_prompter_posix` as the
*prompter* parameter and a null pointer as the *data* parameter. For
example::
ret = krb5_get_init_creds_password(context, &creds, client_princ,
NULL, krb5_prompter_posix, NULL, 0,
NULL, NULL);
Responder callback
~~~~~~~~~~~~~~~~~~
A responder callback can be specified through the init_creds options
using the :c:func:`krb5_get_init_creds_opt_set_responder` function.
Responder callbacks can present a more sophisticated user interface
for authentication secrets. The responder callback is usually invoked
only once per authentication, with a list of questions produced by all
of the allowed preauthentication mechanisms.
When the responder callback is invoked, the *rctx* argument can be
accessed to obtain the list of questions and to answer them. The
:c:func:`krb5_responder_list_questions` function retrieves an array of
question types. For each question type, the
:c:func:`krb5_responder_get_challenge` function retrieves additional
information about the question, if applicable, and the
:c:func:`krb5_responder_set_answer` function sets the answer.
Responder question types, challenges, and answers are UTF-8 strings.
The question type is a well-known string; the meaning of the challenge
and answer depend on the question type. If an application does not
understand a question type, it cannot interpret the challenge or
provide an answer. Failing to answer a question typically results in
the prompter callback being used as a fallback.
Password question
#################
The :c:macro:`KRB5_RESPONDER_QUESTION_PASSWORD` (or ``"password"``)
question type requests the user's password. This question does not
have a challenge, and the response is simply the password string.
One-time password question
##########################
The :c:macro:`KRB5_RESPONDER_QUESTION_OTP` (or ``"otp"``) question
type requests a choice among one-time password tokens and the PIN and
value for the chosen token. The challenge and answer are JSON-encoded
strings, but an application can use convenience functions to avoid
doing any JSON processing itself.
The :c:func:`krb5_responder_otp_get_challenge` function decodes the
challenge into a krb5_responder_otp_challenge structure. The
:c:func:`krb5_responder_otp_set_answer` function selects one of the
token information elements from the challenge and supplies the value
and pin for that token.
PKINIT password or PIN question
###############################
The :c:macro:`KRB5_RESPONDER_QUESTION_PKINIT` (or ``"pkinit"``) question
type requests PINs for hardware devices and/or passwords for encrypted
credentials which are stored on disk, potentially also supplying
information about the state of the hardware devices. The challenge and
answer are JSON-encoded strings, but an application can use convenience
functions to avoid doing any JSON processing itself.
The :c:func:`krb5_responder_pkinit_get_challenge` function decodes the
challenges into a krb5_responder_pkinit_challenge structure. The
:c:func:`krb5_responder_pkinit_set_answer` function can be used to
supply the PIN or password for a particular client credential, and can
be called multiple times.
Example
#######
Here is an example of using a responder callback::
static krb5_error_code
my_responder(krb5_context context, void *data,
krb5_responder_context rctx)
{
krb5_error_code ret;
krb5_responder_otp_challenge *chl;
if (krb5_responder_get_challenge(context, rctx,
KRB5_RESPONDER_QUESTION_PASSWORD)) {
ret = krb5_responder_set_answer(context, rctx,
KRB5_RESPONDER_QUESTION_PASSWORD,
"open sesame");
if (ret)
return ret;
}
ret = krb5_responder_otp_get_challenge(context, rctx, &chl);
if (ret == 0 && chl != NULL) {
ret = krb5_responder_otp_set_answer(context, rctx, 0, "1234",
NULL);
krb5_responder_otp_challenge_free(context, rctx, chl);
if (ret)
return ret;
}
return 0;
}
static krb5_error_code
get_creds(krb5_context context, krb5_principal client_princ)
{
krb5_error_code ret;
krb5_get_init_creds_opt *opt = NULL;
krb5_creds creds;
memset(&creds, 0, sizeof(creds));
ret = krb5_get_init_creds_opt_alloc(context, &opt);
if (ret)
goto cleanup;
ret = krb5_get_init_creds_opt_set_responder(context, opt, my_responder,
NULL);
if (ret)
goto cleanup;
ret = krb5_get_init_creds_password(context, &creds, client_princ,
NULL, NULL, NULL, 0, NULL, opt);
cleanup:
krb5_get_init_creds_opt_free(context, opt);
krb5_free_cred_contents(context, &creds);
return ret;
}
Verifying initial credentials
-----------------------------
Use the function :c:func:`krb5_verify_init_creds` to verify initial
credentials. It takes an options structure (which can be a null
pointer). Use :c:func:`krb5_verify_init_creds_opt_init` to initialize
the caller-allocated options structure, and
:c:func:`krb5_verify_init_creds_opt_set_ap_req_nofail` to set the
"nofail" option. For example::
krb5_verify_init_creds_opt vopt;
krb5_verify_init_creds_opt_init(&vopt);
krb5_verify_init_creds_opt_set_ap_req_nofail(&vopt, 1);
ret = krb5_verify_init_creds(context, &creds, NULL, NULL, NULL, &vopt);
The confusingly named "nofail" option, when set, means that the
verification must actually succeed in order for
:c:func:`krb5_verify_init_creds` to indicate success. The default
state of this option (cleared) means that if there is no key material
available to verify the user credentials, the verification will
succeed anyway. (The default can be changed by a configuration file
setting.)
This accommodates a use case where a large number of unkeyed shared
desktop workstations need to allow users to log in using Kerberos.
The security risks from this practice are mitigated by the absence of
valuable state on the shared workstations---any valuable resources
that the users would access reside on networked servers.

View file

@ -0,0 +1,79 @@
Principal manipulation and parsing
==================================
Kerberos principal structure
..
:c:type:`krb5_principal_data`
:c:type:`krb5_principal`
..
Create and free principal
..
:c:func:`krb5_build_principal()`
:c:func:`krb5_build_principal_alloc_va()`
:c:func:`krb5_build_principal_ext()`
:c:func:`krb5_copy_principal()`
:c:func:`krb5_free_principal()`
:c:func:`krb5_cc_get_principal()`
..
Comparing
..
:c:func:`krb5_principal_compare()`
:c:func:`krb5_principal_compare_flags()`
:c:func:`krb5_principal_compare_any_realm()`
:c:func:`krb5_sname_match()`
:c:func:`krb5_sname_to_principal()`
..
Parsing:
..
:c:func:`krb5_parse_name()`
:c:func:`krb5_parse_name_flags()`
:c:func:`krb5_unparse_name()`
:c:func:`krb5_unparse_name_flags()`
..
Utilities:
..
:c:func:`krb5_is_config_principal()`
:c:func:`krb5_kuserok()`
:c:func:`krb5_set_password()`
:c:func:`krb5_set_password_using_ccache()`
:c:func:`krb5_set_principal_realm()`
:c:func:`krb5_realm_compare()`
..

View file

@ -0,0 +1,413 @@
krb5 API
========
Frequently used public interfaces
----------------------------------
.. toctree::
:maxdepth: 1
krb5_build_principal.rst
krb5_build_principal_alloc_va.rst
krb5_build_principal_ext.rst
krb5_cc_close.rst
krb5_cc_default.rst
krb5_cc_default_name.rst
krb5_cc_destroy.rst
krb5_cc_dup.rst
krb5_cc_get_name.rst
krb5_cc_get_principal.rst
krb5_cc_get_type.rst
krb5_cc_initialize.rst
krb5_cc_new_unique.rst
krb5_cc_resolve.rst
krb5_change_password.rst
krb5_chpw_message.rst
krb5_expand_hostname.rst
krb5_free_context.rst
krb5_free_error_message.rst
krb5_free_principal.rst
krb5_fwd_tgt_creds.rst
krb5_get_default_realm.rst
krb5_get_error_message.rst
krb5_get_host_realm.rst
krb5_get_credentials.rst
krb5_get_fallback_host_realm.rst
krb5_get_init_creds_keytab.rst
krb5_get_init_creds_opt_alloc.rst
krb5_get_init_creds_opt_free.rst
krb5_get_init_creds_opt_get_fast_flags.rst
krb5_get_init_creds_opt_set_address_list.rst
krb5_get_init_creds_opt_set_anonymous.rst
krb5_get_init_creds_opt_set_canonicalize.rst
krb5_get_init_creds_opt_set_change_password_prompt.rst
krb5_get_init_creds_opt_set_etype_list.rst
krb5_get_init_creds_opt_set_expire_callback.rst
krb5_get_init_creds_opt_set_fast_ccache.rst
krb5_get_init_creds_opt_set_fast_ccache_name.rst
krb5_get_init_creds_opt_set_fast_flags.rst
krb5_get_init_creds_opt_set_forwardable.rst
krb5_get_init_creds_opt_set_in_ccache.rst
krb5_get_init_creds_opt_set_out_ccache.rst
krb5_get_init_creds_opt_set_pa.rst
krb5_get_init_creds_opt_set_pac_request.rst
krb5_get_init_creds_opt_set_preauth_list.rst
krb5_get_init_creds_opt_set_proxiable.rst
krb5_get_init_creds_opt_set_renew_life.rst
krb5_get_init_creds_opt_set_responder.rst
krb5_get_init_creds_opt_set_salt.rst
krb5_get_init_creds_opt_set_tkt_life.rst
krb5_get_init_creds_password.rst
krb5_get_profile.rst
krb5_get_prompt_types.rst
krb5_get_renewed_creds.rst
krb5_get_validated_creds.rst
krb5_init_context.rst
krb5_init_secure_context.rst
krb5_is_config_principal.rst
krb5_is_thread_safe.rst
krb5_kt_close.rst
krb5_kt_client_default.rst
krb5_kt_default.rst
krb5_kt_default_name.rst
krb5_kt_dup.rst
krb5_kt_get_name.rst
krb5_kt_get_type.rst
krb5_kt_resolve.rst
krb5_kuserok.rst
krb5_parse_name.rst
krb5_parse_name_flags.rst
krb5_principal_compare.rst
krb5_principal_compare_any_realm.rst
krb5_principal_compare_flags.rst
krb5_prompter_posix.rst
krb5_realm_compare.rst
krb5_responder_get_challenge.rst
krb5_responder_list_questions.rst
krb5_responder_set_answer.rst
krb5_responder_otp_get_challenge.rst
krb5_responder_otp_set_answer.rst
krb5_responder_otp_challenge_free.rst
krb5_responder_pkinit_get_challenge.rst
krb5_responder_pkinit_set_answer.rst
krb5_responder_pkinit_challenge_free.rst
krb5_set_default_realm.rst
krb5_set_password.rst
krb5_set_password_using_ccache.rst
krb5_set_principal_realm.rst
krb5_set_trace_callback.rst
krb5_set_trace_filename.rst
krb5_sname_match.rst
krb5_sname_to_principal.rst
krb5_unparse_name.rst
krb5_unparse_name_ext.rst
krb5_unparse_name_flags.rst
krb5_unparse_name_flags_ext.rst
krb5_us_timeofday.rst
krb5_verify_authdata_kdc_issued.rst
Rarely used public interfaces
--------------------------------
.. toctree::
:maxdepth: 1
krb5_425_conv_principal.rst
krb5_524_conv_principal.rst
krb5_address_compare.rst
krb5_address_order.rst
krb5_address_search.rst
krb5_allow_weak_crypto.rst
krb5_aname_to_localname.rst
krb5_anonymous_principal.rst
krb5_anonymous_realm.rst
krb5_appdefault_boolean.rst
krb5_appdefault_string.rst
krb5_auth_con_free.rst
krb5_auth_con_genaddrs.rst
krb5_auth_con_get_checksum_func.rst
krb5_auth_con_getaddrs.rst
krb5_auth_con_getauthenticator.rst
krb5_auth_con_getflags.rst
krb5_auth_con_getkey.rst
krb5_auth_con_getkey_k.rst
krb5_auth_con_getlocalseqnumber.rst
krb5_auth_con_getrcache.rst
krb5_auth_con_getrecvsubkey.rst
krb5_auth_con_getrecvsubkey_k.rst
krb5_auth_con_getremoteseqnumber.rst
krb5_auth_con_getsendsubkey.rst
krb5_auth_con_getsendsubkey_k.rst
krb5_auth_con_init.rst
krb5_auth_con_set_checksum_func.rst
krb5_auth_con_set_req_cksumtype.rst
krb5_auth_con_setaddrs.rst
krb5_auth_con_setflags.rst
krb5_auth_con_setports.rst
krb5_auth_con_setrcache.rst
krb5_auth_con_setrecvsubkey.rst
krb5_auth_con_setrecvsubkey_k.rst
krb5_auth_con_setsendsubkey.rst
krb5_auth_con_setsendsubkey_k.rst
krb5_auth_con_setuseruserkey.rst
krb5_cc_cache_match.rst
krb5_cc_copy_creds.rst
krb5_cc_end_seq_get.rst
krb5_cc_get_config.rst
krb5_cc_get_flags.rst
krb5_cc_get_full_name.rst
krb5_cc_move.rst
krb5_cc_next_cred.rst
krb5_cc_remove_cred.rst
krb5_cc_retrieve_cred.rst
krb5_cc_select.rst
krb5_cc_set_config.rst
krb5_cc_set_default_name.rst
krb5_cc_set_flags.rst
krb5_cc_start_seq_get.rst
krb5_cc_store_cred.rst
krb5_cc_support_switch.rst
krb5_cc_switch.rst
krb5_cccol_cursor_free.rst
krb5_cccol_cursor_new.rst
krb5_cccol_cursor_next.rst
krb5_cccol_have_content.rst
krb5_clear_error_message.rst
krb5_check_clockskew.rst
krb5_copy_addresses.rst
krb5_copy_authdata.rst
krb5_copy_authenticator.rst
krb5_copy_checksum.rst
krb5_copy_context.rst
krb5_copy_creds.rst
krb5_copy_data.rst
krb5_copy_error_message.rst
krb5_copy_keyblock.rst
krb5_copy_keyblock_contents.rst
krb5_copy_principal.rst
krb5_copy_ticket.rst
krb5_find_authdata.rst
krb5_free_addresses.rst
krb5_free_ap_rep_enc_part.rst
krb5_free_authdata.rst
krb5_free_authenticator.rst
krb5_free_cred_contents.rst
krb5_free_creds.rst
krb5_free_data.rst
krb5_free_data_contents.rst
krb5_free_default_realm.rst
krb5_free_enctypes.rst
krb5_free_error.rst
krb5_free_host_realm.rst
krb5_free_keyblock.rst
krb5_free_keyblock_contents.rst
krb5_free_keytab_entry_contents.rst
krb5_free_string.rst
krb5_free_ticket.rst
krb5_free_unparsed_name.rst
krb5_get_etype_info.rst
krb5_get_permitted_enctypes.rst
krb5_get_server_rcache.rst
krb5_get_time_offsets.rst
krb5_init_context_profile.rst
krb5_init_creds_free.rst
krb5_init_creds_get.rst
krb5_init_creds_get_creds.rst
krb5_init_creds_get_error.rst
krb5_init_creds_get_times.rst
krb5_init_creds_init.rst
krb5_init_creds_set_keytab.rst
krb5_init_creds_set_password.rst
krb5_init_creds_set_service.rst
krb5_init_creds_step.rst
krb5_init_keyblock.rst
krb5_is_referral_realm.rst
krb5_kdc_sign_ticket.rst
krb5_kdc_verify_ticket.rst
krb5_kt_add_entry.rst
krb5_kt_end_seq_get.rst
krb5_kt_get_entry.rst
krb5_kt_have_content.rst
krb5_kt_next_entry.rst
krb5_kt_read_service_key.rst
krb5_kt_remove_entry.rst
krb5_kt_start_seq_get.rst
krb5_make_authdata_kdc_issued.rst
krb5_marshal_credentials.rst
krb5_merge_authdata.rst
krb5_mk_1cred.rst
krb5_mk_error.rst
krb5_mk_ncred.rst
krb5_mk_priv.rst
krb5_mk_rep.rst
krb5_mk_rep_dce.rst
krb5_mk_req.rst
krb5_mk_req_extended.rst
krb5_mk_safe.rst
krb5_os_localaddr.rst
krb5_pac_add_buffer.rst
krb5_pac_free.rst
krb5_pac_get_buffer.rst
krb5_pac_get_types.rst
krb5_pac_init.rst
krb5_pac_parse.rst
krb5_pac_sign.rst
krb5_pac_sign_ext.rst
krb5_pac_verify.rst
krb5_pac_verify_ext.rst
krb5_pac_get_client_info.rst
krb5_prepend_error_message.rst
krb5_principal2salt.rst
krb5_rd_cred.rst
krb5_rd_error.rst
krb5_rd_priv.rst
krb5_rd_rep.rst
krb5_rd_rep_dce.rst
krb5_rd_req.rst
krb5_rd_safe.rst
krb5_read_password.rst
krb5_salttype_to_string.rst
krb5_server_decrypt_ticket_keytab.rst
krb5_set_default_tgs_enctypes.rst
krb5_set_error_message.rst
krb5_set_kdc_recv_hook.rst
krb5_set_kdc_send_hook.rst
krb5_set_real_time.rst
krb5_string_to_cksumtype.rst
krb5_string_to_deltat.rst
krb5_string_to_enctype.rst
krb5_string_to_salttype.rst
krb5_string_to_timestamp.rst
krb5_timeofday.rst
krb5_timestamp_to_sfstring.rst
krb5_timestamp_to_string.rst
krb5_tkt_creds_free.rst
krb5_tkt_creds_get.rst
krb5_tkt_creds_get_creds.rst
krb5_tkt_creds_get_times.rst
krb5_tkt_creds_init.rst
krb5_tkt_creds_step.rst
krb5_unmarshal_credentials.rst
krb5_verify_init_creds.rst
krb5_verify_init_creds_opt_init.rst
krb5_verify_init_creds_opt_set_ap_req_nofail.rst
krb5_vprepend_error_message.rst
krb5_vset_error_message.rst
krb5_vwrap_error_message.rst
krb5_wrap_error_message.rst
Public interfaces that should not be called directly
-------------------------------------------------------
.. toctree::
:maxdepth: 1
krb5_c_block_size.rst
krb5_c_checksum_length.rst
krb5_c_crypto_length.rst
krb5_c_crypto_length_iov.rst
krb5_c_decrypt.rst
krb5_c_decrypt_iov.rst
krb5_c_derive_prfplus.rst
krb5_c_encrypt.rst
krb5_c_encrypt_iov.rst
krb5_c_encrypt_length.rst
krb5_c_enctype_compare.rst
krb5_c_free_state.rst
krb5_c_fx_cf2_simple.rst
krb5_c_init_state.rst
krb5_c_is_coll_proof_cksum.rst
krb5_c_is_keyed_cksum.rst
krb5_c_keyed_checksum_types.rst
krb5_c_keylengths.rst
krb5_c_make_checksum.rst
krb5_c_make_checksum_iov.rst
krb5_c_make_random_key.rst
krb5_c_padding_length.rst
krb5_c_prf.rst
krb5_c_prfplus.rst
krb5_c_prf_length.rst
krb5_c_random_add_entropy.rst
krb5_c_random_make_octets.rst
krb5_c_random_os_entropy.rst
krb5_c_random_to_key.rst
krb5_c_string_to_key.rst
krb5_c_string_to_key_with_params.rst
krb5_c_valid_cksumtype.rst
krb5_c_valid_enctype.rst
krb5_c_verify_checksum.rst
krb5_c_verify_checksum_iov.rst
krb5_cksumtype_to_string.rst
krb5_decode_authdata_container.rst
krb5_decode_ticket.rst
krb5_deltat_to_string.rst
krb5_encode_authdata_container.rst
krb5_enctype_to_name.rst
krb5_enctype_to_string.rst
krb5_free_checksum.rst
krb5_free_checksum_contents.rst
krb5_free_cksumtypes.rst
krb5_free_tgt_creds.rst
krb5_k_create_key.rst
krb5_k_decrypt.rst
krb5_k_decrypt_iov.rst
krb5_k_encrypt.rst
krb5_k_encrypt_iov.rst
krb5_k_free_key.rst
krb5_k_key_enctype.rst
krb5_k_key_keyblock.rst
krb5_k_make_checksum.rst
krb5_k_make_checksum_iov.rst
krb5_k_prf.rst
krb5_k_reference_key.rst
krb5_k_verify_checksum.rst
krb5_k_verify_checksum_iov.rst
Legacy convenience interfaces
------------------------------
.. toctree::
:maxdepth: 1
krb5_recvauth.rst
krb5_recvauth_version.rst
krb5_sendauth.rst
Deprecated public interfaces
------------------------------
.. toctree::
:maxdepth: 1
krb5_524_convert_creds.rst
krb5_auth_con_getlocalsubkey.rst
krb5_auth_con_getremotesubkey.rst
krb5_auth_con_initivector.rst
krb5_build_principal_va.rst
krb5_c_random_seed.rst
krb5_calculate_checksum.rst
krb5_checksum_size.rst
krb5_encrypt.rst
krb5_decrypt.rst
krb5_eblock_enctype.rst
krb5_encrypt_size.rst
krb5_finish_key.rst
krb5_finish_random_key.rst
krb5_cc_gen_new.rst
krb5_get_credentials_renew.rst
krb5_get_credentials_validate.rst
krb5_get_in_tkt_with_password.rst
krb5_get_in_tkt_with_skey.rst
krb5_get_in_tkt_with_keytab.rst
krb5_get_init_creds_opt_init.rst
krb5_init_random_key.rst
krb5_kt_free_entry.rst
krb5_random_key.rst
krb5_process_key.rst
krb5_string_to_key.rst
krb5_use_enctype.rst
krb5_verify_checksum.rst

View file

@ -0,0 +1,9 @@
Complete reference - API and datatypes
======================================
.. toctree::
:maxdepth: 1
api/index.rst
types/index.rst
macros/index.rst

View file

@ -0,0 +1,397 @@
krb5 simple macros
=========================
Public
-------
.. toctree::
:maxdepth: 1
ADDRTYPE_ADDRPORT.rst
ADDRTYPE_CHAOS.rst
ADDRTYPE_DDP.rst
ADDRTYPE_INET.rst
ADDRTYPE_INET6.rst
ADDRTYPE_IPPORT.rst
ADDRTYPE_ISO.rst
ADDRTYPE_IS_LOCAL.rst
ADDRTYPE_NETBIOS.rst
ADDRTYPE_XNS.rst
AD_TYPE_EXTERNAL.rst
AD_TYPE_FIELD_TYPE_MASK.rst
AD_TYPE_REGISTERED.rst
AD_TYPE_RESERVED.rst
AP_OPTS_ETYPE_NEGOTIATION.rst
AP_OPTS_MUTUAL_REQUIRED.rst
AP_OPTS_RESERVED.rst
AP_OPTS_USE_SESSION_KEY.rst
AP_OPTS_USE_SUBKEY.rst
AP_OPTS_WIRE_MASK.rst
CKSUMTYPE_CMAC_CAMELLIA128.rst
CKSUMTYPE_CMAC_CAMELLIA256.rst
CKSUMTYPE_CRC32.rst
CKSUMTYPE_DESCBC.rst
CKSUMTYPE_HMAC_MD5_ARCFOUR.rst
CKSUMTYPE_HMAC_SHA1_96_AES128.rst
CKSUMTYPE_HMAC_SHA1_96_AES256.rst
CKSUMTYPE_HMAC_SHA256_128_AES128.rst
CKSUMTYPE_HMAC_SHA384_192_AES256.rst
CKSUMTYPE_HMAC_SHA1_DES3.rst
CKSUMTYPE_MD5_HMAC_ARCFOUR.rst
CKSUMTYPE_NIST_SHA.rst
CKSUMTYPE_RSA_MD4.rst
CKSUMTYPE_RSA_MD4_DES.rst
CKSUMTYPE_RSA_MD5.rst
CKSUMTYPE_RSA_MD5_DES.rst
CKSUMTYPE_SHA1.rst
ENCTYPE_AES128_CTS_HMAC_SHA1_96.rst
ENCTYPE_AES128_CTS_HMAC_SHA256_128.rst
ENCTYPE_AES256_CTS_HMAC_SHA1_96.rst
ENCTYPE_AES256_CTS_HMAC_SHA384_192.rst
ENCTYPE_ARCFOUR_HMAC.rst
ENCTYPE_ARCFOUR_HMAC_EXP.rst
ENCTYPE_CAMELLIA128_CTS_CMAC.rst
ENCTYPE_CAMELLIA256_CTS_CMAC.rst
ENCTYPE_DES3_CBC_ENV.rst
ENCTYPE_DES3_CBC_RAW.rst
ENCTYPE_DES3_CBC_SHA.rst
ENCTYPE_DES3_CBC_SHA1.rst
ENCTYPE_DES_CBC_CRC.rst
ENCTYPE_DES_CBC_MD4.rst
ENCTYPE_DES_CBC_MD5.rst
ENCTYPE_DES_CBC_RAW.rst
ENCTYPE_DES_HMAC_SHA1.rst
ENCTYPE_DSA_SHA1_CMS.rst
ENCTYPE_MD5_RSA_CMS.rst
ENCTYPE_NULL.rst
ENCTYPE_RC2_CBC_ENV.rst
ENCTYPE_RSA_ENV.rst
ENCTYPE_RSA_ES_OAEP_ENV.rst
ENCTYPE_SHA1_RSA_CMS.rst
ENCTYPE_UNKNOWN.rst
KDC_OPT_ALLOW_POSTDATE.rst
KDC_OPT_CANONICALIZE.rst
KDC_OPT_CNAME_IN_ADDL_TKT.rst
KDC_OPT_DISABLE_TRANSITED_CHECK.rst
KDC_OPT_ENC_TKT_IN_SKEY.rst
KDC_OPT_FORWARDABLE.rst
KDC_OPT_FORWARDED.rst
KDC_OPT_POSTDATED.rst
KDC_OPT_PROXIABLE.rst
KDC_OPT_PROXY.rst
KDC_OPT_RENEW.rst
KDC_OPT_RENEWABLE.rst
KDC_OPT_RENEWABLE_OK.rst
KDC_OPT_REQUEST_ANONYMOUS.rst
KDC_OPT_VALIDATE.rst
KDC_TKT_COMMON_MASK.rst
KRB5_ALTAUTH_ATT_CHALLENGE_RESPONSE.rst
KRB5_ANONYMOUS_PRINCSTR.rst
KRB5_ANONYMOUS_REALMSTR.rst
KRB5_AP_REP.rst
KRB5_AP_REQ.rst
KRB5_AS_REP.rst
KRB5_AS_REQ.rst
KRB5_AUTHDATA_AND_OR.rst
KRB5_AUTHDATA_AP_OPTIONS.rst
KRB5_AUTHDATA_AUTH_INDICATOR.rst
KRB5_AUTHDATA_CAMMAC.rst
KRB5_AUTHDATA_ETYPE_NEGOTIATION.rst
KRB5_AUTHDATA_FX_ARMOR.rst
KRB5_AUTHDATA_IF_RELEVANT.rst
KRB5_AUTHDATA_INITIAL_VERIFIED_CAS.rst
KRB5_AUTHDATA_KDC_ISSUED.rst
KRB5_AUTHDATA_MANDATORY_FOR_KDC.rst
KRB5_AUTHDATA_OSF_DCE.rst
KRB5_AUTHDATA_SESAME.rst
KRB5_AUTHDATA_SIGNTICKET.rst
KRB5_AUTHDATA_WIN2K_PAC.rst
KRB5_AUTH_CONTEXT_DO_SEQUENCE.rst
KRB5_AUTH_CONTEXT_DO_TIME.rst
KRB5_AUTH_CONTEXT_GENERATE_LOCAL_ADDR.rst
KRB5_AUTH_CONTEXT_GENERATE_LOCAL_FULL_ADDR.rst
KRB5_AUTH_CONTEXT_GENERATE_REMOTE_ADDR.rst
KRB5_AUTH_CONTEXT_GENERATE_REMOTE_FULL_ADDR.rst
KRB5_AUTH_CONTEXT_PERMIT_ALL.rst
KRB5_AUTH_CONTEXT_RET_SEQUENCE.rst
KRB5_AUTH_CONTEXT_RET_TIME.rst
KRB5_AUTH_CONTEXT_USE_SUBKEY.rst
KRB5_CRED.rst
KRB5_CRYPTO_TYPE_CHECKSUM.rst
KRB5_CRYPTO_TYPE_DATA.rst
KRB5_CRYPTO_TYPE_EMPTY.rst
KRB5_CRYPTO_TYPE_HEADER.rst
KRB5_CRYPTO_TYPE_PADDING.rst
KRB5_CRYPTO_TYPE_SIGN_ONLY.rst
KRB5_CRYPTO_TYPE_STREAM.rst
KRB5_CRYPTO_TYPE_TRAILER.rst
KRB5_CYBERSAFE_SECUREID.rst
KRB5_DOMAIN_X500_COMPRESS.rst
KRB5_ENCPADATA_REQ_ENC_PA_REP.rst
KRB5_ERROR.rst
KRB5_FAST_REQUIRED.rst
KRB5_GC_CACHED.rst
KRB5_GC_CANONICALIZE.rst
KRB5_GC_CONSTRAINED_DELEGATION.rst
KRB5_GC_FORWARDABLE.rst
KRB5_GC_NO_STORE.rst
KRB5_GC_NO_TRANSIT_CHECK.rst
KRB5_GC_USER_USER.rst
KRB5_GET_INIT_CREDS_OPT_ADDRESS_LIST.rst
KRB5_GET_INIT_CREDS_OPT_ANONYMOUS.rst
KRB5_GET_INIT_CREDS_OPT_CANONICALIZE.rst
KRB5_GET_INIT_CREDS_OPT_CHG_PWD_PRMPT.rst
KRB5_GET_INIT_CREDS_OPT_ETYPE_LIST.rst
KRB5_GET_INIT_CREDS_OPT_FORWARDABLE.rst
KRB5_GET_INIT_CREDS_OPT_PREAUTH_LIST.rst
KRB5_GET_INIT_CREDS_OPT_PROXIABLE.rst
KRB5_GET_INIT_CREDS_OPT_RENEW_LIFE.rst
KRB5_GET_INIT_CREDS_OPT_SALT.rst
KRB5_GET_INIT_CREDS_OPT_TKT_LIFE.rst
KRB5_INIT_CONTEXT_SECURE.rst
KRB5_INIT_CONTEXT_KDC.rst
KRB5_INIT_CREDS_STEP_FLAG_CONTINUE.rst
KRB5_INT16_MAX.rst
KRB5_INT16_MIN.rst
KRB5_INT32_MAX.rst
KRB5_INT32_MIN.rst
KRB5_KEYUSAGE_AD_ITE.rst
KRB5_KEYUSAGE_AD_KDCISSUED_CKSUM.rst
KRB5_KEYUSAGE_AD_MTE.rst
KRB5_KEYUSAGE_AD_SIGNEDPATH.rst
KRB5_KEYUSAGE_APP_DATA_CKSUM.rst
KRB5_KEYUSAGE_APP_DATA_ENCRYPT.rst
KRB5_KEYUSAGE_AP_REP_ENCPART.rst
KRB5_KEYUSAGE_AP_REQ_AUTH.rst
KRB5_KEYUSAGE_AP_REQ_AUTH_CKSUM.rst
KRB5_KEYUSAGE_AS_REP_ENCPART.rst
KRB5_KEYUSAGE_AS_REQ.rst
KRB5_KEYUSAGE_AS_REQ_PA_ENC_TS.rst
KRB5_KEYUSAGE_CAMMAC.rst
KRB5_KEYUSAGE_ENC_CHALLENGE_CLIENT.rst
KRB5_KEYUSAGE_ENC_CHALLENGE_KDC.rst
KRB5_KEYUSAGE_FAST_ENC.rst
KRB5_KEYUSAGE_FAST_FINISHED.rst
KRB5_KEYUSAGE_FAST_REP.rst
KRB5_KEYUSAGE_FAST_REQ_CHKSUM.rst
KRB5_KEYUSAGE_GSS_TOK_MIC.rst
KRB5_KEYUSAGE_GSS_TOK_WRAP_INTEG.rst
KRB5_KEYUSAGE_GSS_TOK_WRAP_PRIV.rst
KRB5_KEYUSAGE_IAKERB_FINISHED.rst
KRB5_KEYUSAGE_KDC_REP_TICKET.rst
KRB5_KEYUSAGE_KRB_CRED_ENCPART.rst
KRB5_KEYUSAGE_KRB_ERROR_CKSUM.rst
KRB5_KEYUSAGE_KRB_PRIV_ENCPART.rst
KRB5_KEYUSAGE_KRB_SAFE_CKSUM.rst
KRB5_KEYUSAGE_PA_AS_FRESHNESS.rst
KRB5_KEYUSAGE_PA_FX_COOKIE.rst
KRB5_KEYUSAGE_PA_OTP_REQUEST.rst
KRB5_KEYUSAGE_PA_PKINIT_KX.rst
KRB5_KEYUSAGE_PA_S4U_X509_USER_REPLY.rst
KRB5_KEYUSAGE_PA_S4U_X509_USER_REQUEST.rst
KRB5_KEYUSAGE_PA_SAM_CHALLENGE_CKSUM.rst
KRB5_KEYUSAGE_PA_SAM_CHALLENGE_TRACKID.rst
KRB5_KEYUSAGE_PA_SAM_RESPONSE.rst
KRB5_KEYUSAGE_SPAKE.rst
KRB5_KEYUSAGE_TGS_REP_ENCPART_SESSKEY.rst
KRB5_KEYUSAGE_TGS_REP_ENCPART_SUBKEY.rst
KRB5_KEYUSAGE_TGS_REQ_AD_SESSKEY.rst
KRB5_KEYUSAGE_TGS_REQ_AD_SUBKEY.rst
KRB5_KEYUSAGE_TGS_REQ_AUTH.rst
KRB5_KEYUSAGE_TGS_REQ_AUTH_CKSUM.rst
KRB5_KPASSWD_ACCESSDENIED.rst
KRB5_KPASSWD_AUTHERROR.rst
KRB5_KPASSWD_BAD_VERSION.rst
KRB5_KPASSWD_HARDERROR.rst
KRB5_KPASSWD_INITIAL_FLAG_NEEDED.rst
KRB5_KPASSWD_MALFORMED.rst
KRB5_KPASSWD_SOFTERROR.rst
KRB5_KPASSWD_SUCCESS.rst
KRB5_LRQ_ALL_ACCT_EXPTIME.rst
KRB5_LRQ_ALL_LAST_INITIAL.rst
KRB5_LRQ_ALL_LAST_RENEWAL.rst
KRB5_LRQ_ALL_LAST_REQ.rst
KRB5_LRQ_ALL_LAST_TGT.rst
KRB5_LRQ_ALL_LAST_TGT_ISSUED.rst
KRB5_LRQ_ALL_PW_EXPTIME.rst
KRB5_LRQ_NONE.rst
KRB5_LRQ_ONE_ACCT_EXPTIME.rst
KRB5_LRQ_ONE_LAST_INITIAL.rst
KRB5_LRQ_ONE_LAST_RENEWAL.rst
KRB5_LRQ_ONE_LAST_REQ.rst
KRB5_LRQ_ONE_LAST_TGT.rst
KRB5_LRQ_ONE_LAST_TGT_ISSUED.rst
KRB5_LRQ_ONE_PW_EXPTIME.rst
KRB5_NT_ENTERPRISE_PRINCIPAL.rst
KRB5_NT_ENT_PRINCIPAL_AND_ID.rst
KRB5_NT_MS_PRINCIPAL.rst
KRB5_NT_MS_PRINCIPAL_AND_ID.rst
KRB5_NT_PRINCIPAL.rst
KRB5_NT_SMTP_NAME.rst
KRB5_NT_SRV_HST.rst
KRB5_NT_SRV_INST.rst
KRB5_NT_SRV_XHST.rst
KRB5_NT_UID.rst
KRB5_NT_UNKNOWN.rst
KRB5_NT_WELLKNOWN.rst
KRB5_NT_X500_PRINCIPAL.rst
KRB5_PAC_ATTRIBUTES_INFO.rst
KRB5_PAC_CLIENT_INFO.rst
KRB5_PAC_CLIENT_CLAIMS.rst
KRB5_PAC_CREDENTIALS_INFO.rst
KRB5_PAC_DELEGATION_INFO.rst
KRB5_PAC_DEVICE_CLAIMS.rst
KRB5_PAC_DEVICE_INFO.rst
KRB5_PAC_LOGON_INFO.rst
KRB5_PAC_PRIVSVR_CHECKSUM.rst
KRB5_PAC_REQUESTOR.rst
KRB5_PAC_SERVER_CHECKSUM.rst
KRB5_PAC_TICKET_CHECKSUM.rst
KRB5_PAC_UPN_DNS_INFO.rst
KRB5_PAC_FULL_CHECKSUM.rst
KRB5_PADATA_AFS3_SALT.rst
KRB5_PADATA_AP_REQ.rst
KRB5_PADATA_AS_CHECKSUM.rst
KRB5_PADATA_AS_FRESHNESS.rst
KRB5_PADATA_ENCRYPTED_CHALLENGE.rst
KRB5_PADATA_ENC_SANDIA_SECURID.rst
KRB5_PADATA_ENC_TIMESTAMP.rst
KRB5_PADATA_ENC_UNIX_TIME.rst
KRB5_PADATA_ETYPE_INFO.rst
KRB5_PADATA_ETYPE_INFO2.rst
KRB5_PADATA_FOR_USER.rst
KRB5_PADATA_FX_COOKIE.rst
KRB5_PADATA_FX_ERROR.rst
KRB5_PADATA_FX_FAST.rst
KRB5_PADATA_GET_FROM_TYPED_DATA.rst
KRB5_PADATA_NONE.rst
KRB5_PADATA_OSF_DCE.rst
KRB5_PADATA_OTP_CHALLENGE.rst
KRB5_PADATA_OTP_PIN_CHANGE.rst
KRB5_PADATA_OTP_REQUEST.rst
KRB5_PADATA_PAC_OPTIONS.rst
KRB5_PADATA_PAC_REQUEST.rst
KRB5_PADATA_PKINIT_KX.rst
KRB5_PADATA_PK_AS_REP.rst
KRB5_PADATA_PK_AS_REP_OLD.rst
KRB5_PADATA_PK_AS_REQ.rst
KRB5_PADATA_PK_AS_REQ_OLD.rst
KRB5_PADATA_PW_SALT.rst
KRB5_PADATA_REFERRAL.rst
KRB5_PADATA_S4U_X509_USER.rst
KRB5_PADATA_SAM_CHALLENGE.rst
KRB5_PADATA_SAM_CHALLENGE_2.rst
KRB5_PADATA_SAM_REDIRECT.rst
KRB5_PADATA_SAM_RESPONSE.rst
KRB5_PADATA_SAM_RESPONSE_2.rst
KRB5_PADATA_SESAME.rst
KRB5_PADATA_SPAKE.rst
KRB5_PADATA_REDHAT_IDP_OAUTH2.rst
KRB5_PADATA_REDHAT_PASSKEY.rst
KRB5_PADATA_SVR_REFERRAL_INFO.rst
KRB5_PADATA_TGS_REQ.rst
KRB5_PADATA_USE_SPECIFIED_KVNO.rst
KRB5_PRINCIPAL_COMPARE_CASEFOLD.rst
KRB5_PRINCIPAL_COMPARE_ENTERPRISE.rst
KRB5_PRINCIPAL_COMPARE_IGNORE_REALM.rst
KRB5_PRINCIPAL_COMPARE_UTF8.rst
KRB5_PRINCIPAL_PARSE_ENTERPRISE.rst
KRB5_PRINCIPAL_PARSE_IGNORE_REALM.rst
KRB5_PRINCIPAL_PARSE_NO_DEF_REALM.rst
KRB5_PRINCIPAL_PARSE_NO_REALM.rst
KRB5_PRINCIPAL_PARSE_REQUIRE_REALM.rst
KRB5_PRINCIPAL_UNPARSE_DISPLAY.rst
KRB5_PRINCIPAL_UNPARSE_NO_REALM.rst
KRB5_PRINCIPAL_UNPARSE_SHORT.rst
KRB5_PRIV.rst
KRB5_PROMPT_TYPE_NEW_PASSWORD.rst
KRB5_PROMPT_TYPE_NEW_PASSWORD_AGAIN.rst
KRB5_PROMPT_TYPE_PASSWORD.rst
KRB5_PROMPT_TYPE_PREAUTH.rst
KRB5_PVNO.rst
KRB5_REALM_BRANCH_CHAR.rst
KRB5_RECVAUTH_BADAUTHVERS.rst
KRB5_RECVAUTH_SKIP_VERSION.rst
KRB5_REFERRAL_REALM.rst
KRB5_RESPONDER_PKINIT_FLAGS_TOKEN_USER_PIN_COUNT_LOW.rst
KRB5_RESPONDER_PKINIT_FLAGS_TOKEN_USER_PIN_FINAL_TRY.rst
KRB5_RESPONDER_PKINIT_FLAGS_TOKEN_USER_PIN_LOCKED.rst
KRB5_RESPONDER_QUESTION_PKINIT.rst
KRB5_RESPONDER_OTP_FLAGS_COLLECT_PIN.rst
KRB5_RESPONDER_OTP_FLAGS_COLLECT_TOKEN.rst
KRB5_RESPONDER_OTP_FLAGS_NEXTOTP.rst
KRB5_RESPONDER_OTP_FLAGS_SEPARATE_PIN.rst
KRB5_RESPONDER_OTP_FORMAT_ALPHANUMERIC.rst
KRB5_RESPONDER_OTP_FORMAT_DECIMAL.rst
KRB5_RESPONDER_OTP_FORMAT_HEXADECIMAL.rst
KRB5_RESPONDER_QUESTION_OTP.rst
KRB5_RESPONDER_QUESTION_PASSWORD.rst
KRB5_SAFE.rst
KRB5_SAM_MUST_PK_ENCRYPT_SAD.rst
KRB5_SAM_SEND_ENCRYPTED_SAD.rst
KRB5_SAM_USE_SAD_AS_KEY.rst
KRB5_TC_MATCH_2ND_TKT.rst
KRB5_TC_MATCH_AUTHDATA.rst
KRB5_TC_MATCH_FLAGS.rst
KRB5_TC_MATCH_FLAGS_EXACT.rst
KRB5_TC_MATCH_IS_SKEY.rst
KRB5_TC_MATCH_KTYPE.rst
KRB5_TC_MATCH_SRV_NAMEONLY.rst
KRB5_TC_MATCH_TIMES.rst
KRB5_TC_MATCH_TIMES_EXACT.rst
KRB5_TC_NOTICKET.rst
KRB5_TC_OPENCLOSE.rst
KRB5_TC_SUPPORTED_KTYPES.rst
KRB5_TGS_NAME.rst
KRB5_TGS_NAME_SIZE.rst
KRB5_TGS_REP.rst
KRB5_TGS_REQ.rst
KRB5_TKT_CREDS_STEP_FLAG_CONTINUE.rst
KRB5_VERIFY_INIT_CREDS_OPT_AP_REQ_NOFAIL.rst
KRB5_WELLKNOWN_NAMESTR.rst
LR_TYPE_INTERPRETATION_MASK.rst
LR_TYPE_THIS_SERVER_ONLY.rst
MAX_KEYTAB_NAME_LEN.rst
MSEC_DIRBIT.rst
MSEC_VAL_MASK.rst
SALT_TYPE_AFS_LENGTH.rst
SALT_TYPE_NO_LENGTH.rst
THREEPARAMOPEN.rst
TKT_FLG_ANONYMOUS.rst
TKT_FLG_ENC_PA_REP.rst
TKT_FLG_FORWARDABLE.rst
TKT_FLG_FORWARDED.rst
TKT_FLG_HW_AUTH.rst
TKT_FLG_INITIAL.rst
TKT_FLG_INVALID.rst
TKT_FLG_MAY_POSTDATE.rst
TKT_FLG_OK_AS_DELEGATE.rst
TKT_FLG_POSTDATED.rst
TKT_FLG_PRE_AUTH.rst
TKT_FLG_PROXIABLE.rst
TKT_FLG_PROXY.rst
TKT_FLG_RENEWABLE.rst
TKT_FLG_TRANSIT_POLICY_CHECKED.rst
VALID_INT_BITS.rst
VALID_UINT_BITS.rst
krb5_const.rst
krb5_princ_component.rst
krb5_princ_name.rst
krb5_princ_realm.rst
krb5_princ_set_realm.rst
krb5_princ_set_realm_data.rst
krb5_princ_set_realm_length.rst
krb5_princ_size.rst
krb5_princ_type.rst
krb5_roundup.rst
krb5_x.rst
krb5_xc.rst
Deprecated macros
------------------------------
.. toctree::
:maxdepth: 1
krb524_convert_creds_kdc.rst
krb524_init_ets.rst

View file

@ -0,0 +1,108 @@
krb5 types and structures
=========================
Public
-------
.. toctree::
:maxdepth: 1
krb5_address.rst
krb5_addrtype.rst
krb5_ap_req.rst
krb5_ap_rep.rst
krb5_ap_rep_enc_part.rst
krb5_authdata.rst
krb5_authdatatype.rst
krb5_authenticator.rst
krb5_boolean.rst
krb5_checksum.rst
krb5_const_pointer.rst
krb5_const_principal.rst
krb5_cred.rst
krb5_cred_enc_part.rst
krb5_cred_info.rst
krb5_creds.rst
krb5_crypto_iov.rst
krb5_cryptotype.rst
krb5_data.rst
krb5_deltat.rst
krb5_enc_data.rst
krb5_enc_kdc_rep_part.rst
krb5_enc_tkt_part.rst
krb5_encrypt_block.rst
krb5_enctype.rst
krb5_error.rst
krb5_error_code.rst
krb5_expire_callback_func.rst
krb5_flags.rst
krb5_get_init_creds_opt.rst
krb5_gic_opt_pa_data.rst
krb5_int16.rst
krb5_int32.rst
krb5_kdc_rep.rst
krb5_kdc_req.rst
krb5_keyblock.rst
krb5_keytab_entry.rst
krb5_keyusage.rst
krb5_kt_cursor.rst
krb5_kvno.rst
krb5_last_req_entry.rst
krb5_magic.rst
krb5_mk_req_checksum_func.rst
krb5_msgtype.rst
krb5_octet.rst
krb5_pa_pac_req.rst
krb5_pa_server_referral_data.rst
krb5_pa_svr_referral_data.rst
krb5_pa_data.rst
krb5_pointer.rst
krb5_post_recv_fn.rst
krb5_pre_send_fn.rst
krb5_preauthtype.rst
krb5_principal.rst
krb5_principal_data.rst
krb5_prompt.rst
krb5_prompt_type.rst
krb5_prompter_fct.rst
krb5_pwd_data.rst
krb5_responder_context.rst
krb5_responder_fn.rst
krb5_responder_otp_challenge.rst
krb5_responder_otp_tokeninfo.rst
krb5_responder_pkinit_challenge.rst
krb5_responder_pkinit_identity.rst
krb5_response.rst
krb5_replay_data.rst
krb5_ticket.rst
krb5_ticket_times.rst
krb5_timestamp.rst
krb5_tkt_authent.rst
krb5_trace_callback.rst
krb5_trace_info.rst
krb5_transited.rst
krb5_typed_data.rst
krb5_ui_2.rst
krb5_ui_4.rst
krb5_verify_init_creds_opt.rst
passwd_phrase_element.rst
Internal
---------
.. toctree::
:maxdepth: 1
krb5_auth_context.rst
krb5_cksumtype
krb5_context.rst
krb5_cc_cursor.rst
krb5_ccache.rst
krb5_cccol_cursor.rst
krb5_init_creds_context.rst
krb5_key.rst
krb5_keytab.rst
krb5_pac.rst
krb5_rcache.rst
krb5_tkt_creds_context.rst

View file

@ -0,0 +1,12 @@
.. highlight:: c
.. _krb5-int32-struct:
krb5_int32
==========
..
.. c:type:: krb5_int32
..
krb5_int32 is a signed 32-bit integer type

View file

@ -0,0 +1,12 @@
.. highlight:: c
.. _krb5-ui4-struct:
krb5_ui_4
==========
..
.. c:type:: krb5_ui_4
..
krb5_ui_4 is an unsigned 32-bit integer type.

View file

@ -0,0 +1,28 @@
Year 2038 considerations for uses of krb5_timestamp
===================================================
POSIX time values, which measure the number of seconds since January 1
1970, will exceed the maximum value representable in a signed 32-bit
integer in January 2038. This documentation describes considerations
for consumers of the MIT krb5 libraries.
Applications or libraries which use libkrb5 and consume the timestamps
included in credentials or other structures make use of the
:c:type:`krb5_timestamp` type. For historical reasons, krb5_timestamp
is a signed 32-bit integer, even on platforms where a larger type is
natively used to represent time values. To behave properly for time
values after January 2038, calling code should cast krb5_timestamp
values to uint32_t, and then to time_t::
(time_t)(uint32_t)timestamp
Used in this way, krb5_timestamp values can represent time values up
until February 2106, provided that the platform uses a 64-bit or
larger time_t type. This usage will also remain safe if a later
version of MIT krb5 changes krb5_timestamp to an unsigned 32-bit
integer.
The GSSAPI only uses representations of time intervals, not absolute
times. Callers of the GSSAPI should require no changes to behave
correctly after January 2038, provided that they use MIT krb5 release
1.16 or later.

View file

@ -0,0 +1,160 @@
.. _ccache_definition:
Credential cache
================
A credential cache (or "ccache") holds Kerberos credentials while they
remain valid and, generally, while the user's session lasts, so that
authenticating to a service multiple times (e.g., connecting to a web
or mail server more than once) doesn't require contacting the KDC
every time.
A credential cache usually contains one initial ticket which is
obtained using a password or another form of identity verification.
If this ticket is a ticket-granting ticket, it can be used to obtain
additional credentials without the password. Because the credential
cache does not store the password, less long-term damage can be done
to the user's account if the machine is compromised.
A credentials cache stores a default client principal name, set when
the cache is created. This is the name shown at the top of the
:ref:`klist(1)` *-A* output.
Each normal cache entry includes a service principal name, a client
principal name (which, in some ccache types, need not be the same as
the default), lifetime information, and flags, along with the
credential itself. There are also other entries, indicated by special
names, that store additional information.
ccache types
------------
The credential cache interface, like the :ref:`keytab_definition` and
:ref:`rcache_definition` interfaces, uses `TYPE:value` strings to
indicate the type of credential cache and any associated cache naming
data to use.
There are several kinds of credentials cache supported in the MIT
Kerberos library. Not all are supported on every platform. In most
cases, it should be correct to use the default type built into the
library.
#. **API** is only implemented on Windows. It communicates with a
server process that holds the credentials in memory for the user,
rather than writing them to disk.
#. **DIR** points to the storage location of the collection of the
credential caches in *FILE:* format. It is most useful when dealing
with multiple Kerberos realms and KDCs. For release 1.10 the
directory must already exist. In post-1.10 releases the
requirement is for parent directory to exist and the current
process must have permissions to create the directory if it does
not exist. See :ref:`col_ccache` for details. New in release 1.10.
The following residual forms are supported:
* DIR:dirname
* DIR::dirpath/filename - a single cache within the directory
Switching to a ccache of the latter type causes it to become the
primary for the directory.
#. **FILE** caches are the simplest and most portable. A simple flat
file format is used to store one credential after another. This is
the default ccache type if no type is specified in a ccache name.
#. **KCM** caches work by contacting a daemon process called ``kcm``
to perform cache operations. If the cache name is just ``KCM:``,
the default cache as determined by the KCM daemon will be used.
Newly created caches must generally be named ``KCM:uid:name``,
where *uid* is the effective user ID of the running process.
KCM client support is new in release 1.13. A KCM daemon has not
yet been implemented in MIT krb5, but the client will interoperate
with the KCM daemon implemented by Heimdal. macOS 10.7 and higher
provides a KCM daemon as part of the operating system, and the
**KCM** cache type is used as the default cache on that platform in
a default build.
#. **KEYRING** is Linux-specific, and uses the kernel keyring support
to store credential data in unswappable kernel memory where only
the current user should be able to access it. The following
residual forms are supported:
* KEYRING:name
* KEYRING:process:name - process keyring
* KEYRING:thread:name - thread keyring
Starting with release 1.12 the *KEYRING* type supports collections.
The following new residual forms were added:
* KEYRING:session:name - session keyring
* KEYRING:user:name - user keyring
* KEYRING:persistent:uidnumber - persistent per-UID collection.
Unlike the user keyring, this collection survives after the user
logs out, until the cache credentials expire. This type of
ccache requires support from the kernel; otherwise, it will fall
back to the user keyring.
See :ref:`col_ccache` for details.
#. **MEMORY** caches are for storage of credentials that don't need to
be made available outside of the current process. For example, a
memory ccache is used by :ref:`kadmin(1)` to store the
administrative ticket used to contact the admin server. Memory
ccaches are faster than file ccaches and are automatically
destroyed when the process exits.
#. **MSLSA** is a Windows-specific cache type that accesses the
Windows credential store.
.. _col_ccache:
Collections of caches
---------------------
Some credential cache types can support collections of multiple
caches. One of the caches in the collection is designated as the
*primary* and will be used when the collection is resolved as a cache.
When a collection-enabled cache type is the default cache for a
process, applications can search the specified collection for a
specific client principal, and GSSAPI applications will automatically
select between the caches in the collection based on criteria such as
the target service realm.
Credential cache collections are new in release 1.10, with support
from the **DIR** and **API** ccache types. Starting in release 1.12,
collections are also supported by the **KEYRING** ccache type.
Collections are supported by the **KCM** ccache type in release 1.13.
Tool alterations to use cache collection
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
* :ref:`kdestroy(1)` *-A* will destroy all caches in the collection.
* If the default cache type supports switching, :ref:`kinit(1)`
*princname* will search the collection for a matching cache and
store credentials there, or will store credentials in a new unique
cache of the default type if no existing cache for the principal
exists. Either way, kinit will switch to the selected cache.
* :ref:`klist(1)` *-l* will list the caches in the collection.
* :ref:`klist(1)` *-A* will show the content of all caches in the
collection.
* :ref:`kswitch(1)` *-p princname* will search the collection for a
matching cache and switch to it.
* :ref:`kswitch(1)` *-c cachename* will switch to a specified cache.
Default ccache name
-------------------
The default credential cache name is determined by the following, in
descending order of priority:
#. The **KRB5CCNAME** environment variable. For example,
``KRB5CCNAME=DIR:/mydir/``.
#. The **default_ccache_name** profile variable in :ref:`libdefaults`.
#. The hardcoded default, |ccache|.

View file

@ -0,0 +1,140 @@
.. _datetime:
Supported date and time formats
===============================
.. _duration:
Time duration
-------------
This format is used to express a time duration in the Kerberos
configuration files and user commands. The allowed formats are:
====================== ============== ============
Format Example Value
---------------------- -------------- ------------
h:m[:s] 36:00 36 hours
NdNhNmNs 8h30s 8 hours 30 seconds
N (number of seconds) 3600 1 hour
====================== ============== ============
Here *N* denotes a number, *d* - days, *h* - hours, *m* - minutes,
*s* - seconds.
.. note::
The time interval should not exceed 2147483647 seconds.
Examples::
Request a ticket valid for one hour, five hours, 30 minutes
and 10 days respectively:
kinit -l 3600
kinit -l 5:00
kinit -l 30m
kinit -l "10d 0h 0m 0s"
.. _getdate:
getdate time
------------
Some of the kadmin and kdb5_util commands take a date-time in a
human-readable format. Some of the acceptable date-time
strings are:
+-----------+------------------+-----------------+
| | Format | Example |
+===========+==================+=================+
| Date | mm/dd/yy | 07/27/12 |
| +------------------+-----------------+
| | month dd, yyyy | Jul 27, 2012 |
| +------------------+-----------------+
| | yyyy-mm-dd | 2012-07-27 |
+-----------+------------------+-----------------+
| Absolute | HH:mm[:ss]pp | 08:30 PM |
| time +------------------+-----------------+
| | hh:mm[:ss] | 20:30 |
+-----------+------------------+-----------------+
| Relative | N tt | 30 sec |
| time | | |
+-----------+------------------+-----------------+
| Time zone | Z | EST |
| +------------------+-----------------+
| | z | -0400 |
+-----------+------------------+-----------------+
(See :ref:`abbreviation`.)
Examples::
Create a principal that expires on the date indicated:
addprinc test1 -expire "3/27/12 10:00:07 EST"
addprinc test2 -expire "January 23, 2015 10:05pm"
addprinc test3 -expire "22:00 GMT"
Add a principal that will expire in 30 minutes:
addprinc test4 -expire "30 minutes"
.. _abstime:
Absolute time
-------------
This rarely used date-time format can be noted in one of the
following ways:
+------------------------+----------------------+--------------+
| Format | Example | Value |
+========================+======================+==============+
| yyyymmddhhmmss | 20141231235900 | One minute |
+------------------------+----------------------+ before 2015 |
| yyyy.mm.dd.hh.mm.ss | 2014.12.31.23.59.00 | |
+------------------------+----------------------+ |
| yymmddhhmmss | 141231235900 | |
+------------------------+----------------------+ |
| yy.mm.dd.hh.mm.ss | 14.12.31.23.59.00 | |
+------------------------+----------------------+ |
| dd-month-yyyy:hh:mm:ss | 31-Dec-2014:23:59:00 | |
+------------------------+----------------------+--------------+
| hh:mm:ss | 20:00:00 | 8 o'clock in |
+------------------------+----------------------+ the evening |
| hhmmss | 200000 | |
+------------------------+----------------------+--------------+
(See :ref:`abbreviation`.)
Example::
Set the default expiration date to July 27, 2012 at 20:30
default_principal_expiration = 20120727203000
.. _abbreviation:
Abbreviations used in this document
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
| *month* : locales month name or its abbreviation;
| *dd* : day of month (01-31);
| *HH* : hours (00-12);
| *hh* : hours (00-23);
| *mm* : in time - minutes (00-59); in date - month (01-12);
| *N* : number;
| *pp* : AM or PM;
| *ss* : seconds (00-60);
| *tt* : time units (hours, minutes, min, seconds, sec);
| *yyyy* : year;
| *yy* : last two digits of the year;
| *Z* : alphabetic time zone abbreviation;
| *z* : numeric time zone;
.. note::
- If the date specification contains spaces, you may need to
enclose it in double quotes;
- All keywords are case-insensitive.

View file

@ -0,0 +1,14 @@
.. _basic_concepts:
Kerberos V5 concepts
====================
.. toctree::
:maxdepth: 1
ccache_def
keytab_def
rcache_def
stash_file_def
date_format

View file

@ -0,0 +1,59 @@
.. _keytab_definition:
keytab
======
A keytab (short for "key table") stores long-term keys for one or more
principals. Keytabs are normally represented by files in a standard
format, although in rare cases they can be represented in other ways.
Keytabs are used most often to allow server applications to accept
authentications from clients, but can also be used to obtain initial
credentials for client applications.
Keytabs are named using the format *type*\ ``:``\ *value*. Usually
*type* is ``FILE`` and *value* is the absolute pathname of the file.
The other possible value for *type* is ``MEMORY``, which indicates a
temporary keytab stored in the memory of the current process.
A keytab contains one or more entries, where each entry consists of a
timestamp (indicating when the entry was written to the keytab), a
principal name, a key version number, an encryption type, and the
encryption key itself.
A keytab can be displayed using the :ref:`klist(1)` command with the
``-k`` option. Keytabs can be created or appended to by extracting
keys from the KDC database using the :ref:`kadmin(1)` :ref:`ktadd`
command. Keytabs can be manipulated using the :ref:`ktutil(1)` and
:ref:`k5srvutil(1)` commands.
Default keytab
--------------
The default keytab is used by server applications if the application
does not request a specific keytab. The name of the default keytab is
determined by the following, in decreasing order of preference:
#. The **KRB5_KTNAME** environment variable.
#. The **default_keytab_name** profile variable in :ref:`libdefaults`.
#. The hardcoded default, |keytab|.
Default client keytab
---------------------
The default client keytab is used, if it is present and readable, to
automatically obtain initial credentials for GSSAPI client
applications. The principal name of the first entry in the client
keytab is used by default when obtaining initial credentials. The
name of the default client keytab is determined by the following, in
decreasing order of preference:
#. The **KRB5_CLIENT_KTNAME** environment variable.
#. The **default_client_keytab_name** profile variable in
:ref:`libdefaults`.
#. The hardcoded default, |ckeytab|.

View file

@ -0,0 +1,111 @@
.. _rcache_definition:
replay cache
============
A replay cache (or "rcache") keeps track of all authenticators
recently presented to a service. If a duplicate authentication
request is detected in the replay cache, an error message is sent to
the application program.
The replay cache interface, like the credential cache and
:ref:`keytab_definition` interfaces, uses `type:residual` strings to
indicate the type of replay cache and any associated cache naming
data to use.
Background information
----------------------
Some Kerberos or GSSAPI services use a simple authentication mechanism
where a message is sent containing an authenticator, which establishes
the encryption key that the client will use for talking to the
service. But nothing about that prevents an eavesdropper from
recording the messages sent by the client, establishing a new
connection, and re-sending or "replaying" the same messages; the
replayed authenticator will establish the same encryption key for the
new session, and the following messages will be decrypted and
processed. The attacker may not know what the messages say, and can't
generate new messages under the same encryption key, but in some
instances it may be harmful to the user (or helpful to the attacker)
to cause the server to see the same messages again a second time. For
example, if the legitimate client sends "delete first message in
mailbox", a replay from an attacker may delete another, different
"first" message. (Protocol design to guard against such problems has
been discussed in :rfc:`4120#section-10`.)
Even if one protocol uses further protection to verify that the client
side of the connection actually knows the encryption keys (and thus is
presumably a legitimate user), if another service uses the same
service principal name, it may be possible to record an authenticator
used with the first protocol and "replay" it against the second.
The replay cache mitigates these attacks somewhat, by keeping track of
authenticators that have been seen until their five-minute window
expires. Different authenticators generated by multiple connections
from the same legitimate client will generally have different
timestamps, and thus will not be considered the same.
This mechanism isn't perfect. If a message is sent to one application
server but a man-in-the-middle attacker can prevent it from actually
arriving at that server, the attacker could then use the authenticator
(once!) against a different service on the same host. This could be a
problem if the message from the client included something more than
authentication in the first message that could be useful to the
attacker (which is uncommon; in most protocols the server has to
indicate a successful authentication before the client sends
additional messages), or if the simple act of presenting the
authenticator triggers some interesting action in the service being
attacked.
Replay cache types
------------------
Unlike the credential cache and keytab interfaces, replay cache types
are in lowercase. The following types are defined:
#. **none** disables the replay cache. The residual value is ignored.
#. **file2** (new in release 1.18) uses a hash-based format to store
replay records. The file may grow to accommodate hash collisions.
The residual value is the filename.
#. **dfl** is the default type if no environment variable or
configuration specifies a different type. It stores replay data in
a file2 replay cache with a filename based on the effective uid.
The residual value is ignored.
For the dfl type, the location of the replay cache file is determined
as follows:
#. The directory is taken from the **KRB5RCACHEDIR** environment
variable, or the **TMPDIR** environment variable, or a temporary
directory determined at configuration time such as ``/var/tmp``, in
descending order of preference.
#. The filename is ``krb5_EUID.rcache2`` where EUID is the effective
uid of the process.
#. The file is opened without following symbolic links, and ownership
of the file is verified to match the effective uid.
On Windows, the directory for the dfl type is the local appdata
directory, unless overridden by the **KRB5RCACHEDIR** environment
variable. The filename on Windows is ``krb5.rcache2``, and the file
is opened normally.
Default replay cache name
-------------------------
The default replay cache name is determined by the following, in
descending order of priority:
#. The **KRB5RCACHENAME** environment variable (new in release 1.18).
#. The **KRB5RCACHETYPE** environment variable. If this variable is
set, the residual value is empty.
#. The **default_rcache_name** profile variable in :ref:`libdefaults`
(new in release 1.18).
#. If none of the above are set, the default replay cache name is
``dfl:``.

View file

@ -0,0 +1,25 @@
.. _stash_definition:
stash file
============
The stash file is a local copy of the master key that resides in
encrypted form on the KDC's local disk. The stash file is used to
authenticate the KDC to itself automatically before starting the
:ref:`kadmind(8)` and :ref:`krb5kdc(8)` daemons (e.g., as part of the
machine's boot sequence). The stash file, like the keytab file (see
:ref:`keytab_file`) is a potential point-of-entry for a break-in, and
if compromised, would allow unrestricted access to the Kerberos
database. If you choose to install a stash file, it should be
readable only by root, and should exist only on the KDC's local disk.
The file should not be part of any backup of the machine, unless
access to the backup data is secured as tightly as access to the
master password itself.
.. note::
If you choose not to install a stash file, the KDC will prompt you for the master key each time it starts up.
This means that the KDC will not be able to start automatically, such as after a system reboot.

75
crypto/krb5/doc/build/directory_org.rst vendored Normal file
View file

@ -0,0 +1,75 @@
Organization of the source directory
====================================
Below is a brief overview of the organization of the complete source
directory. More detailed descriptions follow.
=============== ==============================================
appl Kerberos application client and server programs
ccapi Credential cache services
clients Kerberos V5 user programs (See :ref:`user_commands`)
config Configure scripts
config-files Sample Kerberos configuration files
include include files needed to build the Kerberos system
kadmin Administrative interface to the Kerberos database: :ref:`kadmin(1)`, :ref:`kdb5_util(8)`, :ref:`ktutil(1)`.
kdc Kerberos V5 Authentication Service and Key Distribution Center
lib_ Libraries for use with/by Kerberos V5
plugins Kerberos plugins directory
po Localization infrastructure
prototype Templates files containing the MIT copyright message and a placeholder for the title and description of the file.
kprop Utilities for propagating the database to replica KDCs :ref:`kprop(8)` and :ref:`kpropd(8)`
tests Test suite
util_ Various utilities for building/configuring the code, sending bug reports, etc.
windows Source code for building Kerberos V5 on Windows (see windows/README)
=============== ==============================================
.. _lib:
lib
---
The lib directory contain several subdirectories as well as some
definition and glue files.
- The apputils directory contains the code for the generic network
servicing.
- The crypto subdirectory contains the Kerberos V5 encryption
library.
- The gssapi library contains the Generic Security Services API,
which is a library of commands to be used in secure client-server
communication.
- The kadm5 directory contains the libraries for the KADM5
administration utilities.
- The Kerberos 5 database libraries are contained in kdb.
- The krb5 directory contains Kerberos 5 API.
- The rpc directory contains the API for the Kerberos Remote
Procedure Call protocol.
.. _util:
util
----
The util directory contains several utility programs and libraries.
- the programs used to configure and build the code, such as
autoconf, lndir, kbuild, reconf, and makedepend, are in this
directory.
- the profile directory contains most of the functions which parse
the Kerberos configuration files (krb5.conf and kdc.conf).
- the Kerberos error table library and utilities (et);
- the Sub-system library and utilities (ss);
- database utilities (db2);
- pseudo-terminal utilities (pty);
- bug-reporting program send-pr;
- a generic support library support used by several of our other
libraries;
- the build infrastructure for building lightweight Kerberos client
(collected-client-lib)
- the tool for validating Kerberos configuration files
(confvalidator);
- the toolkit for kernel integrators for building krb5 code subsets
(gss-kernel-lib);
- source code for building Kerberos V5 on MacOS (mac)
- Windows getopt operations (windows)

148
crypto/krb5/doc/build/doing_build.rst vendored Normal file
View file

@ -0,0 +1,148 @@
Doing the build
===============
.. _do_build:
Building within a single tree
-----------------------------
If you only need to build Kerberos for one platform, using a single
directory tree which contains both the source files and the object
files is the simplest. However, if you need to maintain Kerberos for
a large number of platforms, you will probably want to use separate
build trees for each platform. We recommend that you look at OS
Incompatibilities, for notes that we have on particular operating
systems.
If you don't want separate build trees for each architecture, then use
the following abbreviated procedure::
cd /u1/krb5-VERSION/src
./configure
make
That's it!
Building with separate build directories
----------------------------------------
If you wish to keep separate build directories for each platform, you
can do so using the following procedure. (Note, this requires that
your make program support VPATH. GNU's make will provide this
functionality, for example.) If your make program does not support
this, see the next section.
For example, if you wish to store the binaries in ``tmpbuild`` build
directory you might use the following procedure::
mkdir /u1/tmpbuild
cd /u1/tmpbuild
/u1/krb5-VERSION/src/configure
make
Building using lndir
--------------------
If you wish to keep separate build directories for each platform, and
you do not have access to a make program which supports VPATH, all is
not lost. You can use the lndir program to create symbolic link trees
in your build directory.
For example, if you wish to create a build directory for solaris
binaries you might use the following procedure::
mkdir /u1/krb5-VERSION/solaris
cd /u1/krb5-VERSION/solaris
/u1/krb5-VERSION/src/util/lndir `pwd`/../src
./configure
make
You must give an absolute pathname to lndir because it has a bug that
makes it fail for relative pathnames. Note that this version differs
from the latest version as distributed and installed by the
XConsortium with X11R6. Either version should be acceptable.
Installing the binaries
-----------------------
Once you have built Kerberos, you should install the binaries. You can
do this by running::
make install
If you want to install the binaries into a destination directory that
is not their final destination, which may be convenient if you want to
build a binary distribution to be deployed on multiple hosts, you may
use::
make install DESTDIR=/path/to/destdir
This will install the binaries under *DESTDIR/PREFIX*, e.g., the user
programs will install into *DESTDIR/PREFIX/bin*, the libraries into
*DESTDIR/PREFIX/lib*, etc. *DESTDIR* must be an absolute path.
Some implementations of make allow multiple commands to be run in
parallel, for faster builds. We test our Makefiles in parallel builds
with GNU make only; they may not be compatible with other parallel
build implementations.
Testing the build
-----------------
The Kerberos V5 distribution comes with built-in regression tests. To
run them, simply type the following command while in the top-level
build directory (i.e., the directory where you sent typed make to
start building Kerberos; see :ref:`do_build`)::
make check
On some operating systems, you have to run ``make install`` before
running ``make check``, or the test suite will pick up installed
versions of Kerberos libraries rather than the newly built ones. You
can install into a prefix that isn't in the system library search
path, though. Alternatively, you can configure with
**-**\ **-disable-rpath**, which renders the build tree less suitable
for installation, but allows testing without interference from
previously installed libraries.
There are additional regression tests available, which are not run
by ``make check``. These tests require manual setup and teardown of
support infrastructure which is not easily automated, or require
excessive resources for ordinary use. The procedure for running
the manual tests is documented at
https://k5wiki.kerberos.org/wiki/Manual_Testing.
Cleaning up the build
---------------------
* Use ``make clean`` to remove all files generated by running make
command.
* Use ``make distclean`` to remove all files generated by running
./configure script. After running ``make distclean`` your source
tree (ideally) should look like the raw (just un-tarred) source
tree.
Using autoconf
--------------
(If you are not a developer, you can ignore this section.)
In the Kerberos V5 source directory, there is a configure script which
automatically determines the compilation environment and creates the
proper Makefiles for a particular platform. This configure script is
generated using autoconf, which you should already have installed if
you will be making changes to ``src/configure.in``.
Normal users will not need to worry about running autoconf; the
distribution comes with the configure script already prebuilt.
The autoconf package comes with a script called ``autoreconf`` that
will automatically run ``autoconf`` and ``autoheader`` as needed. You
should run ``autoreconf`` from the top source directory, e.g.::
cd /u1/krb5-VERSION/src
autoreconf --verbose

63
crypto/krb5/doc/build/index.rst vendored Normal file
View file

@ -0,0 +1,63 @@
.. _build_V5:
Building Kerberos V5
====================
This section details how to build and install MIT Kerberos software
from the source.
Prerequisites
-------------
In order to build Kerberos V5, you will need approximately 60-70
megabytes of disk space. The exact amount will vary depending on the
platform and whether the distribution is compiled with debugging
symbol tables or not.
Your C compiler must conform to ANSI C (ISO/IEC 9899:1990, "c89").
Some operating systems do not have an ANSI C compiler, or their
default compiler requires extra command-line options to enable ANSI C
conformance.
If you wish to keep a separate build tree, which contains the compiled
\*.o file and executables, separate from your source tree, you will
need a make program which supports **VPATH**, or you will need to use
a tool such as lndir to produce a symbolic link tree for your build
tree.
Obtaining the software
----------------------
The source code can be obtained from MIT Kerberos Distribution page,
at https://kerberos.org/dist/index.html.
The MIT Kerberos distribution comes in an archive file, generally
named krb5-VERSION-signed.tar, where *VERSION* is a placeholder for
the major and minor versions of MIT Kerberos. (For example, MIT
Kerberos 1.9 has major version "1" and minor version "9".)
The krb5-VERSION-signed.tar contains a compressed tar file consisting
of the sources for all of Kerberos (generally named
krb5-VERSION.tar.gz) and a PGP signature file for this source tree
(generally named krb5-VERSION.tar.gz.asc). MIT highly recommends that
you verify the integrity of the source code using this signature,
e.g., by running::
tar xf krb5-VERSION-signed.tar
gpg --verify krb5-VERSION.tar.gz.asc
Unpack krb5-VERSION.tar.gz in some directory. In this section we will assume
that you have chosen the top directory of the distribution the directory
``/u1/krb5-VERSION``.
Review the README file for the license, copyright and other sprecific to the
distribution information.
Contents
--------
.. toctree::
:maxdepth: 1
directory_org.rst
doing_build.rst
options2configure.rst
osconf.rst

View file

@ -0,0 +1,397 @@
.. _options2configure:
Options to *configure*
======================
There are a number of options to configure which you can use to
control how the Kerberos distribution is built.
Most commonly used options
--------------------------
**-**\ **-help**
Provides help to configure. This will list the set of commonly
used options for building Kerberos.
**-**\ **-prefix=**\ *PREFIX*
By default, Kerberos will install the package's files rooted at
``/usr/local``. If you desire to place the binaries into the
directory *PREFIX*, use this option.
**-**\ **-exec-prefix=**\ *EXECPREFIX*
This option allows one to separate the architecture independent
programs from the host-dependent files (configuration files,
manual pages). Use this option to install architecture-dependent
programs in *EXECPREFIX*. The default location is the value of
specified by **-**\ **-prefix** option.
**-**\ **-localstatedir=**\ *LOCALSTATEDIR*
This option sets the directory for locally modifiable
single-machine data. In Kerberos, this mostly is useful for
setting a location for the KDC data files, as they will be
installed in ``LOCALSTATEDIR/krb5kdc``, which is by default
``PREFIX/var/krb5kdc``.
**-**\ **-with-netlib**\ [=\ *libs*]
Allows for suppression of or replacement of network libraries. By
default, Kerberos V5 configuration will look for ``-lnsl`` and
``-lsocket``. If your operating system has a broken resolver
library or fails to pass the tests in ``src/tests/resolv``, you
will need to use this option.
**-**\ **-enable-dns-for-realm**
Enable the use of DNS to look up a host's Kerberos realm,
if the information is not provided in
:ref:`krb5.conf(5)`. See :ref:`mapping_hostnames`
for information about using DNS to determine the default realm.
DNS lookups for realm names are disabled by default.
**-**\ **-with-system-et**
Use an installed version of the error-table (et) support software,
the compile_et program, the com_err.h header file and the com_err
library. If these are not in the default locations, you may wish
to specify ``CPPFLAGS=-I/some/dir`` and
``LDFLAGS=-L/some/other/dir`` options at configuration time as
well.
If this option is not given, a version supplied with the Kerberos
sources will be built and installed along with the rest of the
Kerberos tree, for Kerberos applications to link against.
**-**\ **-with-system-ss**
Use an installed version of the subsystem command-line interface
software, the mk_cmds program, the ``ss/ss.h`` header file and the
ss library. If these are not in the default locations, you may
wish to specify ``CPPFLAGS=-I/some/dir`` and
``LDFLAGS=-L/some/other/dir`` options at configuration time as
well. See also the **SS_LIB** option.
If this option is not given, the ss library supplied with the
Kerberos sources will be compiled and linked into those programs
that need it; it will not be installed separately.
**-**\ **-with-system-db**
Use an installed version of the Berkeley DB package, which must
provide an API compatible with version 1.85. This option is
unsupported and untested. In particular, we do not know if the
database-rename code used in the dumpfile load operation will
behave properly.
If this option is not given, a version supplied with the Kerberos
sources will be built and installed. (We are not updating this
version at this time because of licensing issues with newer
versions that we haven't investigated sufficiently yet.)
Environment variables
---------------------
**CC=**\ *COMPILER*
Use *COMPILER* as the C compiler.
**CFLAGS=**\ *FLAGS*
Use *FLAGS* as the default set of C compiler flags.
**CPP=**\ *CPP*
C preprocessor to use. (e.g., ``CPP='gcc -E'``)
**CPPFLAGS=**\ *CPPOPTS*
Use *CPPOPTS* as the default set of C preprocessor flags. The
most common use of this option is to select certain #define's for
use with the operating system's include files.
**DB_HEADER=**\ *headername*
If db.h is not the correct header file to include to compile
against the Berkeley DB 1.85 API, specify the correct header file
name with this option. For example, ``DB_HEADER=db3/db_185.h``.
**DB_LIB=**\ *libs*...
If ``-ldb`` is not the correct library specification for the
Berkeley DB library version to be used, override it with this
option. For example, ``DB_LIB=-ldb-3.3``.
**DEFCCNAME=**\ *ccachename*
Override the built-in default credential cache name.
For example, ``DEFCCNAME=DIR:/var/run/user/%{USERID}/ccache``
See :ref:`parameter_expansion` for information about supported
parameter expansions.
**DEFCKTNAME=**\ *keytabname*
Override the built-in default client keytab name.
The format is the same as for *DEFCCNAME*.
**DEFKTNAME=**\ *keytabname*
Override the built-in default keytab name.
The format is the same as for *DEFCCNAME*.
**LD=**\ *LINKER*
Use *LINKER* as the default loader if it should be different from
C compiler as specified above.
**LDFLAGS=**\ *LDOPTS*
This option informs the linker where to get additional libraries
(e.g., ``-L<lib dir>``).
**LIBS=**\ *LDNAME*
This option allows one to specify libraries to be passed to the
linker (e.g., ``-l<library>``)
**PKCS11_MODNAME=**\ *library*
Override the built-in default PKCS11 library name.
**SS_LIB=**\ *libs*...
If ``-lss`` is not the correct way to link in your installed ss
library, for example if additional support libraries are needed,
specify the correct link options here. Some variants of this
library are around which allow for Emacs-like line editing, but
different versions require different support libraries to be
explicitly specified.
This option is ignored if **-**\ **-with-system-ss** is not specified.
**YACC**
The 'Yet Another C Compiler' implementation to use. Defaults to
the first program found out of: '`bison -y`', '`byacc`',
'`yacc`'.
**YFLAGS**
The list of arguments that will be passed by default to $YACC.
This script will default YFLAGS to the empty string to avoid a
default value of ``-d`` given by some make applications.
Fine tuning of the installation directories
-------------------------------------------
**-**\ **-bindir=**\ *DIR*
User executables. Defaults to ``EXECPREFIX/bin``, where
*EXECPREFIX* is the path specified by **-**\ **-exec-prefix**
configuration option.
**-**\ **-sbindir=**\ *DIR*
System admin executables. Defaults to ``EXECPREFIX/sbin``, where
*EXECPREFIX* is the path specified by **-**\ **-exec-prefix**
configuration option.
**-**\ **-sysconfdir=**\ *DIR*
Read-only single-machine data such as krb5.conf.
Defaults to ``PREFIX/etc``, where
*PREFIX* is the path specified by **-**\ **-prefix** configuration
option.
**-**\ **-libdir=**\ *DIR*
Object code libraries. Defaults to ``EXECPREFIX/lib``, where
*EXECPREFIX* is the path specified by **-**\ **-exec-prefix**
configuration option.
**-**\ **-includedir=**\ *DIR*
C header files. Defaults to ``PREFIX/include``, where *PREFIX* is
the path specified by **-**\ **-prefix** configuration option.
**-**\ **-datarootdir=**\ *DATAROOTDIR*
Read-only architecture-independent data root. Defaults to
``PREFIX/share``, where *PREFIX* is the path specified by
**-**\ **-prefix** configuration option.
**-**\ **-datadir=**\ *DIR*
Read-only architecture-independent data. Defaults to path
specified by **-**\ **-datarootdir** configuration option.
**-**\ **-localedir=**\ *DIR*
Locale-dependent data. Defaults to ``DATAROOTDIR/locale``, where
*DATAROOTDIR* is the path specified by **-**\ **-datarootdir**
configuration option.
**-**\ **-mandir=**\ *DIR*
Man documentation. Defaults to ``DATAROOTDIR/man``, where
*DATAROOTDIR* is the path specified by **-**\ **-datarootdir**
configuration option.
Program names
-------------
**-**\ **-program-prefix=**\ *PREFIX*
Prepend *PREFIX* to the names of the programs when installing
them. For example, specifying ``--program-prefix=mit-`` at the
configure time will cause the program named ``abc`` to be
installed as ``mit-abc``.
**-**\ **-program-suffix=**\ *SUFFIX*
Append *SUFFIX* to the names of the programs when installing them.
For example, specifying ``--program-suffix=-mit`` at the configure
time will cause the program named ``abc`` to be installed as
``abc-mit``.
**-**\ **-program-transform-name=**\ *PROGRAM*
Run ``sed -e PROGRAM`` on installed program names. (*PROGRAM* is a
sed script).
System types
------------
**-**\ **-build=**\ *BUILD*
Configure for building on *BUILD*
(e.g., ``--build=x86_64-linux-gnu``).
**-**\ **-host=**\ *HOST*
Cross-compile to build programs to run on *HOST*
(e.g., ``--host=x86_64-linux-gnu``). By default, Kerberos V5
configuration will look for "build" option.
Optional features
-----------------
**-**\ **-disable-option-checking**
Ignore unrecognized --enable/--with options.
**-**\ **-disable-**\ *FEATURE*
Do not include *FEATURE* (same as --enable-FEATURE=no).
**-**\ **-enable-**\ *FEATURE*\ [=\ *ARG*]
Include *FEATURE* [ARG=yes].
**-**\ **-enable-maintainer-mode**
Enable rebuilding of source files, Makefiles, etc.
**-**\ **-disable-delayed-initialization**
Initialize library code when loaded. Defaults to delay until
first use.
**-**\ **-disable-thread-support**
Don't enable thread support. Defaults to enabled.
**-**\ **-disable-rpath**
Suppress run path flags in link lines.
**-**\ **-enable-athena**
Build with MIT Project Athena configuration.
**-**\ **-disable-kdc-lookaside-cache**
Disable the cache which detects client retransmits.
**-**\ **-disable-pkinit**
Disable PKINIT plugin support.
**-**\ **-disable-aesni**
Disable support for using AES instructions on x86 platforms.
**-**\ **-enable-asan**\ [=\ *ARG*]
Enable building with asan memory error checking. If *ARG* is
given, it controls the -fsanitize compilation flag value (the
default is "address").
Optional packages
-----------------
**-**\ **-with-**\ *PACKAGE*\ [=ARG\]
Use *PACKAGE* (e.g., ``--with-imap``). The default value of *ARG*
is ``yes``.
**-**\ **-without-**\ *PACKAGE*
Do not use *PACKAGE* (same as ``--with-PACKAGE=no``)
(e.g., ``--without-libedit``).
**-**\ **-with-size-optimizations**
Enable a few optimizations to reduce code size possibly at some
run-time cost.
**-**\ **-with-system-et**
Use the com_err library and compile_et utility that are already
installed on the system, instead of building and installing
local versions.
**-**\ **-with-system-ss**
Use the ss library and mk_cmds utility that are already installed
on the system, instead of building and using private versions.
**-**\ **-with-system-db**
Use the berkeley db utility already installed on the system,
instead of using a private version. This option is not
recommended; enabling it may result in incompatibility with key
databases originating on other systems.
**-**\ **-with-netlib=**\ *LIBS*
Use the resolver library specified in *LIBS*. Use this variable
if the C library resolver is insufficient or broken.
**-**\ **-with-hesiod=**\ *path*
Compile with Hesiod support. The *path* points to the Hesiod
directory. By default Hesiod is unsupported.
**-**\ **-with-ldap**
Compile OpenLDAP database backend module.
**-**\ **-with-lmdb**
Compile LMDB database backend module.
**-**\ **-with-vague-errors**
Do not send helpful errors to client. For example, if the KDC
should return only vague error codes to clients.
**-**\ **-with-crypto-impl=**\ *IMPL*
Use specified crypto implementation (e.g., **-**\
**-with-crypto-impl=**\ *openssl*). The default is the native MIT
Kerberos implementation ``builtin``. The other currently
implemented crypto backend is ``openssl``. (See
:ref:`mitK5features`)
**-**\ **-without-libedit**
Do not compile and link against libedit. Some utilities will no
longer offer command history or completion in interactive mode if
libedit is disabled.
**-**\ **-with-readline**
Compile and link against GNU readline, as an alternative to libedit.
**-**\ **-with-system-verto**
Use an installed version of libverto. If the libverto header and
library are not in default locations, you may wish to specify
``CPPFLAGS=-I/some/dir`` and ``LDFLAGS=-L/some/other/dir`` options
at configuration time as well.
If this option is not given, the build system will try to detect
an installed version of libverto and use it if it is found.
Otherwise, a version supplied with the Kerberos sources will be
built and installed. The built-in version does not contain the
full set of back-end modules and is not a suitable general
replacement for the upstream version, but will work for the
purposes of Kerberos.
Specifying **-**\ **-without-system-verto** will cause the built-in
version of libverto to be used unconditionally.
**-**\ **-with-krb5-config=**\ *PATH*
Use the krb5-config program at *PATH* to obtain the build-time
default credential cache, keytab, and client keytab names. The
default is to use ``krb5-config`` from the program path. Specify
``--without-krb5-config`` to disable the use of krb5-config and
use the usual built-in defaults.
**-**\ **-without-keyutils**
Build without libkeyutils support. This disables the KEYRING
credential cache type.
Examples
--------
For example, in order to configure Kerberos on a Solaris machine using
the suncc compiler with the optimizer turned on, run the configure
script with the following options::
% ./configure CC=suncc CFLAGS=-O
For a slightly more complicated example, consider a system where
several packages to be used by Kerberos are installed in
``/usr/foobar``, including Berkeley DB 3.3, and an ss library that
needs to link against the curses library. The configuration of
Kerberos might be done thus::
./configure CPPFLAGS=-I/usr/foobar/include LDFLAGS=-L/usr/foobar/lib \
--with-system-et --with-system-ss --with-system-db \
SS_LIB='-lss -lcurses' DB_HEADER=db3/db_185.h DB_LIB=-ldb-3.3

26
crypto/krb5/doc/build/osconf.rst vendored Normal file
View file

@ -0,0 +1,26 @@
osconf.hin
==========
There is one configuration file which you may wish to edit to control
various compile-time parameters in the Kerberos distribution::
include/osconf.hin
The list that follows is by no means complete, just some of the more
interesting variables.
**DEFAULT_PROFILE_PATH**
The pathname to the file which contains the profiles for the known
realms, their KDCs, etc. The default value is |krb5conf|.
**DEFAULT_KEYTAB_NAME**
The type and pathname to the default server keytab file. The
default is |keytab|.
**DEFAULT_KDC_ENCTYPE**
The default encryption type for the KDC database master key. The
default value is |defmkey|.
**RCTMPDIR**
The directory which stores replay caches. The default is
``/var/tmp``.
**DEFAULT_KDB_FILE**
The location of the default database. The default value is
|kdcdir|\ ``/principal``.

View file

@ -0,0 +1,82 @@
How to build this documentation from the source
===============================================
Pre-requisites for a simple build, or to update man pages:
* Sphinx 1.0.4 or higher (See https://www.sphinx-doc.org) with the
autodoc extension installed.
Additional prerequisites to include the API reference based on Doxygen
markup:
* Python 2.5 with the Cheetah, lxml, and xml modules
* Doxygen
Simple build without API reference
----------------------------------
To test simple changes to the RST sources, you can build the
documentation without the Doxygen reference by running, from the doc
directory::
sphinx-build . test_html
You will see a number of warnings about missing files. This is
expected. If there is not already a ``doc/version.py`` file, you will
need to create one by first running ``make version.py`` in the
``src/doc`` directory of a configured build tree.
Updating man pages
------------------
Man pages are generated from the RST sources and checked into the
``src/man`` directory of the repository. This allows man pages to be
installed without requiring Sphinx when using a source checkout. To
regenerate these files, run ``make man`` from the man subdirectory
of a configured build tree. You can also do this from an unconfigured
source tree with::
cd src/man
make -f Makefile.in top_srcdir=.. srcdir=. man
make clean
As with the simple build, it is normal to see warnings about missing
files when rebuilding the man pages.
Building for a release tarball or web site
------------------------------------------
To generate documentation in HTML format, run ``make html`` in the
``doc`` subdirectory of a configured build tree (the build directory
corresponding to ``src/doc``, not the top-level ``doc`` directory).
The output will be placed in the top-level ``doc/html`` directory.
This build will include the API reference generated from Doxygen
markup in the source tree.
Documentation generated this way will use symbolic names for paths
(like ``BINDIR`` for the directory containing user programs), with the
symbolic names being links to a table showing typical values for those
paths.
You can also do this from an unconfigured source tree with::
cd src/doc
make -f Makefile.in SPHINX_ARGS= htmlsrc
Building for an OS package or site documentation
------------------------------------------------
To generate documentation specific to a build of MIT krb5 as you have
configured it, run ``make substhtml`` in the ``doc`` subdirectory of a
configured build tree (the build directory corresponding to
``src/doc``, not the top-level ``doc`` directory). The output will be
placed in the ``html_subst`` subdirectory of that build directory.
This build will include the API reference.
Documentation generated this way will use concrete paths (like
``/usr/local/bin`` for the directory containing user programs, for a
default custom build).

View file

@ -0,0 +1,5 @@
Please see
https://k5wiki.kerberos.org/wiki/Coding_style
for the current coding style.

320
crypto/krb5/doc/conf.py Normal file
View file

@ -0,0 +1,320 @@
# -*- coding: utf-8 -*-
#
# MIT Kerberos documentation build configuration file, created by
# sphinx-quickstart on Wed Oct 13 09:14:03 2010.
#
# This file is execfile()d with the current directory set to its containing dir.
#
# Note that not all possible configuration values are present in this
# autogenerated file.
#
# All configuration values have a default; values that are commented out
# serve to show the default.
import sys, os
# If extensions (or modules to document with autodoc) are in another directory,
# add these directories to sys.path here. If the directory is relative to the
# documentation root, use os.path.abspath to make it absolute, like shown here.
#sys.path.insert(0, os.path.abspath('.'))
# -- General configuration -----------------------------------------------------
# If your documentation needs a minimal Sphinx version, state it here.
#needs_sphinx = '1.0'
# Add any Sphinx extension module names here, as strings. They can be extensions
# coming with Sphinx (named 'sphinx.ext.*') or your custom ones.
#extensions = ['sphinx.ext.autodoc', 'sphinxcontrib.doxylink']
extensions = ['sphinx.ext.autodoc']
# Add any paths that contain templates here, relative to this directory.
templates_path = ['_templates']
# The suffix of source filenames.
source_suffix = '.rst'
# The encoding of source files.
#source_encoding = 'utf-8-sig'
# The master toctree document.
if 'notice' in tags:
master_doc = 'notice'
else:
master_doc = 'index'
# General information about the project.
project = u'MIT Kerberos'
copyright = u'1985-2024, MIT'
# The version info for the project you're documenting, acts as replacement for
# |version| and |release|, also used in various other places throughout the
# built documents.
exec(open("version.py").read())
# The short X.Y version.
r_list = [r_major, r_minor]
if r_patch:
r_list += [r_patch]
version = '.'.join(map(str, r_list))
# The full version, including alpha/beta/rc tags.
release = version
if r_tail:
release += '-' + r_tail
# The language for content autogenerated by Sphinx. Refer to documentation
# for a list of supported languages.
#language = None
# There are two options for replacing |today|: either, you set today to some
# non-false value, then it is used:
today = ' '
# Else, today_fmt is used as the format for a strftime call.
#today_fmt = '%B %d, %Y'
# List of patterns, relative to source directory, that match files and
# directories to ignore when looking for source files.
exclude_patterns = []
# The reST default role (used for this markup: `text`) to use for all documents.
#default_role = None
# If true, '()' will be appended to :func: etc. cross-reference text.
#add_function_parentheses = True
# If true, the current module name will be prepended to all description
# unit titles (such as .. function::).
#add_module_names = True
# If true, sectionauthor and moduleauthor directives will be shown in the
# output. They are ignored by default.
#show_authors = False
# The name of the Pygments (syntax highlighting) style to use.
pygments_style = 'sphinx'
# A list of ignored prefixes for module index sorting.
#modindex_common_prefix = []
# -- Options for HTML output ---------------------------------------------------
# When we can rely on Sphinx 1.8 (released Sep 2018) we can just set:
# html_css_files = ['kerb.css']
# But in the meantime, we add this file using either a way that works
# after 1.8 or a way that works before 4.0.
def setup(app):
if callable(getattr(app, 'add_css_file', None)):
app.add_css_file('kerb.css')
else:
app.add_stylesheet('kerb.css')
# The theme to use for HTML and HTML Help pages. See the documentation for
# a list of builtin themes.
# html_theme = 'default'
html_theme = 'agogo'
# Theme options are theme-specific and customize the look and feel of a theme
# further. For a list of options available for each theme, see the
# documentation.
html_theme_options = { "linkcolor": "#881f0d", "footerbg": "#5d1509",
"bgcolor": "#5d1509", "documentwidth": "80%",
"pagewidth": "auto", "sidebarwidth": "20%" }
# Add any paths that contain custom themes here, relative to this directory.
#html_theme_path = []
# The name for this set of Sphinx documents. If None, it defaults to
# "<project> v<release> documentation".
html_title = "MIT Kerberos Documentation"
# A shorter title for the navigation bar. Default is the same as html_title.
#html_short_title = None
# The name of an image file (relative to this directory) to place at the top
# of the sidebar.
if os.environ.get('HTML_LOGO'):
html_logo = os.environ['HTML_LOGO']
# The name of an image file (within the static path) to use as favicon of the
# docs. This file should be a Windows icon file (.ico) being 16x16 or 32x32
# pixels large.
#html_favicon = None
# Add any paths that contain custom static files (such as style sheets) here,
# relative to this directory. They are copied after the builtin static files,
# so a file named "default.css" will overwrite the builtin "default.css".
html_static_path = ['_static']
# If not '', a 'Last updated on:' timestamp is inserted at every page bottom,
# using the given strftime format.
#html_last_updated_fmt = '%b %d, %Y'
# Custom sidebar templates, maps document names to template names.
#html_sidebars = {}
# Additional templates that should be rendered to pages, maps page names to
# template names.
#html_additional_pages = {}
# If false, no module index is generated.
#html_domain_indices = True
# If false, no index is generated.
#html_use_index = True
# If true, the index is split into individual pages for each letter.
html_split_index = True
# If true, links to the reST sources are added to the pages.
html_show_sourcelink = False
# If true, "Created using Sphinx" is shown in the HTML footer. Default is True.
#html_show_sphinx = True
# If true, "(C) Copyright ..." is shown in the HTML footer. Default is True.
#html_show_copyright = True
# If true, an OpenSearch description file will be output, and all pages will
# contain a <link> tag referring to it. The value of this option must be the
# base URL from which the finished HTML is served.
#html_use_opensearch = ''
# This is the file name suffix for HTML files (e.g., ".xhtml").
#html_file_suffix = None
# Output file base name for HTML help builder.
htmlhelp_basename = 'MIT Kerberos'
pointsize = '10pt'
# -- Options for LaTeX output --------------------------------------------------
# The paper size ('letter' or 'a4').
#latex_paper_size = 'letter'
# The font size ('10pt', '11pt' or '12pt').
#latex_font_size = '10pt'
# Grouping the document tree into LaTeX files. List of tuples
# (source start file, target name, title, author, documentclass [howto/manual]).
latex_documents = [
('admin/index', 'admin.tex', u"Kerberos Administration Guide", u'MIT',
'manual'),
('appdev/index', 'appdev.tex', u"Kerberos Application Developer Guide",
u'MIT', 'manual'),
('basic/index', 'basic.tex', u"Kerberos Concepts", u'MIT', 'manual'),
('build/index', 'build.tex', u"Building MIT Kerberos", u'MIT', 'manual'),
('plugindev/index', 'plugindev.tex', u"Kerberos Plugin Module Developer Guide",
u'MIT', 'manual'),
('user/index', 'user.tex', u"Kerberos User Guide", u'MIT', 'manual')
]
# The name of an image file (relative to this directory) to place at the top of
# the title page.
#latex_logo = None
# For "manual" documents, if this is true, then toplevel headings are parts,
# not chapters.
#latex_use_parts = False
# If true, show page references after internal links.
#latex_show_pagerefs = False
# If true, show URL addresses after external links.
#latex_show_urls = False
# Additional stuff for the LaTeX preamble.
#latex_preamble = ''
# Documents to append as an appendix to all manuals.
#latex_appendices = []
# If false, no module index is generated.
#latex_domain_indices = True
if 'mansubs' in tags:
bindir = '``@BINDIR@``'
sbindir = '``@SBINDIR@``'
libdir = '``@LIBDIR@``'
localstatedir = '``@LOCALSTATEDIR@``'
runstatedir = '``@RUNSTATEDIR@``'
sysconfdir = '``@SYSCONFDIR@``'
ccache = '``@CCNAME@``'
keytab = '``@KTNAME@``'
ckeytab = '``@CKTNAME@``'
pkcs11_modname = '``@PKCS11MOD@``'
elif 'pathsubs' in tags:
# Read configured paths from a file produced by the build system.
exec(open("paths.py").read())
else:
bindir = ':ref:`BINDIR <paths>`'
sbindir = ':ref:`SBINDIR <paths>`'
libdir = ':ref:`LIBDIR <paths>`'
localstatedir = ':ref:`LOCALSTATEDIR <paths>`'
runstatedir = ':ref:`RUNSTATEDIR <paths>`'
sysconfdir = ':ref:`SYSCONFDIR <paths>`'
ccache = ':ref:`DEFCCNAME <paths>`'
keytab = ':ref:`DEFKTNAME <paths>`'
ckeytab = ':ref:`DEFCKTNAME <paths>`'
pkcs11_modname = ':ref:`PKCS11_MODNAME <paths>`'
rst_epilog = '\n'
if 'notice' in tags:
exclude_patterns = [ 'admin', 'appdev', 'basic', 'build',
'plugindev', 'user' ]
exclude_patterns += [ 'about.rst', 'build_this.rst', 'copyright.rst',
'index.rst', 'mitK5*.rst', 'resources.rst' ]
rst_epilog += '.. |copy| replace:: \(C\)'
else:
exclude_patterns += [ 'notice.rst' ]
rst_epilog += '.. |bindir| replace:: %s\n' % bindir
rst_epilog += '.. |sbindir| replace:: %s\n' % sbindir
rst_epilog += '.. |libdir| replace:: %s\n' % libdir
rst_epilog += '.. |kdcdir| replace:: %s\\ ``/krb5kdc``\n' % localstatedir
rst_epilog += '.. |kdcrundir| replace:: %s\\ ``/krb5kdc``\n' % runstatedir
rst_epilog += '.. |sysconfdir| replace:: %s\n' % sysconfdir
rst_epilog += '.. |ccache| replace:: %s\n' % ccache
rst_epilog += '.. |keytab| replace:: %s\n' % keytab
rst_epilog += '.. |ckeytab| replace:: %s\n' % ckeytab
rst_epilog += '.. |pkcs11_modname| replace:: %s\n' % pkcs11_modname
rst_epilog += '''
.. |krb5conf| replace:: ``/etc/krb5.conf``
.. |defkeysalts| replace:: ``aes256-cts-hmac-sha1-96:normal aes128-cts-hmac-sha1-96:normal``
.. |defetypes| replace:: ``aes256-cts-hmac-sha1-96 aes128-cts-hmac-sha1-96 aes256-cts-hmac-sha384-192 aes128-cts-hmac-sha256-128 des3-cbc-sha1 arcfour-hmac-md5 camellia256-cts-cmac camellia128-cts-cmac``
.. |defmkey| replace:: ``aes256-cts-hmac-sha1-96``
.. |copy| unicode:: U+000A9
'''
# -- Options for manual page output --------------------------------------------
# One entry per manual page. List of tuples
# (source start file, name, description, authors, manual section).
man_pages = [
('user/user_commands/kinit', 'kinit', u'obtain and cache Kerberos ticket-granting ticket', [u'MIT'], 1),
('user/user_commands/klist', 'klist', u'list cached Kerberos tickets', [u'MIT'], 1),
('user/user_commands/kdestroy', 'kdestroy', u'destroy Kerberos tickets', [u'MIT'], 1),
('user/user_commands/kswitch', 'kswitch', u'switch primary ticket cache', [u'MIT'], 1),
('user/user_commands/kpasswd', 'kpasswd', u'change a user\'s Kerberos password', [u'MIT'], 1),
('user/user_commands/kvno', 'kvno', u'print key version numbers of Kerberos principals', [u'MIT'], 1),
('user/user_commands/ksu', 'ksu', u'Kerberized super-user', [u'MIT'], 1),
('user/user_commands/krb5-config', 'krb5-config', u'tool for linking against MIT Kerberos libraries', [u'MIT'], 1),
('user/user_config/k5login', 'k5login', u'Kerberos V5 acl file for host access', [u'MIT'], 5),
('user/user_config/k5identity', 'k5identity', u'Kerberos V5 client principal selection rules', [u'MIT'], 5),
('user/user_config/kerberos', 'kerberos', u'Overview of using Kerberos', [u'MIT'], 7),
('admin/admin_commands/krb5kdc', 'krb5kdc', u'Kerberos V5 KDC', [u'MIT'], 8),
('admin/admin_commands/kadmin_local', 'kadmin', u'Kerberos V5 database administration program', [u'MIT'], 1),
('admin/admin_commands/kprop', 'kprop', u'propagate a Kerberos V5 principal database to a replica server', [u'MIT'], 8),
('admin/admin_commands/kproplog', 'kproplog', u'display the contents of the Kerberos principal update log', [u'MIT'], 8),
('admin/admin_commands/kpropd', 'kpropd', u'Kerberos V5 replica KDC update server', [u'MIT'], 8),
('admin/admin_commands/kdb5_util', 'kdb5_util', u'Kerberos database maintenance utility', [u'MIT'], 8),
('admin/admin_commands/ktutil', 'ktutil', u'Kerberos keytab file maintenance utility', [u'MIT'], 1),
('admin/admin_commands/k5srvutil', 'k5srvutil', u'host key table (keytab) manipulation utility', [u'MIT'], 1),
('admin/admin_commands/kadmind', 'kadmind', u'KADM5 administration server', [u'MIT'], 8),
('admin/admin_commands/kdb5_ldap_util', 'kdb5_ldap_util', u'Kerberos configuration utility', [u'MIT'], 8),
('admin/conf_files/krb5_conf', 'krb5.conf', u'Kerberos configuration file', [u'MIT'], 5),
('admin/conf_files/kdc_conf', 'kdc.conf', u'Kerberos V5 KDC configuration file', [u'MIT'], 5),
('admin/conf_files/kadm5_acl', 'kadm5.acl', u'Kerberos ACL file', [u'MIT'], 5),
('user/user_commands/sclient', 'sclient', u'sample Kerberos version 5 client', [u'MIT'], 1),
('admin/admin_commands/sserver', 'sserver', u'sample Kerberos version 5 server', [u'MIT'], 8),
]

View file

@ -0,0 +1,70 @@
Contributing to MIT Kerberos
DESIGN
======
If you are planning to contribute a substantial amount of work, please
ensure that you have a discussion about the design on the
krbdev@mit.edu list. Some changes may require coordination with
standards groups. For example, interface changes and extensions for
the GSS-API should be discussed in the IETF KITTEN Working Group.
STYLE
=====
Please follow the guidelines in doc/coding-style for new code. For
existing code, please preserve its existing indentation and brace
conventions. These existing conventions usually resemble the
guidelines in doc/coding-style. Exceptions to the style in
doc/coding-style are usually large past contributions or imports from
other parties. These include (not an exhaustive list):
src/appl/bsd
src/appl/gssftp
src/appl/telnet
src/kadmin
src/lib/kadm5
src/lib/gssapi/mechglue
src/lib/rpc
PATCHES
=======
We prefer patches in either unified or context diff format (diff -u or
diff -c). As is usual practice, please specify the original file
before the modified file on the diff command line. It's also useful
to perform the diff from the top level of the tree, e.g.,
diff -ur src.orig src
It's even more useful if you use our anonymous Subversion repository
at
svn://anonsvn.mit.edu/krb5
and use "svn diff" (or "svk diff" if you prefer to use SVK) to
generate your patches.
It is much easier for us to integrate patches which are generated
against current code on the trunk. Please ensure that your source
tree is up-to-date before generating your patch.
COPYRIGHT
=========
If you are submitting substantial quantities of new code, or are
substantially modifying existing code, please be clear about the
copyright status of your contributions. Note that if your
contribution was created in the course of your employment, your
employer may own copyright in your contribution.
We prefer that MIT receives the ownership of the contributions, but
will generally accept contributed code with copyright owned by other
parties provided that the license conditions are substantially
identical to the existing license on the MIT krb5 code.
Appropriate copyright notices and license terms should be added to new
or changed files, unless the contributed code is being assigned to the
already-listed copyright holder in the file, or the contribution is
being released to the public domain. Please make sure that the
year in the copyright statement is kept up-to-date.

View file

@ -0,0 +1,8 @@
Copyright
=========
Copyright |copy| 1985-2024 by the Massachusetts Institute of
Technology and its contributors. All rights reserved.
See :ref:`mitK5license` for additional copyright and license
information.

View file

@ -0,0 +1,33 @@
/** @example cc_set_config.c
*
* Usage examples for krb5_cc_set_config and krb5_cc_get_config functions
*/
#include <k5-int.h>
krb5_error_code
func_set(krb5_context context, krb5_ccache id,
krb5_const_principal principal, const char *key)
{
krb5_data config_data;
config_data.data = "yes";
config_data.length = strlen(config_data.data);
return krb5_cc_set_config(context, id, principal, key, &config_data);
}
krb5_error_code
func_get(krb5_context context, krb5_ccache id,
krb5_const_principal principal, const char *key)
{
krb5_data config_data;
krb5_error_code ret;
config_data.data = NULL;
ret = krb5_cc_get_config(context, id, principal, key, &config_data);
if (ret){
return ret;
}
/* do something */
krb5_free_data_contents(context, &config_data);
return ret;
}

View file

@ -0,0 +1,23 @@
/** @example cc_unique.c
*
* Usage example for krb5_cc_new_unique function
*/
#include "k5-int.h"
krb5_error_code
func(krb5_context context)
{
krb5_error_code ret;
krb5_ccache ccache = NULL;
ret = krb5_cc_new_unique(context, "MEMORY", NULL, &ccache);
if (ret){
ccache = NULL;
return ret;
}
/* do something */
if (ccache)
(void)krb5_cc_destroy(context, ccache);
return 0;
}

View file

@ -0,0 +1,20 @@
/** @example error_message.c
*
* Demo for krb5_get/set/free_error_message function family
*/
#include <k5-int.h>
krb5_error_code
func(krb5_context context)
{
krb5_error_code ret;
ret = krb5_func(context);
if (ret) {
const char *err_str = krb5_get_error_message(context, ret);
krb5_set_error_message(context, ret,
"Failed krb5_func: %s", err_str);
krb5_free_error_message(context, err_str);
}
}

View file

@ -0,0 +1,55 @@
/** @example tkt_creds.c
*
* Usage example for krb5_tkt_creds function family
*/
#include "krb5.h"
krb5_error_code
func(krb5_context context, krb5_flags options,
krb5_ccache ccache, krb5_creds *in_creds,
krb5_creds **out_creds)
{
krb5_error_code code = KRB5_OK;
krb5_creds *ncreds = NULL;
krb5_tkt_creds_context ctx = NULL;
*out_creds = NULL;
/* Allocate a container. */
ncreds = k5alloc(sizeof(*ncreds), &code);
if (ncreds == NULL)
goto cleanup;
/* Make and execute a krb5_tkt_creds context to get the credential. */
code = krb5_tkt_creds_init(context, ccache, in_creds, options, &ctx);
if (code != KRB5_OK)
goto cleanup;
code = krb5_tkt_creds_get(context, ctx);
if (code != KRB5_OK)
goto cleanup;
code = krb5_tkt_creds_get_creds(context, ctx, ncreds);
if (code != KRB5_OK)
goto cleanup;
*out_creds = ncreds;
ncreds = NULL;
cleanup:
krb5_free_creds(context, ncreds);
krb5_tkt_creds_free(context, ctx);
return code;
}
/* Allocate zeroed memory; set *code to 0 on success or ENOMEM on failure. */
static inline void *
k5alloc(size_t len, krb5_error_code *code)
{
void *ptr;
/* Allocate at least one byte since zero-byte allocs may return NULL. */
ptr = calloc((len > 0) ? len : 1, 1);
*code = (ptr == NULL) ? ENOMEM : 0;
return ptr;
}

View file

@ -0,0 +1,28 @@
/** @example verify_init_creds.c
*
* Usage example for krb5_verify_init_creds function family
*/
#include "k5-int.h"
krb5_error_code
func(krb5_context context, krb5_creds *creds, krb5_principal server_principal)
{
krb5_error_code ret = KRB5_OK;
krb5_verify_init_creds_opt options;
krb5_verify_init_creds_opt_init (&options);
krb5_verify_init_creds_opt_set_ap_req_nofail (&options, 1);
ret = krb5_verify_init_creds(context,
creds,
server_principal,
NULL /* use default keytab */,
NULL /* don't store creds in ccache */,
&options);
if (ret) {
/* error while verifying credentials for server */
}
return ret;
}

View file

@ -0,0 +1,182 @@
.. _ccache_file_format:
Credential cache file format
============================
There are four versions of the file format used by the FILE credential
cache type. The first byte of the file always has the value 5, and
the value of the second byte contains the version number (1 through
4). Versions 1 and 2 of the file format use native byte order for integer
representations. Versions 3 and 4 always use big-endian byte order.
After the two-byte version indicator, the file has three parts: the
header (in version 4 only), the default principal name, and a sequence
of credentials.
Header format
-------------
The header appears only in format version 4. It begins with a 16-bit
integer giving the length of the entire header, followed by a sequence
of fields. Each field consists of a 16-bit tag, a 16-bit length, and
a value of the given length. A file format implementation should
ignore fields with unknown tags.
At this time there is only one defined header field. Its tag value is
1, its length is always 8, and its contents are two 32-bit integers
giving the seconds and microseconds of the time offset of the KDC
relative to the client. Adding this offset to the current time on the
client should give the current time on the KDC, if that offset has not
changed since the initial authentication.
.. _cache_principal_format:
Principal format
----------------
The default principal is marshalled using the following informal
grammar::
principal ::=
name type (32 bits) [omitted in version 1]
count of components (32 bits) [includes realm in version 1]
realm (data)
component1 (data)
component2 (data)
...
data ::=
length (32 bits)
value (length bytes)
There is no external framing on the default principal, so it must be
parsed according to the above grammar in order to find the sequence of
credentials which follows.
.. _ccache_credential_format:
Credential format
-----------------
The credential format uses the following informal grammar (referencing
the ``principal`` and ``data`` types from the previous section)::
credential ::=
client (principal)
server (principal)
keyblock (keyblock)
authtime (32 bits)
starttime (32 bits)
endtime (32 bits)
renew_till (32 bits)
is_skey (1 byte, 0 or 1)
ticket_flags (32 bits)
addresses (addresses)
authdata (authdata)
ticket (data)
second_ticket (data)
keyblock ::=
enctype (16 bits) [repeated twice in version 3]
data
addresses ::=
count (32 bits)
address1
address2
...
address ::=
addrtype (16 bits)
data
authdata ::=
count (32 bits)
authdata1
authdata2
...
authdata ::=
ad_type (16 bits)
data
There is no external framing on a marshalled credential, so it must be
parsed according to the above grammar in order to find the next
credential. There is also no count of credentials or marker at the
end of the sequence of credentials; the sequence ends when the file
ends.
Credential cache configuration entries
--------------------------------------
Configuration entries are encoded as credential entries. The client
principal of the entry is the default principal of the cache. The
server principal has the realm ``X-CACHECONF:`` and two or three
components, the first of which is ``krb5_ccache_conf_data``. The
server principal's second component is the configuration key. The
third component, if it exists, is a principal to which the
configuration key is associated. The configuration value is stored in
the ticket field of the entry. All other entry fields are zeroed.
Programs using credential caches must be aware of configuration
entries for several reasons:
* A program which displays the contents of a cache should not
generally display configuration entries.
* The ticket field of a configuration entry is not (usually) a valid
encoding of a Kerberos ticket. An implementation must not treat the
cache file as malformed if it cannot decode the ticket field.
* Configuration entries have an endtime field of 0 and might therefore
always be considered expired, but they should not be treated as
unimportant as a result. For instance, a program which copies
credentials from one cache to another should not omit configuration
entries because of the endtime.
The following configuration keys are currently used in MIT krb5:
fast_avail
The presence of this key with a non-empty value indicates that the
KDC asserted support for FAST (see :rfc:`6113`) during the initial
authentication, using the negotiation method described in
:rfc:`6806` section 11. This key is not associated with any
principal.
pa_config_data
The value of this key contains a JSON object representation of
parameters remembered by the preauthentication mechanism used
during the initial authentication. These parameters may be used
when refreshing credentials. This key is associated with the
server principal of the initial authentication (usually the local
krbtgt principal of the client realm).
pa_type
The value of this key is the ASCII decimal representation of the
preauth type number used during the initial authentication. This
key is associated with the server principal of the initial
authentication.
proxy_impersonator
The presence of this key indicates that the cache is a synthetic
delegated credential for use with S4U2Proxy. The value is the
name of the intermediate service whose TGT can be used to make
S4U2Proxy requests for target services. This key is not
associated with any principal.
refresh_time
The presence of this key indicates that the cache was acquired by
the GSS mechanism using a client keytab. The value is the ASCII
decimal representation of a timestamp at which the GSS mechanism
should attempt to refresh the credential cache from the client
keytab.
start_realm
This key indicates the realm of the ticket-granting ticket to be
used for TGS requests, when making a referrals request or
beginning a cross-realm request. If it is not present, the client
realm is used.

View file

@ -0,0 +1,97 @@
KDC cookie format
=================
:rfc:`6113` section 5.2 specifies a pa-data type PA-FX-COOKIE, which
clients are required to reflect back to the KDC during
pre-authentication. The MIT krb5 KDC uses the following formats for
cookies.
Trivial cookie (version 0)
--------------------------
If there is no pre-authentication mechanism state information to save,
a trivial cookie containing the value "MIT" is used. A trivial cookie
is needed to indicate that the conversation can continue.
Secure cookie (version 1)
-------------------------
In release 1.14 and later, a secure cookie can be sent if there is any
mechanism state to save for the next request. A secure cookie
contains the concatenation of the following:
* the four bytes "MIT1"
* a four-byte big-endian kvno value
* an :rfc:`3961` ciphertext
The ciphertext is encrypted in the cookie key with key usage
number 513. The cookie key is derived from a key in the local krbtgt
principal entry for the realm (e.g. ``krbtgt/KRBTEST.COM@KRBTEST.COM``
if the request is to the ``KRBTEST.COM`` realm). The first krbtgt key
for the indicated kvno value is combined with the client principal as
follows::
cookie-key <- random-to-key(PRF+(tgt-key, "COOKIE" | client-princ))
where **random-to-key** is the :rfc:`3961` random-to-key operation for
the krbtgt key's encryption type, **PRF+** is defined in :rfc:`6113`,
and ``|`` denotes concatenation. *client-princ* is the request client
principal name with realm, marshalled according to :rfc:`1964` section
2.1.1.
The plain text of the encrypted part of a cookie is the DER encoding
of the following ASN.1 type::
SecureCookie ::= SEQUENCE {
time INTEGER,
data SEQUENCE OF PA-DATA,
...
}
The time field represents the cookie creation time; for brevity, it is
encoded as an integer giving the POSIX timestamp rather than as an
ASN.1 GeneralizedTime value. The data field contains one element for
each pre-authentication type which requires saved state. For
mechanisms which have separate request and reply types, the request
type is used; this allows the KDC to determine whether a cookie is
relevant to a request by comparing the request pa-data types to the
cookie data types.
SPAKE cookie format (version 1)
-------------------------------
Inside the SecureCookie wrapper, a data value of type 151 contains
state for SPAKE pre-authentication. This data is the concatenation of
the following:
* a two-byte big-endian version number with the value 1
* a two-byte big-endian stage number
* a four-byte big-endian group number
* a four-byte big-endian length and data for the SPAKE value
* a four-byte big-endian length and data for the transcript hash
* zero or more second factor records, each consisting of:
- a four-byte big-endian second-factor type
- a four-byte big-endian length and data
The stage value is 0 if the cookie was sent with a challenge message.
Otherwise it is 1 for the first encdata message sent by the KDC during
an exchange, 2 for the second, etc..
The group value indicates the group number used in the SPAKE challenge.
For a stage-0 cookie, the SPAKE value is the KDC private key,
represented in the scalar marshalling form of the group. For other
cookies, the SPAKE value is the SPAKE result K, represented in the
group element marshalling form.
For a stage-0 cookie, the transcript hash is the intermediate hash
after updating with the client support message (if one was sent) and
challenge. For other cookies it is the final hash.
For a stage-0 cookie, there may be any number of second-factor
records, including none; a second-factor type need not create a state
field if it does not need one, and no record is created for SF-NONE.
For other cookies, there must be exactly one second-factor record
corresponding to the factor type chosen by the client.

View file

@ -0,0 +1,19 @@
PKINIT freshness tokens
=======================
:rfc:`8070` specifies a pa-data type PA_AS_FRESHNESS, which clients
should reflect within signed PKINIT data to prove recent access to the
client certificate private key. The contents of a freshness token are
left to the KDC implementation. The MIT krb5 KDC uses the following
format for freshness tokens (starting in release 1.17):
* a four-byte big-endian POSIX timestamp
* a four-byte big-endian key version number
* an :rfc:`3961` checksum, with no ASN.1 wrapper
The checksum is computed using the first key in the local krbtgt
principal entry for the realm (e.g. ``krbtgt/KRBTEST.COM@KRBTEST.COM``
if the request is to the ``KRBTEST.COM`` realm) of the indicated key
version. The checksum type must be the mandatory checksum type for
the encryption type of the krbtgt key. The key usage value for the
checksum is 514.

View file

@ -0,0 +1,11 @@
Protocols and file formats
==========================
.. toctree::
:maxdepth: 1
ccache_file_format
keytab_file_format
rcache_file_format
cookie
freshness_token

View file

@ -0,0 +1,51 @@
.. _keytab_file_format:
Keytab file format
==================
There are two versions of the file format used by the FILE keytab
type. The first byte of the file always has the value 5, and the
value of the second byte contains the version number (1 or 2).
Version 1 of the file format uses native byte order for integer
representations. Version 2 always uses big-endian byte order.
After the two-byte version indicator, the file contains a sequence of
signed 32-bit record lengths followed by key records or holes. A
positive record length indicates a valid key entry whose size is equal
to or less than the record length. A negative length indicates a
zero-filled hole whose size is the inverse of the length. A length of
0 indicates the end of the file.
Key entry format
----------------
A key entry may be smaller in size than the record length which
precedes it, because it may have replaced a hole which is larger than
the key entry. Key entries use the following informal grammar::
entry ::=
principal
timestamp (32 bits)
key version (8 bits)
enctype (16 bits)
key length (16 bits)
key contents
key version (32 bits) [in release 1.14 and later]
principal ::=
count of components (16 bits) [includes realm in version 1]
realm (data)
component1 (data)
component2 (data)
...
name type (32 bits) [omitted in version 1]
data ::=
length (16 bits)
value (length bytes)
The 32-bit key version overrides the 8-bit key version. To determine
if it is present, the implementation must check that at least 4 bytes
remain in the record after the other fields are read, and that the
value of the 32-bit integer contained in those bytes is non-zero.

View file

@ -0,0 +1,50 @@
Replay cache file format
========================
This section documents the second version of the replay cache file
format, used by the "file2" replay cache type (new in release 1.18).
The first version of the file replay cache format is not documented.
All accesses to the replay cache file take place under an exclusive
POSIX or Windows file lock, obtained when the file is opened and
released when it is closed. Replay cache files are automatically
created when first accessed.
For each store operation, a tag is derived from the checksum part of
the :RFC:`3961` ciphertext of the authenticator. The checksum is
coerced to a fixed length of 12 bytes, either through truncation or
right-padding with zero bytes. A four-byte timestamp is appended to
the tag to produce a total record length of 16 bytes.
Bytes 0 through 15 of the file contain a hash seed for the SipHash-2-4
algorithm (siphash_); this field is populated with random bytes when
the file is first created. All remaining bytes are divided into a
series of expanding hash tables:
* Bytes 16-16383: hash table 1 (1023 slots)
* Bytes 16384-49151: hash table 2 (2048 slots)
* Bytes 49152-114687: hash table 3 (4096 slots)
* ...
Only some hash tables will be present in the file at any specific
time, and the final table may be only partially filled. Replay cache
files may be sparse if the filesystem supports it.
For each table present in the file, the tag is hashed with SipHash-2-4
using the seed recorded in the file. The first byte of the seed is
incremented by one (modulo 256) for each table after the first. The
resulting hash value is taken modulo one less than the table size
(1022 for the first hash table, 2047 for the second) to produce the
index. The record may be found at the slot given by the index or at
the next slot.
All candidate locations for the record must be searched until a slot
is found with a timestamp of zero (indicating a slot which has never
been written to) or an offset is reached at or beyond the end of the
file. Any candidate location with a timestamp value of zero, with a
timestamp value less than the current time minus clockskew, or at or
beyond the end of the file is available for writing. When all
candidate locations have been searched without finding a match, the
new entry is written to the earliest candidate available for writing.
.. _siphash: https://131002.net/siphash/siphash.pdf

View file

@ -0,0 +1,4 @@
# Sphinx build info version 1
# This file hashes the configuration used when building these files. When it is not found, a full rebuild will be done.
config: 9ca503a4e24138fa47d4451ee9426bb5
tags: 645f666f9bcd5a90fca523b33c5a78b7

View file

@ -0,0 +1,35 @@
Contributing to the MIT Kerberos Documentation
==============================================
We are looking for documentation writers and editors who could contribute
towards improving the MIT KC documentation content. If you are an experienced
Kerberos developer and/or administrator, please consider sharing your knowledge
and experience with the Kerberos Community. You can suggest your own topic or
write about any of the topics listed
`here <https://k5wiki.kerberos.org/wiki/Projects/Documentation_Tasks>`__.
If you have any questions, comments, or suggestions on the existing documents,
please send your feedback via email to krb5-bugs@mit.edu. The HTML version of
this documentation has a "FEEDBACK" link to the krb5-bugs@mit.edu email
address with a pre-constructed subject line.
Background
----------
Starting with release 1.11, the Kerberos documentation set is
unified in a central form. Man pages, HTML documentation, and PDF
documents are compiled from reStructuredText sources, and the application
developer documentation incorporates Doxygen markup from the source
tree. This project was undertaken along the outline described
`here <https://k5wiki.kerberos.org/wiki/Projects/Kerberos_Documentation>`__.
Previous versions of Kerberos 5 attempted to maintain separate documentation
in the texinfo format, with separate groff manual pages. Having the API
documentation disjoint from the source code implementing that API
resulted in the documentation becoming stale, and over time the documentation
ceased to match reality. With a fresh start and a source format that is
easier to use and maintain, reStructuredText-based documents should provide
an improved experience for the user. Consolidating all the documentation
formats into a single source document makes the documentation set easier
to maintain.

View file

@ -0,0 +1,17 @@
Administration programs
========================
.. toctree::
:maxdepth: 1
kadmin_local.rst
kadmind.rst
kdb5_util.rst
kdb5_ldap_util.rst
krb5kdc.rst
kprop.rst
kpropd.rst
kproplog.rst
ktutil.rst
k5srvutil.rst
sserver.rst

View file

@ -0,0 +1,69 @@
.. _k5srvutil(1):
k5srvutil
=========
SYNOPSIS
--------
**k5srvutil** *operation*
[**-i**]
[**-f** *filename*]
[**-e** *keysalts*]
DESCRIPTION
-----------
k5srvutil allows an administrator to list keys currently in
a keytab, to obtain new keys for a principal currently in a keytab,
or to delete non-current keys from a keytab.
*operation* must be one of the following:
**list**
Lists the keys in a keytab, showing version number and principal
name.
**change**
Uses the kadmin protocol to update the keys in the Kerberos
database to new randomly-generated keys, and updates the keys in
the keytab to match. If a key's version number doesn't match the
version number stored in the Kerberos server's database, then the
operation will fail. If the **-i** flag is given, k5srvutil will
prompt for confirmation before changing each key. If the **-k**
option is given, the old and new keys will be displayed.
Ordinarily, keys will be generated with the default encryption
types and key salts. This can be overridden with the **-e**
option. Old keys are retained in the keytab so that existing
tickets continue to work, but **delold** should be used after
such tickets expire, to prevent attacks against the old keys.
**delold**
Deletes keys that are not the most recent version from the keytab.
This operation should be used some time after a change operation
to remove old keys, after existing tickets issued for the service
have expired. If the **-i** flag is given, then k5srvutil will
prompt for confirmation for each principal.
**delete**
Deletes particular keys in the keytab, interactively prompting for
each key.
In all cases, the default keytab is used unless this is overridden by
the **-f** option.
k5srvutil uses the :ref:`kadmin(1)` program to edit the keytab in
place.
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kadmin(1)`, :ref:`ktutil(1)`, :ref:`kerberos(7)`

View file

@ -0,0 +1,985 @@
.. _kadmin(1):
kadmin
======
SYNOPSIS
--------
.. _kadmin_synopsis:
**kadmin**
[**-O**\|\ **-N**]
[**-r** *realm*]
[**-p** *principal*]
[**-q** *query*]
[[**-c** *cache_name*]\|[**-k** [**-t** *keytab*]]\|\ **-n**]
[**-w** *password*]
[**-s** *admin_server*\ [:*port*]]
[command args...]
**kadmin.local**
[**-r** *realm*]
[**-p** *principal*]
[**-q** *query*]
[**-d** *dbname*]
[**-e** *enc*:*salt* ...]
[**-m**]
[**-x** *db_args*]
[command args...]
DESCRIPTION
-----------
kadmin and kadmin.local are command-line interfaces to the Kerberos V5
administration system. They provide nearly identical functionalities;
the difference is that kadmin.local directly accesses the KDC
database, while kadmin performs operations using :ref:`kadmind(8)`.
Except as explicitly noted otherwise, this man page will use "kadmin"
to refer to both versions. kadmin provides for the maintenance of
Kerberos principals, password policies, and service key tables
(keytabs).
The remote kadmin client uses Kerberos to authenticate to kadmind
using the service principal ``kadmin/admin`` or ``kadmin/ADMINHOST``
(where *ADMINHOST* is the fully-qualified hostname of the admin
server). If the credentials cache contains a ticket for one of these
principals, and the **-c** credentials_cache option is specified, that
ticket is used to authenticate to kadmind. Otherwise, the **-p** and
**-k** options are used to specify the client Kerberos principal name
used to authenticate. Once kadmin has determined the principal name,
it requests a service ticket from the KDC, and uses that service
ticket to authenticate to kadmind.
Since kadmin.local directly accesses the KDC database, it usually must
be run directly on the primary KDC with sufficient permissions to read
the KDC database. If the KDC database uses the LDAP database module,
kadmin.local can be run on any host which can access the LDAP server.
OPTIONS
-------
.. _kadmin_options:
**-r** *realm*
Use *realm* as the default database realm.
**-p** *principal*
Use *principal* to authenticate. Otherwise, kadmin will append
``/admin`` to the primary principal name of the default ccache,
the value of the **USER** environment variable, or the username as
obtained with getpwuid, in order of preference.
**-k**
Use a keytab to decrypt the KDC response instead of prompting for
a password. In this case, the default principal will be
``host/hostname``. If there is no keytab specified with the
**-t** option, then the default keytab will be used.
**-t** *keytab*
Use *keytab* to decrypt the KDC response. This can only be used
with the **-k** option.
**-n**
Requests anonymous processing. Two types of anonymous principals
are supported. For fully anonymous Kerberos, configure PKINIT on
the KDC and configure **pkinit_anchors** in the client's
:ref:`krb5.conf(5)`. Then use the **-n** option with a principal
of the form ``@REALM`` (an empty principal name followed by the
at-sign and a realm name). If permitted by the KDC, an anonymous
ticket will be returned. A second form of anonymous tickets is
supported; these realm-exposed tickets hide the identity of the
client but not the client's realm. For this mode, use ``kinit
-n`` with a normal principal name. If supported by the KDC, the
principal (but not realm) will be replaced by the anonymous
principal. As of release 1.8, the MIT Kerberos KDC only supports
fully anonymous operation.
**-c** *credentials_cache*
Use *credentials_cache* as the credentials cache. The cache
should contain a service ticket for the ``kadmin/admin`` or
``kadmin/ADMINHOST`` (where *ADMINHOST* is the fully-qualified
hostname of the admin server) service; it can be acquired with the
:ref:`kinit(1)` program. If this option is not specified, kadmin
requests a new service ticket from the KDC, and stores it in its
own temporary ccache.
**-w** *password*
Use *password* instead of prompting for one. Use this option with
care, as it may expose the password to other users on the system
via the process list.
**-q** *query*
Perform the specified query and then exit.
**-d** *dbname*
Specifies the name of the KDC database. This option does not
apply to the LDAP database module.
**-s** *admin_server*\ [:*port*]
Specifies the admin server which kadmin should contact.
**-m**
If using kadmin.local, prompt for the database master password
instead of reading it from a stash file.
**-e** "*enc*:*salt* ..."
Sets the keysalt list to be used for any new keys created. See
:ref:`Keysalt_lists` in :ref:`kdc.conf(5)` for a list of possible
values.
**-O**
Force use of old AUTH_GSSAPI authentication flavor.
**-N**
Prevent fallback to AUTH_GSSAPI authentication flavor.
**-x** *db_args*
Specifies the database specific arguments. See the next section
for supported options.
Starting with release 1.14, if any command-line arguments remain after
the options, they will be treated as a single query to be executed.
This mode of operation is intended for scripts and behaves differently
from the interactive mode in several respects:
* Query arguments are split by the shell, not by kadmin.
* Informational and warning messages are suppressed. Error messages
and query output (e.g. for **get_principal**) will still be
displayed.
* Confirmation prompts are disabled (as if **-force** was given).
Password prompts will still be issued as required.
* The exit status will be non-zero if the query fails.
The **-q** option does not carry these behavior differences; the query
will be processed as if it was entered interactively. The **-q**
option cannot be used in combination with a query in the remaining
arguments.
.. _dboptions:
DATABASE OPTIONS
----------------
Database options can be used to override database-specific defaults.
Supported options for the DB2 module are:
**-x dbname=**\ \*filename*
Specifies the base filename of the DB2 database.
**-x lockiter**
Make iteration operations hold the lock for the duration of
the entire operation, rather than temporarily releasing the
lock while handling each principal. This is the default
behavior, but this option exists to allow command line
override of a [dbmodules] setting. First introduced in
release 1.13.
**-x unlockiter**
Make iteration operations unlock the database for each
principal, instead of holding the lock for the duration of the
entire operation. First introduced in release 1.13.
Supported options for the LDAP module are:
**-x host=**\ *ldapuri*
Specifies the LDAP server to connect to by a LDAP URI.
**-x binddn=**\ *bind_dn*
Specifies the DN used to bind to the LDAP server.
**-x bindpwd=**\ *password*
Specifies the password or SASL secret used to bind to the LDAP
server. Using this option may expose the password to other
users on the system via the process list; to avoid this,
instead stash the password using the **stashsrvpw** command of
:ref:`kdb5_ldap_util(8)`.
**-x sasl_mech=**\ *mechanism*
Specifies the SASL mechanism used to bind to the LDAP server.
The bind DN is ignored if a SASL mechanism is used. New in
release 1.13.
**-x sasl_authcid=**\ *name*
Specifies the authentication name used when binding to the
LDAP server with a SASL mechanism, if the mechanism requires
one. New in release 1.13.
**-x sasl_authzid=**\ *name*
Specifies the authorization name used when binding to the LDAP
server with a SASL mechanism. New in release 1.13.
**-x sasl_realm=**\ *realm*
Specifies the realm used when binding to the LDAP server with
a SASL mechanism, if the mechanism uses one. New in release
1.13.
**-x debug=**\ *level*
sets the OpenLDAP client library debug level. *level* is an
integer to be interpreted by the library. Debugging messages
are printed to standard error. New in release 1.12.
COMMANDS
--------
When using the remote client, available commands may be restricted
according to the privileges specified in the :ref:`kadm5.acl(5)` file
on the admin server.
.. _add_principal:
add_principal
~~~~~~~~~~~~~
**add_principal** [*options*] *newprinc*
Creates the principal *newprinc*, prompting twice for a password. If
no password policy is specified with the **-policy** option, and the
policy named ``default`` is assigned to the principal if it exists.
However, creating a policy named ``default`` will not automatically
assign this policy to previously existing principals. This policy
assignment can be suppressed with the **-clearpolicy** option.
This command requires the **add** privilege.
Aliases: **addprinc**, **ank**
Options:
**-expire** *expdate*
(:ref:`getdate` string) The expiration date of the principal.
**-pwexpire** *pwexpdate*
(:ref:`getdate` string) The password expiration date.
**-maxlife** *maxlife*
(:ref:`duration` or :ref:`getdate` string) The maximum ticket life
for the principal.
**-maxrenewlife** *maxrenewlife*
(:ref:`duration` or :ref:`getdate` string) The maximum renewable
life of tickets for the principal.
**-kvno** *kvno*
The initial key version number.
**-policy** *policy*
The password policy used by this principal. If not specified, the
policy ``default`` is used if it exists (unless **-clearpolicy**
is specified).
**-clearpolicy**
Prevents any policy from being assigned when **-policy** is not
specified.
{-\|+}\ **allow_postdated**
**-allow_postdated** prohibits this principal from obtaining
postdated tickets. **+allow_postdated** clears this flag.
{-\|+}\ **allow_forwardable**
**-allow_forwardable** prohibits this principal from obtaining
forwardable tickets. **+allow_forwardable** clears this flag.
{-\|+}\ **allow_renewable**
**-allow_renewable** prohibits this principal from obtaining
renewable tickets. **+allow_renewable** clears this flag.
{-\|+}\ **allow_proxiable**
**-allow_proxiable** prohibits this principal from obtaining
proxiable tickets. **+allow_proxiable** clears this flag.
{-\|+}\ **allow_dup_skey**
**-allow_dup_skey** disables user-to-user authentication for this
principal by prohibiting others from obtaining a service ticket
encrypted in this principal's TGT session key.
**+allow_dup_skey** clears this flag.
{-\|+}\ **requires_preauth**
**+requires_preauth** requires this principal to preauthenticate
before being allowed to kinit. **-requires_preauth** clears this
flag. When **+requires_preauth** is set on a service principal,
the KDC will only issue service tickets for that service principal
if the client's initial authentication was performed using
preauthentication.
{-\|+}\ **requires_hwauth**
**+requires_hwauth** requires this principal to preauthenticate
using a hardware device before being allowed to kinit.
**-requires_hwauth** clears this flag. When **+requires_hwauth** is
set on a service principal, the KDC will only issue service tickets
for that service principal if the client's initial authentication was
performed using a hardware device to preauthenticate.
{-\|+}\ **ok_as_delegate**
**+ok_as_delegate** sets the **okay as delegate** flag on tickets
issued with this principal as the service. Clients may use this
flag as a hint that credentials should be delegated when
authenticating to the service. **-ok_as_delegate** clears this
flag.
{-\|+}\ **allow_svr**
**-allow_svr** prohibits the issuance of service tickets for this
principal. In release 1.17 and later, user-to-user service
tickets are still allowed unless the **-allow_dup_skey** flag is
also set. **+allow_svr** clears this flag.
{-\|+}\ **allow_tgs_req**
**-allow_tgs_req** specifies that a Ticket-Granting Service (TGS)
request for a service ticket for this principal is not permitted.
**+allow_tgs_req** clears this flag.
{-\|+}\ **allow_tix**
**-allow_tix** forbids the issuance of any tickets for this
principal. **+allow_tix** clears this flag.
{-\|+}\ **needchange**
**+needchange** forces a password change on the next initial
authentication to this principal. **-needchange** clears this
flag.
{-\|+}\ **password_changing_service**
**+password_changing_service** marks this principal as a password
change service principal.
{-\|+}\ **ok_to_auth_as_delegate**
**+ok_to_auth_as_delegate** allows this principal to acquire
forwardable tickets to itself from arbitrary users, for use with
constrained delegation.
{-\|+}\ **no_auth_data_required**
**+no_auth_data_required** prevents PAC or AD-SIGNEDPATH data from
being added to service tickets for the principal.
{-\|+}\ **lockdown_keys**
**+lockdown_keys** prevents keys for this principal from leaving
the KDC via kadmind. The chpass and extract operations are denied
for a principal with this attribute. The chrand operation is
allowed, but will not return the new keys. The delete and rename
operations are also denied if this attribute is set, in order to
prevent a malicious administrator from replacing principals like
krbtgt/* or kadmin/* with new principals without the attribute.
This attribute can be set via the network protocol, but can only
be removed using kadmin.local.
**-randkey**
Sets the key of the principal to a random value.
**-nokey**
Causes the principal to be created with no key. New in release
1.12.
**-pw** *password*
Sets the password of the principal to the specified string and
does not prompt for a password. Note: using this option in a
shell script may expose the password to other users on the system
via the process list.
**-e** *enc*:*salt*,...
Uses the specified keysalt list for setting the keys of the
principal. See :ref:`Keysalt_lists` in :ref:`kdc.conf(5)` for a
list of possible values.
**-x** *db_princ_args*
Indicates database-specific options. The options for the LDAP
database module are:
**-x dn=**\ *dn*
Specifies the LDAP object that will contain the Kerberos
principal being created.
**-x linkdn=**\ *dn*
Specifies the LDAP object to which the newly created Kerberos
principal object will point.
**-x containerdn=**\ *container_dn*
Specifies the container object under which the Kerberos
principal is to be created.
**-x tktpolicy=**\ *policy*
Associates a ticket policy to the Kerberos principal.
.. note::
- The **containerdn** and **linkdn** options cannot be
specified with the **dn** option.
- If the *dn* or *containerdn* options are not specified while
adding the principal, the principals are created under the
principal container configured in the realm or the realm
container.
- *dn* and *containerdn* should be within the subtrees or
principal container configured in the realm.
Example::
kadmin: addprinc jennifer
No policy specified for "jennifer@ATHENA.MIT.EDU";
defaulting to no policy.
Enter password for principal jennifer@ATHENA.MIT.EDU:
Re-enter password for principal jennifer@ATHENA.MIT.EDU:
Principal "jennifer@ATHENA.MIT.EDU" created.
kadmin:
.. _modify_principal:
modify_principal
~~~~~~~~~~~~~~~~
**modify_principal** [*options*] *principal*
Modifies the specified principal, changing the fields as specified.
The options to **add_principal** also apply to this command, except
for the **-randkey**, **-pw**, and **-e** options. In addition, the
option **-clearpolicy** will clear the current policy of a principal.
This command requires the *modify* privilege.
Alias: **modprinc**
Options (in addition to the **addprinc** options):
**-unlock**
Unlocks a locked principal (one which has received too many failed
authentication attempts without enough time between them according
to its password policy) so that it can successfully authenticate.
.. _rename_principal:
rename_principal
~~~~~~~~~~~~~~~~
**rename_principal** [**-force**] *old_principal* *new_principal*
Renames the specified *old_principal* to *new_principal*. This
command prompts for confirmation, unless the **-force** option is
given.
This command requires the **add** and **delete** privileges.
Alias: **renprinc**
.. _delete_principal:
delete_principal
~~~~~~~~~~~~~~~~
**delete_principal** [**-force**] *principal*
Deletes the specified *principal* from the database. This command
prompts for deletion, unless the **-force** option is given.
This command requires the **delete** privilege.
Alias: **delprinc**
.. _change_password:
change_password
~~~~~~~~~~~~~~~
**change_password** [*options*] *principal*
Changes the password of *principal*. Prompts for a new password if
neither **-randkey** or **-pw** is specified.
This command requires the **changepw** privilege, or that the
principal running the program is the same as the principal being
changed.
Alias: **cpw**
The following options are available:
**-randkey**
Sets the key of the principal to a random value.
**-pw** *password*
Set the password to the specified string. Using this option in a
script may expose the password to other users on the system via
the process list.
**-e** *enc*:*salt*,...
Uses the specified keysalt list for setting the keys of the
principal. See :ref:`Keysalt_lists` in :ref:`kdc.conf(5)` for a
list of possible values.
**-keepold**
Keeps the existing keys in the database. This flag is usually not
necessary except perhaps for ``krbtgt`` principals.
Example::
kadmin: cpw systest
Enter password for principal systest@BLEEP.COM:
Re-enter password for principal systest@BLEEP.COM:
Password for systest@BLEEP.COM changed.
kadmin:
.. _purgekeys:
purgekeys
~~~~~~~~~
**purgekeys** [**-all**\|\ **-keepkvno** *oldest_kvno_to_keep*] *principal*
Purges previously retained old keys (e.g., from **change_password
-keepold**) from *principal*. If **-keepkvno** is specified, then
only purges keys with kvnos lower than *oldest_kvno_to_keep*. If
**-all** is specified, then all keys are purged. The **-all** option
is new in release 1.12.
This command requires the **modify** privilege.
.. _get_principal:
get_principal
~~~~~~~~~~~~~
**get_principal** [**-terse**] *principal*
Gets the attributes of principal. With the **-terse** option, outputs
fields as quoted tab-separated strings.
This command requires the **inquire** privilege, or that the principal
running the the program to be the same as the one being listed.
Alias: **getprinc**
Examples::
kadmin: getprinc tlyu/admin
Principal: tlyu/admin@BLEEP.COM
Expiration date: [never]
Last password change: Mon Aug 12 14:16:47 EDT 1996
Password expiration date: [never]
Maximum ticket life: 0 days 10:00:00
Maximum renewable life: 7 days 00:00:00
Last modified: Mon Aug 12 14:16:47 EDT 1996 (bjaspan/admin@BLEEP.COM)
Last successful authentication: [never]
Last failed authentication: [never]
Failed password attempts: 0
Number of keys: 1
Key: vno 1, aes256-cts-hmac-sha384-192
MKey: vno 1
Attributes:
Policy: [none]
kadmin: getprinc -terse systest
systest@BLEEP.COM 3 86400 604800 1
785926535 753241234 785900000
tlyu/admin@BLEEP.COM 786100034 0 0
kadmin:
.. _list_principals:
list_principals
~~~~~~~~~~~~~~~
**list_principals** [*expression*]
Retrieves all or some principal names. *expression* is a shell-style
glob expression that can contain the wild-card characters ``?``,
``*``, and ``[]``. All principal names matching the expression are
printed. If no expression is provided, all principal names are
printed. If the expression does not contain an ``@`` character, an
``@`` character followed by the local realm is appended to the
expression.
This command requires the **list** privilege.
Alias: **listprincs**, **get_principals**, **getprincs**
Example::
kadmin: listprincs test*
test3@SECURE-TEST.OV.COM
test2@SECURE-TEST.OV.COM
test1@SECURE-TEST.OV.COM
testuser@SECURE-TEST.OV.COM
kadmin:
.. _get_strings:
get_strings
~~~~~~~~~~~
**get_strings** *principal*
Displays string attributes on *principal*.
This command requires the **inquire** privilege.
Alias: **getstrs**
.. _set_string:
set_string
~~~~~~~~~~
**set_string** *principal* *name* *value*
Sets a string attribute on *principal*. String attributes are used to
supply per-principal configuration to the KDC and some KDC plugin
modules. The following string attribute names are recognized by the
KDC:
**require_auth**
Specifies an authentication indicator which is required to
authenticate to the principal as a service. Multiple indicators
can be specified, separated by spaces; in this case any of the
specified indicators will be accepted. (New in release 1.14.)
**session_enctypes**
Specifies the encryption types supported for session keys when the
principal is authenticated to as a server. See
:ref:`Encryption_types` in :ref:`kdc.conf(5)` for a list of the
accepted values.
**otp**
Enables One Time Passwords (OTP) preauthentication for a client
*principal*. The *value* is a JSON string representing an array
of objects, each having optional ``type`` and ``username`` fields.
**pkinit_cert_match**
Specifies a matching expression that defines the certificate
attributes required for the client certificate used by the
principal during PKINIT authentication. The matching expression
is in the same format as those used by the **pkinit_cert_match**
option in :ref:`krb5.conf(5)`. (New in release 1.16.)
**pac_privsvr_enctype**
Forces the encryption type of the PAC KDC checksum buffers to the
specified encryption type for tickets issued to this server, by
deriving a key from the local krbtgt key if it is of a different
encryption type. It may be necessary to set this value to
"aes256-sha1" on the cross-realm krbtgt entry for an Active
Directory realm when using aes-sha2 keys on the local krbtgt
entry.
This command requires the **modify** privilege.
Alias: **setstr**
Example::
set_string host/foo.mit.edu session_enctypes aes128-cts
set_string user@FOO.COM otp "[{""type"":""hotp"",""username"":""al""}]"
.. _del_string:
del_string
~~~~~~~~~~
**del_string** *principal* *key*
Deletes a string attribute from *principal*.
This command requires the **delete** privilege.
Alias: **delstr**
.. _add_policy:
add_policy
~~~~~~~~~~
**add_policy** [*options*] *policy*
Adds a password policy named *policy* to the database.
This command requires the **add** privilege.
Alias: **addpol**
The following options are available:
**-maxlife** *time*
(:ref:`duration` or :ref:`getdate` string) Sets the maximum
lifetime of a password.
**-minlife** *time*
(:ref:`duration` or :ref:`getdate` string) Sets the minimum
lifetime of a password.
**-minlength** *length*
Sets the minimum length of a password.
**-minclasses** *number*
Sets the minimum number of character classes required in a
password. The five character classes are lower case, upper case,
numbers, punctuation, and whitespace/unprintable characters.
**-history** *number*
Sets the number of past keys kept for a principal. This option is
not supported with the LDAP KDC database module.
.. _policy_maxfailure:
**-maxfailure** *maxnumber*
Sets the number of authentication failures before the principal is
locked. Authentication failures are only tracked for principals
which require preauthentication. The counter of failed attempts
resets to 0 after a successful attempt to authenticate. A
*maxnumber* value of 0 (the default) disables lockout.
.. _policy_failurecountinterval:
**-failurecountinterval** *failuretime*
(:ref:`duration` or :ref:`getdate` string) Sets the allowable time
between authentication failures. If an authentication failure
happens after *failuretime* has elapsed since the previous
failure, the number of authentication failures is reset to 1. A
*failuretime* value of 0 (the default) means forever.
.. _policy_lockoutduration:
**-lockoutduration** *lockouttime*
(:ref:`duration` or :ref:`getdate` string) Sets the duration for
which the principal is locked from authenticating if too many
authentication failures occur without the specified failure count
interval elapsing. A duration of 0 (the default) means the
principal remains locked out until it is administratively unlocked
with ``modprinc -unlock``.
**-allowedkeysalts**
Specifies the key/salt tuples supported for long-term keys when
setting or changing a principal's password/keys. See
:ref:`Keysalt_lists` in :ref:`kdc.conf(5)` for a list of the
accepted values, but note that key/salt tuples must be separated
with commas (',') only. To clear the allowed key/salt policy use
a value of '-'.
Example::
kadmin: add_policy -maxlife "2 days" -minlength 5 guests
kadmin:
.. _modify_policy:
modify_policy
~~~~~~~~~~~~~
**modify_policy** [*options*] *policy*
Modifies the password policy named *policy*. Options are as described
for **add_policy**.
This command requires the **modify** privilege.
Alias: **modpol**
.. _delete_policy:
delete_policy
~~~~~~~~~~~~~
**delete_policy** [**-force**] *policy*
Deletes the password policy named *policy*. Prompts for confirmation
before deletion. The command will fail if the policy is in use by any
principals.
This command requires the **delete** privilege.
Alias: **delpol**
Example::
kadmin: del_policy guests
Are you sure you want to delete the policy "guests"?
(yes/no): yes
kadmin:
.. _get_policy:
get_policy
~~~~~~~~~~
**get_policy** [ **-terse** ] *policy*
Displays the values of the password policy named *policy*. With the
**-terse** flag, outputs the fields as quoted strings separated by
tabs.
This command requires the **inquire** privilege.
Alias: **getpol**
Examples::
kadmin: get_policy admin
Policy: admin
Maximum password life: 180 days 00:00:00
Minimum password life: 00:00:00
Minimum password length: 6
Minimum number of password character classes: 2
Number of old keys kept: 5
Reference count: 17
kadmin: get_policy -terse admin
admin 15552000 0 6 2 5 17
kadmin:
The "Reference count" is the number of principals using that policy.
With the LDAP KDC database module, the reference count field is not
meaningful.
.. _list_policies:
list_policies
~~~~~~~~~~~~~
**list_policies** [*expression*]
Retrieves all or some policy names. *expression* is a shell-style
glob expression that can contain the wild-card characters ``?``,
``*``, and ``[]``. All policy names matching the expression are
printed. If no expression is provided, all existing policy names are
printed.
This command requires the **list** privilege.
Aliases: **listpols**, **get_policies**, **getpols**.
Examples::
kadmin: listpols
test-pol
dict-only
once-a-min
test-pol-nopw
kadmin: listpols t*
test-pol
test-pol-nopw
kadmin:
.. _ktadd:
ktadd
~~~~~
| **ktadd** [options] *principal*
| **ktadd** [options] **-glob** *princ-exp*
Adds a *principal*, or all principals matching *princ-exp*, to a
keytab file. Each principal's keys are randomized in the process.
The rules for *princ-exp* are described in the **list_principals**
command.
This command requires the **inquire** and **changepw** privileges.
With the **-glob** form, it also requires the **list** privilege.
The options are:
**-k[eytab]** *keytab*
Use *keytab* as the keytab file. Otherwise, the default keytab is
used.
**-e** *enc*:*salt*,...
Uses the specified keysalt list for setting the new keys of the
principal. See :ref:`Keysalt_lists` in :ref:`kdc.conf(5)` for a
list of possible values.
**-q**
Display less verbose information.
**-norandkey**
Do not randomize the keys. The keys and their version numbers stay
unchanged. This option cannot be specified in combination with the
**-e** option.
An entry for each of the principal's unique encryption types is added,
ignoring multiple keys with the same encryption type but different
salt types.
Alias: **xst**
Example::
kadmin: ktadd -k /tmp/foo-new-keytab host/foo.mit.edu
Entry for principal host/foo.mit.edu@ATHENA.MIT.EDU with kvno 3,
encryption type aes256-cts-hmac-sha1-96 added to keytab
FILE:/tmp/foo-new-keytab
kadmin:
.. _ktremove:
ktremove
~~~~~~~~
**ktremove** [options] *principal* [*kvno* | *all* | *old*]
Removes entries for the specified *principal* from a keytab. Requires
no permissions, since this does not require database access.
If the string "all" is specified, all entries for that principal are
removed; if the string "old" is specified, all entries for that
principal except those with the highest kvno are removed. Otherwise,
the value specified is parsed as an integer, and all entries whose
kvno match that integer are removed.
The options are:
**-k[eytab]** *keytab*
Use *keytab* as the keytab file. Otherwise, the default keytab is
used.
**-q**
Display less verbose information.
Alias: **ktrem**
Example::
kadmin: ktremove kadmin/admin all
Entry for principal kadmin/admin with kvno 3 removed from keytab
FILE:/etc/krb5.keytab
kadmin:
lock
~~~~
Lock database exclusively. Use with extreme caution! This command
only works with the DB2 KDC database module.
unlock
~~~~~~
Release the exclusive database lock.
list_requests
~~~~~~~~~~~~~
Lists available for kadmin requests.
Aliases: **lr**, **?**
quit
~~~~
Exit program. If the database was locked, the lock is released.
Aliases: **exit**, **q**
HISTORY
-------
The kadmin program was originally written by Tom Yu at MIT, as an
interface to the OpenVision Kerberos administration program.
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kpasswd(1)`, :ref:`kadmind(8)`, :ref:`kerberos(7)`

View file

@ -0,0 +1,129 @@
.. _kadmind(8):
kadmind
=======
SYNOPSIS
--------
**kadmind**
[**-x** *db_args*]
[**-r** *realm*]
[**-m**]
[**-nofork**]
[**-proponly**]
[**-port** *port-number*]
[**-P** *pid_file*]
[**-p** *kdb5_util_path*]
[**-K** *kprop_path*]
[**-k** *kprop_port*]
[**-F** *dump_file*]
DESCRIPTION
-----------
kadmind starts the Kerberos administration server. kadmind typically
runs on the primary Kerberos server, which stores the KDC database.
If the KDC database uses the LDAP module, the administration server
and the KDC server need not run on the same machine. kadmind accepts
remote requests from programs such as :ref:`kadmin(1)` and
:ref:`kpasswd(1)` to administer the information in these database.
kadmind requires a number of configuration files to be set up in order
for it to work:
:ref:`kdc.conf(5)`
The KDC configuration file contains configuration information for
the KDC and admin servers. kadmind uses settings in this file to
locate the Kerberos database, and is also affected by the
**acl_file**, **dict_file**, **kadmind_port**, and iprop-related
settings.
:ref:`kadm5.acl(5)`
kadmind's ACL (access control list) tells it which principals are
allowed to perform administration actions. The pathname to the
ACL file can be specified with the **acl_file** :ref:`kdc.conf(5)`
variable; by default, it is |kdcdir|\ ``/kadm5.acl``.
After the server begins running, it puts itself in the background and
disassociates itself from its controlling terminal.
kadmind can be configured for incremental database propagation.
Incremental propagation allows replica KDC servers to receive
principal and policy updates incrementally instead of receiving full
dumps of the database. This facility can be enabled in the
:ref:`kdc.conf(5)` file with the **iprop_enable** option. Incremental
propagation requires the principal ``kiprop/PRIMARY\@REALM`` (where
PRIMARY is the primary KDC's canonical host name, and REALM the realm
name). In release 1.13, this principal is automatically created and
registered into the datebase.
OPTIONS
-------
**-r** *realm*
specifies the realm that kadmind will serve; if it is not
specified, the default realm of the host is used.
**-m**
causes the master database password to be fetched from the
keyboard (before the server puts itself in the background, if not
invoked with the **-nofork** option) rather than from a file on
disk.
**-nofork**
causes the server to remain in the foreground and remain
associated to the terminal.
**-proponly**
causes the server to only listen and respond to Kerberos replica
incremental propagation polling requests. This option can be used
to set up a hierarchical propagation topology where a replica KDC
provides incremental updates to other Kerberos replicas.
**-port** *port-number*
specifies the port on which the administration server listens for
connections. The default port is determined by the
**kadmind_port** configuration variable in :ref:`kdc.conf(5)`.
**-P** *pid_file*
specifies the file to which the PID of kadmind process should be
written after it starts up. This file can be used to identify
whether kadmind is still running and to allow init scripts to stop
the correct process.
**-p** *kdb5_util_path*
specifies the path to the kdb5_util command to use when dumping the
KDB in response to full resync requests when iprop is enabled.
**-K** *kprop_path*
specifies the path to the kprop command to use to send full dumps
to replicas in response to full resync requests.
**-k** *kprop_port*
specifies the port by which the kprop process that is spawned by
kadmind connects to the replica kpropd, in order to transfer the
dump file during an iprop full resync request.
**-F** *dump_file*
specifies the file path to be used for dumping the KDB in response
to full resync requests when iprop is enabled.
**-x** *db_args*
specifies database-specific arguments. See :ref:`Database Options
<dboptions>` in :ref:`kadmin(1)` for supported arguments.
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kpasswd(1)`, :ref:`kadmin(1)`, :ref:`kdb5_util(8)`,
:ref:`kdb5_ldap_util(8)`, :ref:`kadm5.acl(5)`, :ref:`kerberos(7)`

View file

@ -0,0 +1,449 @@
.. _kdb5_ldap_util(8):
kdb5_ldap_util
===============
SYNOPSIS
--------
.. _kdb5_ldap_util_synopsis:
**kdb5_ldap_util**
[**-D** *user_dn* [**-w** *passwd*]]
[**-H** *ldapuri*]
**command**
[*command_options*]
.. _kdb5_ldap_util_synopsis_end:
DESCRIPTION
-----------
kdb5_ldap_util allows an administrator to manage realms, Kerberos
services and ticket policies.
COMMAND-LINE OPTIONS
--------------------
.. _kdb5_ldap_util_options:
**-r** *realm*
Specifies the realm to be operated on.
**-D** *user_dn*
Specifies the Distinguished Name (DN) of the user who has
sufficient rights to perform the operation on the LDAP server.
**-w** *passwd*
Specifies the password of *user_dn*. This option is not
recommended.
**-H** *ldapuri*
Specifies the URI of the LDAP server.
By default, kdb5_ldap_util operates on the default realm (as specified
in :ref:`krb5.conf(5)`) and connects and authenticates to the LDAP
server in the same manner as :ref:kadmind(8)` would given the
parameters in :ref:`dbdefaults` in :ref:`kdc.conf(5)`.
.. _kdb5_ldap_util_options_end:
COMMANDS
--------
create
~~~~~~
.. _kdb5_ldap_util_create:
**create**
[**-subtrees** *subtree_dn_list*]
[**-sscope** *search_scope*]
[**-containerref** *container_reference_dn*]
[**-k** *mkeytype*]
[**-kv** *mkeyVNO*]
[**-M** *mkeyname*]
[**-m|-P** *password*\|\ **-sf** *stashfilename*]
[**-s**]
[**-maxtktlife** *max_ticket_life*]
[**-maxrenewlife** *max_renewable_ticket_life*]
[*ticket_flags*]
Creates realm in directory. Options:
**-subtrees** *subtree_dn_list*
Specifies the list of subtrees containing the principals of a
realm. The list contains the DNs of the subtree objects separated
by colon (``:``).
**-sscope** *search_scope*
Specifies the scope for searching the principals under the
subtree. The possible values are 1 or one (one level), 2 or sub
(subtrees).
**-containerref** *container_reference_dn*
Specifies the DN of the container object in which the principals
of a realm will be created. If the container reference is not
configured for a realm, the principals will be created in the
realm container.
**-k** *mkeytype*
Specifies the key type of the master key in the database. The
default is given by the **master_key_type** variable in
:ref:`kdc.conf(5)`.
**-kv** *mkeyVNO*
Specifies the version number of the master key in the database;
the default is 1. Note that 0 is not allowed.
**-M** *mkeyname*
Specifies the principal name for the master key in the database.
If not specified, the name is determined by the
**master_key_name** variable in :ref:`kdc.conf(5)`.
**-m**
Specifies that the master database password should be read from
the TTY rather than fetched from a file on the disk.
**-P** *password*
Specifies the master database password. This option is not
recommended.
**-sf** *stashfilename*
Specifies the stash file of the master database password.
**-s**
Specifies that the stash file is to be created.
**-maxtktlife** *max_ticket_life*
(:ref:`getdate` string) Specifies maximum ticket life for
principals in this realm.
**-maxrenewlife** *max_renewable_ticket_life*
(:ref:`getdate` string) Specifies maximum renewable life of
tickets for principals in this realm.
*ticket_flags*
Specifies global ticket flags for the realm. Allowable flags are
documented in the description of the **add_principal** command in
:ref:`kadmin(1)`.
Example::
kdb5_ldap_util -D cn=admin,o=org -H ldaps://ldap-server1.mit.edu
-r ATHENA.MIT.EDU create -subtrees o=org -sscope SUB
Password for "cn=admin,o=org":
Initializing database for realm 'ATHENA.MIT.EDU'
You will be prompted for the database Master Password.
It is important that you NOT FORGET this password.
Enter KDC database master key:
Re-enter KDC database master key to verify:
.. _kdb5_ldap_util_create_end:
modify
~~~~~~
.. _kdb5_ldap_util_modify:
**modify**
[**-subtrees** *subtree_dn_list*]
[**-sscope** *search_scope*]
[**-containerref** *container_reference_dn*]
[**-maxtktlife** *max_ticket_life*]
[**-maxrenewlife** *max_renewable_ticket_life*]
[*ticket_flags*]
Modifies the attributes of a realm. Options:
**-subtrees** *subtree_dn_list*
Specifies the list of subtrees containing the principals of a
realm. The list contains the DNs of the subtree objects separated
by colon (``:``). This list replaces the existing list.
**-sscope** *search_scope*
Specifies the scope for searching the principals under the
subtrees. The possible values are 1 or one (one level), 2 or sub
(subtrees).
**-containerref** *container_reference_dn* Specifies the DN of the
container object in which the principals of a realm will be
created.
**-maxtktlife** *max_ticket_life*
(:ref:`getdate` string) Specifies maximum ticket life for
principals in this realm.
**-maxrenewlife** *max_renewable_ticket_life*
(:ref:`getdate` string) Specifies maximum renewable life of
tickets for principals in this realm.
*ticket_flags*
Specifies global ticket flags for the realm. Allowable flags are
documented in the description of the **add_principal** command in
:ref:`kadmin(1)`.
Example::
shell% kdb5_ldap_util -r ATHENA.MIT.EDU -D cn=admin,o=org -H
ldaps://ldap-server1.mit.edu modify +requires_preauth
Password for "cn=admin,o=org":
shell%
.. _kdb5_ldap_util_modify_end:
view
~~~~
.. _kdb5_ldap_util_view:
**view**
Displays the attributes of a realm.
Example::
kdb5_ldap_util -D cn=admin,o=org -H ldaps://ldap-server1.mit.edu
-r ATHENA.MIT.EDU view
Password for "cn=admin,o=org":
Realm Name: ATHENA.MIT.EDU
Subtree: ou=users,o=org
Subtree: ou=servers,o=org
SearchScope: ONE
Maximum ticket life: 0 days 01:00:00
Maximum renewable life: 0 days 10:00:00
Ticket flags: DISALLOW_FORWARDABLE REQUIRES_PWCHANGE
.. _kdb5_ldap_util_view_end:
destroy
~~~~~~~
.. _kdb5_ldap_util_destroy:
**destroy** [**-f**]
Destroys an existing realm. Options:
**-f**
If specified, will not prompt the user for confirmation.
Example::
shell% kdb5_ldap_util -r ATHENA.MIT.EDU -D cn=admin,o=org -H
ldaps://ldap-server1.mit.edu destroy
Password for "cn=admin,o=org":
Deleting KDC database of 'ATHENA.MIT.EDU', are you sure?
(type 'yes' to confirm)? yes
OK, deleting database of 'ATHENA.MIT.EDU'...
shell%
.. _kdb5_ldap_util_destroy_end:
list
~~~~
.. _kdb5_ldap_util_list:
**list**
Lists the names of realms under the container.
Example::
shell% kdb5_ldap_util -D cn=admin,o=org -H
ldaps://ldap-server1.mit.edu list
Password for "cn=admin,o=org":
ATHENA.MIT.EDU
OPENLDAP.MIT.EDU
MEDIA-LAB.MIT.EDU
shell%
.. _kdb5_ldap_util_list_end:
stashsrvpw
~~~~~~~~~~
.. _kdb5_ldap_util_stashsrvpw:
**stashsrvpw**
[**-f** *filename*]
*name*
Allows an administrator to store the password for service object in a
file so that KDC and Administration server can use it to authenticate
to the LDAP server. Options:
**-f** *filename*
Specifies the complete path of the service password file. By
default, ``/usr/local/var/service_passwd`` is used.
*name*
Specifies the name of the object whose password is to be stored.
If :ref:`krb5kdc(8)` or :ref:`kadmind(8)` are configured for
simple binding, this should be the distinguished name it will
use as given by the **ldap_kdc_dn** or **ldap_kadmind_dn**
variable in :ref:`kdc.conf(5)`. If the KDC or kadmind is
configured for SASL binding, this should be the authentication
name it will use as given by the **ldap_kdc_sasl_authcid** or
**ldap_kadmind_sasl_authcid** variable.
Example::
kdb5_ldap_util stashsrvpw -f /home/andrew/conf_keyfile
cn=service-kdc,o=org
Password for "cn=service-kdc,o=org":
Re-enter password for "cn=service-kdc,o=org":
.. _kdb5_ldap_util_stashsrvpw_end:
create_policy
~~~~~~~~~~~~~
.. _kdb5_ldap_util_create_policy:
**create_policy**
[**-maxtktlife** *max_ticket_life*]
[**-maxrenewlife** *max_renewable_ticket_life*]
[*ticket_flags*]
*policy_name*
Creates a ticket policy in the directory. Options:
**-maxtktlife** *max_ticket_life*
(:ref:`getdate` string) Specifies maximum ticket life for
principals.
**-maxrenewlife** *max_renewable_ticket_life*
(:ref:`getdate` string) Specifies maximum renewable life of
tickets for principals.
*ticket_flags*
Specifies the ticket flags. If this option is not specified, by
default, no restriction will be set by the policy. Allowable
flags are documented in the description of the **add_principal**
command in :ref:`kadmin(1)`.
*policy_name*
Specifies the name of the ticket policy.
Example::
kdb5_ldap_util -D cn=admin,o=org -H ldaps://ldap-server1.mit.edu
-r ATHENA.MIT.EDU create_policy -maxtktlife "1 day"
-maxrenewlife "1 week" -allow_postdated +needchange
-allow_forwardable tktpolicy
Password for "cn=admin,o=org":
.. _kdb5_ldap_util_create_policy_end:
modify_policy
~~~~~~~~~~~~~
.. _kdb5_ldap_util_modify_policy:
**modify_policy**
[**-maxtktlife** *max_ticket_life*]
[**-maxrenewlife** *max_renewable_ticket_life*]
[*ticket_flags*]
*policy_name*
Modifies the attributes of a ticket policy. Options are same as for
**create_policy**.
Example::
kdb5_ldap_util -D cn=admin,o=org -H
ldaps://ldap-server1.mit.edu -r ATHENA.MIT.EDU modify_policy
-maxtktlife "60 minutes" -maxrenewlife "10 hours"
+allow_postdated -requires_preauth tktpolicy
Password for "cn=admin,o=org":
.. _kdb5_ldap_util_modify_policy_end:
view_policy
~~~~~~~~~~~
.. _kdb5_ldap_util_view_policy:
**view_policy**
*policy_name*
Displays the attributes of the named ticket policy.
Example::
kdb5_ldap_util -D cn=admin,o=org -H ldaps://ldap-server1.mit.edu
-r ATHENA.MIT.EDU view_policy tktpolicy
Password for "cn=admin,o=org":
Ticket policy: tktpolicy
Maximum ticket life: 0 days 01:00:00
Maximum renewable life: 0 days 10:00:00
Ticket flags: DISALLOW_FORWARDABLE REQUIRES_PWCHANGE
.. _kdb5_ldap_util_view_policy_end:
destroy_policy
~~~~~~~~~~~~~~
.. _kdb5_ldap_util_destroy_policy:
**destroy_policy**
[**-force**]
*policy_name*
Destroys an existing ticket policy. Options:
**-force**
Forces the deletion of the policy object. If not specified, the
user will be prompted for confirmation before deleting the policy.
*policy_name*
Specifies the name of the ticket policy.
Example::
kdb5_ldap_util -D cn=admin,o=org -H ldaps://ldap-server1.mit.edu
-r ATHENA.MIT.EDU destroy_policy tktpolicy
Password for "cn=admin,o=org":
This will delete the policy object 'tktpolicy', are you sure?
(type 'yes' to confirm)? yes
** policy object 'tktpolicy' deleted.
.. _kdb5_ldap_util_destroy_policy_end:
list_policy
~~~~~~~~~~~
.. _kdb5_ldap_util_list_policy:
**list_policy**
Lists ticket policies.
Example::
kdb5_ldap_util -D cn=admin,o=org -H ldaps://ldap-server1.mit.edu
-r ATHENA.MIT.EDU list_policy
Password for "cn=admin,o=org":
tktpolicy
tmppolicy
userpolicy
.. _kdb5_ldap_util_list_policy_end:
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kadmin(1)`, :ref:`kerberos(7)`

View file

@ -0,0 +1,502 @@
.. _kdb5_util(8):
kdb5_util
=========
SYNOPSIS
--------
.. _kdb5_util_synopsis:
**kdb5_util**
[**-r** *realm*]
[**-d** *dbname*]
[**-k** *mkeytype*]
[**-kv** *mkeyVNO*]
[**-M** *mkeyname*]
[**-m**]
[**-sf** *stashfilename*]
[**-P** *password*]
[**-x** *db_args*]
*command* [*command_options*]
.. _kdb5_util_synopsis_end:
DESCRIPTION
-----------
kdb5_util allows an administrator to perform maintenance procedures on
the KDC database. Databases can be created, destroyed, and dumped to
or loaded from ASCII files. kdb5_util can create a Kerberos master
key stash file or perform live rollover of the master key.
When kdb5_util is run, it attempts to acquire the master key and open
the database. However, execution continues regardless of whether or
not kdb5_util successfully opens the database, because the database
may not exist yet or the stash file may be corrupt.
Note that some KDC database modules may not support all kdb5_util
commands.
COMMAND-LINE OPTIONS
--------------------
.. _kdb5_util_options:
**-r** *realm*
specifies the Kerberos realm of the database.
**-d** *dbname*
specifies the name under which the principal database is stored;
by default the database is that listed in :ref:`kdc.conf(5)`. The
password policy database and lock files are also derived from this
value.
**-k** *mkeytype*
specifies the key type of the master key in the database. The
default is given by the **master_key_type** variable in
:ref:`kdc.conf(5)`.
**-kv** *mkeyVNO*
Specifies the version number of the master key in the database;
the default is 1. Note that 0 is not allowed.
**-M** *mkeyname*
principal name for the master key in the database. If not
specified, the name is determined by the **master_key_name**
variable in :ref:`kdc.conf(5)`.
**-m**
specifies that the master database password should be read from
the keyboard rather than fetched from a file on disk.
**-sf** *stash_file*
specifies the stash filename of the master database password. If
not specified, the filename is determined by the
**key_stash_file** variable in :ref:`kdc.conf(5)`.
**-P** *password*
specifies the master database password. Using this option may
expose the password to other users on the system via the process
list.
**-x** *db_args*
specifies database-specific options. See :ref:`kadmin(1)` for
supported options.
.. _kdb5_util_options_end:
COMMANDS
--------
create
~~~~~~
.. _kdb5_util_create:
**create** [**-s**]
Creates a new database. If the **-s** option is specified, the stash
file is also created. This command fails if the database already
exists. If the command is successful, the database is opened just as
if it had already existed when the program was first run.
.. _kdb5_util_create_end:
destroy
~~~~~~~
.. _kdb5_util_destroy:
**destroy** [**-f**]
Destroys the database, first overwriting the disk sectors and then
unlinking the files, after prompting the user for confirmation. With
the **-f** argument, does not prompt the user.
.. _kdb5_util_destroy_end:
stash
~~~~~
.. _kdb5_util_stash:
**stash** [**-f** *keyfile*]
Stores the master principal's keys in a stash file. The **-f**
argument can be used to override the *keyfile* specified in
:ref:`kdc.conf(5)`.
.. _kdb5_util_stash_end:
dump
~~~~
.. _kdb5_util_dump:
**dump** [**-b7**\|\ **-r13**\|\ **-r18**]
[**-verbose**] [**-mkey_convert**] [**-new_mkey_file**
*mkey_file*] [**-rev**] [**-recurse**] [*filename*
[*principals*...]]
Dumps the current Kerberos and KADM5 database into an ASCII file. By
default, the database is dumped in current format, "kdb5_util
load_dump version 7". If filename is not specified, or is the string
"-", the dump is sent to standard output. Options:
**-b7**
causes the dump to be in the Kerberos 5 Beta 7 format ("kdb5_util
load_dump version 4"). This was the dump format produced on
releases prior to 1.2.2.
**-r13**
causes the dump to be in the Kerberos 5 1.3 format ("kdb5_util
load_dump version 5"). This was the dump format produced on
releases prior to 1.8.
**-r18**
causes the dump to be in the Kerberos 5 1.8 format ("kdb5_util
load_dump version 6"). This was the dump format produced on
releases prior to 1.11.
**-verbose**
causes the name of each principal and policy to be printed as it
is dumped.
**-mkey_convert**
prompts for a new master key. This new master key will be used to
re-encrypt principal key data in the dumpfile. The principal keys
themselves will not be changed.
**-new_mkey_file** *mkey_file*
the filename of a stash file. The master key in this stash file
will be used to re-encrypt the key data in the dumpfile. The key
data in the database will not be changed.
**-rev**
dumps in reverse order. This may recover principals that do not
dump normally, in cases where database corruption has occurred.
**-recurse**
causes the dump to walk the database recursively (btree only).
This may recover principals that do not dump normally, in cases
where database corruption has occurred. In cases of such
corruption, this option will probably retrieve more principals
than the **-rev** option will.
.. versionchanged:: 1.15
Release 1.15 restored the functionality of the **-recurse**
option.
.. versionchanged:: 1.5
The **-recurse** option ceased working until release 1.15,
doing a normal dump instead of a recursive traversal.
.. _kdb5_util_dump_end:
load
~~~~
.. _kdb5_util_load:
**load** [**-b7**\|\ **-r13**\|\ **-r18**] [**-hash**]
[**-verbose**] [**-update**] *filename*
Loads a database dump from the named file into the named database. If
no option is given to determine the format of the dump file, the
format is detected automatically and handled as appropriate. Unless
the **-update** option is given, **load** creates a new database
containing only the data in the dump file, overwriting the contents of
any previously existing database. Note that when using the LDAP KDC
database module, the **-update** flag is required.
Options:
**-b7**
requires the database to be in the Kerberos 5 Beta 7 format
("kdb5_util load_dump version 4"). This was the dump format
produced on releases prior to 1.2.2.
**-r13**
requires the database to be in Kerberos 5 1.3 format ("kdb5_util
load_dump version 5"). This was the dump format produced on
releases prior to 1.8.
**-r18**
requires the database to be in Kerberos 5 1.8 format ("kdb5_util
load_dump version 6"). This was the dump format produced on
releases prior to 1.11.
**-hash**
stores the database in hash format, if using the DB2 database
type. If this option is not specified, the database will be
stored in btree format. This option is not recommended, as
databases stored in hash format are known to corrupt data and lose
principals.
**-verbose**
causes the name of each principal and policy to be printed as it
is dumped.
**-update**
records from the dump file are added to or updated in the existing
database. Otherwise, a new database is created containing only
what is in the dump file and the old one destroyed upon successful
completion.
.. _kdb5_util_load_end:
ark
~~~
**ark** [**-e** *enc*:*salt*,...] *principal*
Adds new random keys to *principal* at the next available key version
number. Keys for the current highest key version number will be
preserved. The **-e** option specifies the list of encryption and
salt types to be used for the new keys.
add_mkey
~~~~~~~~
**add_mkey** [**-e** *etype*] [**-s**]
Adds a new master key to the master key principal, but does not mark
it as active. Existing master keys will remain. The **-e** option
specifies the encryption type of the new master key; see
:ref:`Encryption_types` in :ref:`kdc.conf(5)` for a list of possible
values. The **-s** option stashes the new master key in the stash
file, which will be created if it doesn't already exist.
After a new master key is added, it should be propagated to replica
servers via a manual or periodic invocation of :ref:`kprop(8)`. Then,
the stash files on the replica servers should be updated with the
kdb5_util **stash** command. Once those steps are complete, the key
is ready to be marked active with the kdb5_util **use_mkey** command.
use_mkey
~~~~~~~~
**use_mkey** *mkeyVNO* [*time*]
Sets the activation time of the master key specified by *mkeyVNO*.
Once a master key becomes active, it will be used to encrypt newly
created principal keys. If no *time* argument is given, the current
time is used, causing the specified master key version to become
active immediately. The format for *time* is :ref:`getdate` string.
After a new master key becomes active, the kdb5_util
**update_princ_encryption** command can be used to update all
principal keys to be encrypted in the new master key.
list_mkeys
~~~~~~~~~~
**list_mkeys**
List all master keys, from most recent to earliest, in the master key
principal. The output will show the kvno, enctype, and salt type for
each mkey, similar to the output of :ref:`kadmin(1)` **getprinc**. A
``*`` following an mkey denotes the currently active master key.
purge_mkeys
~~~~~~~~~~~
**purge_mkeys** [**-f**] [**-n**] [**-v**]
Delete master keys from the master key principal that are not used to
protect any principals. This command can be used to remove old master
keys all principal keys are protected by a newer master key.
**-f**
does not prompt for confirmation.
**-n**
performs a dry run, showing master keys that would be purged, but
not actually purging any keys.
**-v**
gives more verbose output.
update_princ_encryption
~~~~~~~~~~~~~~~~~~~~~~~
**update_princ_encryption** [**-f**] [**-n**] [**-v**]
[*princ-pattern*]
Update all principal records (or only those matching the
*princ-pattern* glob pattern) to re-encrypt the key data using the
active database master key, if they are encrypted using a different
version, and give a count at the end of the number of principals
updated. If the **-f** option is not given, ask for confirmation
before starting to make changes. The **-v** option causes each
principal processed to be listed, with an indication as to whether it
needed updating or not. The **-n** option performs a dry run, only
showing the actions which would have been taken.
tabdump
~~~~~~~
**tabdump** [**-H**] [**-c**] [**-e**] [**-n**] [**-o** *outfile*]
*dumptype*
Dump selected fields of the database in a tabular format suitable for
reporting (e.g., using traditional Unix text processing tools) or
importing into relational databases. The data format is tab-separated
(default), or optionally comma-separated (CSV), with a fixed number of
columns. The output begins with a header line containing field names,
unless suppression is requested using the **-H** option.
The *dumptype* parameter specifies the name of an output table (see
below).
Options:
**-H**
suppress writing the field names in a header line
**-c**
use comma separated values (CSV) format, with minimal quoting,
instead of the default tab-separated (unquoted, unescaped) format
**-e**
write empty hexadecimal string fields as empty fields instead of
as "-1".
**-n**
produce numeric output for fields that normally have symbolic
output, such as enctypes and flag names. Also requests output of
time stamps as decimal POSIX time_t values.
**-o** *outfile*
write the dump to the specified output file instead of to standard
output
Dump types:
**keydata**
principal encryption key information, including actual key data
(which is still encrypted in the master key)
**name**
principal name
**keyindex**
index of this key in the principal's key list
**kvno**
key version number
**enctype**
encryption type
**key**
key data as a hexadecimal string
**salttype**
salt type
**salt**
salt data as a hexadecimal string
**keyinfo**
principal encryption key information (as in **keydata** above),
excluding actual key data
**princ_flags**
principal boolean attributes. Flag names print as hexadecimal
numbers if the **-n** option is specified, and all flag positions
are printed regardless of whether or not they are set. If **-n**
is not specified, print all known flag names for each principal,
but only print hexadecimal flag names if the corresponding flag is
set.
**name**
principal name
**flag**
flag name
**value**
boolean value (0 for clear, or 1 for set)
**princ_lockout**
state information used for tracking repeated password failures
**name**
principal name
**last_success**
time stamp of most recent successful authentication
**last_failed**
time stamp of most recent failed authentication
**fail_count**
count of failed attempts
**princ_meta**
principal metadata
**name**
principal name
**modby**
name of last principal to modify this principal
**modtime**
timestamp of last modification
**lastpwd**
timestamp of last password change
**policy**
policy object name
**mkvno**
key version number of the master key that encrypts this
principal's key data
**hist_kvno**
key version number of the history key that encrypts the key
history data for this principal
**princ_stringattrs**
string attributes (key/value pairs)
**name**
principal name
**key**
attribute name
**value**
attribute value
**princ_tktpolicy**
per-principal ticket policy data, including maximum ticket
lifetimes
**name**
principal name
**expiration**
principal expiration date
**pw_expiration**
password expiration date
**max_life**
maximum ticket lifetime
**max_renew_life**
maximum renewable ticket lifetime
Examples::
$ kdb5_util tabdump -o keyinfo.txt keyinfo
$ cat keyinfo.txt
name keyindex kvno enctype salttype salt
K/M@EXAMPLE.COM 0 1 aes256-cts-hmac-sha384-192 normal -1
foo@EXAMPLE.COM 0 1 aes128-cts-hmac-sha1-96 normal -1
bar@EXAMPLE.COM 0 1 aes128-cts-hmac-sha1-96 normal -1
$ sqlite3
sqlite> .mode tabs
sqlite> .import keyinfo.txt keyinfo
sqlite> select * from keyinfo where enctype like 'aes256-%';
K/M@EXAMPLE.COM 1 1 aes256-cts-hmac-sha384-192 normal -1
sqlite> .quit
$ awk -F'\t' '$4 ~ /aes256-/ { print }' keyinfo.txt
K/M@EXAMPLE.COM 1 1 aes256-cts-hmac-sha384-192 normal -1
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kadmin(1)`, :ref:`kerberos(7)`

View file

@ -0,0 +1,60 @@
.. _kprop(8):
kprop
=====
SYNOPSIS
--------
**kprop**
[**-r** *realm*]
[**-f** *file*]
[**-d**]
[**-P** *port*]
[**-s** *keytab*]
*replica_host*
DESCRIPTION
-----------
kprop is used to securely propagate a Kerberos V5 database dump file
from the primary Kerberos server to a replica Kerberos server, which is
specified by *replica_host*. The dump file must be created by
:ref:`kdb5_util(8)`.
OPTIONS
-------
**-r** *realm*
Specifies the realm of the primary server.
**-f** *file*
Specifies the filename where the dumped principal database file is
to be found; by default the dumped database file is normally
|kdcdir|\ ``/replica_datatrans``.
**-P** *port*
Specifies the port to use to contact the :ref:`kpropd(8)` server
on the remote host.
**-d**
Prints debugging information.
**-s** *keytab*
Specifies the location of the keytab file.
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kpropd(8)`, :ref:`kdb5_util(8)`, :ref:`krb5kdc(8)`,
:ref:`kerberos(7)`

View file

@ -0,0 +1,144 @@
.. _kpropd(8):
kpropd
======
SYNOPSIS
--------
**kpropd**
[**-r** *realm*]
[**-A** *admin_server*]
[**-a** *acl_file*]
[**-f** *replica_dumpfile*]
[**-F** *principal_database*]
[**-p** *kdb5_util_prog*]
[**-P** *port*]
[**--pid-file**\ =\ *pid_file*]
[**-D**]
[**-d**]
[**-s** *keytab_file*]
DESCRIPTION
-----------
The *kpropd* command runs on the replica KDC server. It listens for
update requests made by the :ref:`kprop(8)` program. If incremental
propagation is enabled, it periodically requests incremental updates
from the primary KDC.
When the replica receives a kprop request from the primary, kpropd
accepts the dumped KDC database and places it in a file, and then runs
:ref:`kdb5_util(8)` to load the dumped database into the active
database which is used by :ref:`krb5kdc(8)`. This allows the primary
Kerberos server to use :ref:`kprop(8)` to propagate its database to
the replica servers. Upon a successful download of the KDC database
file, the replica Kerberos server will have an up-to-date KDC
database.
Where incremental propagation is not used, kpropd is commonly invoked
out of inetd(8) as a nowait service. This is done by adding a line to
the ``/etc/inetd.conf`` file which looks like this::
kprop stream tcp nowait root /usr/local/sbin/kpropd kpropd
kpropd can also run as a standalone daemon, backgrounding itself and
waiting for connections on port 754 (or the port specified with the
**-P** option if given). Standalone mode is required for incremental
propagation. Starting in release 1.11, kpropd automatically detects
whether it was run from inetd and runs in standalone mode if it is
not. Prior to release 1.11, the **-S** option is required to run
kpropd in standalone mode; this option is now accepted for backward
compatibility but does nothing.
Incremental propagation may be enabled with the **iprop_enable**
variable in :ref:`kdc.conf(5)`. If incremental propagation is
enabled, the replica periodically polls the primary KDC for updates, at
an interval determined by the **iprop_replica_poll** variable. If the
replica receives updates, kpropd updates its log file with any updates
from the primary. :ref:`kproplog(8)` can be used to view a summary of
the update entry log on the replica KDC. If incremental propagation
is enabled, the principal ``kiprop/replicahostname@REALM`` (where
*replicahostname* is the name of the replica KDC host, and *REALM* is
the name of the Kerberos realm) must be present in the replica's
keytab file.
:ref:`kproplog(8)` can be used to force full replication when iprop is
enabled.
OPTIONS
--------
**-r** *realm*
Specifies the realm of the primary server.
**-A** *admin_server*
Specifies the server to be contacted for incremental updates; by
default, the primary admin server is contacted.
**-f** *file*
Specifies the filename where the dumped principal database file is
to be stored; by default the dumped database file is |kdcdir|\
``/from_master``.
**-F** *kerberos_db*
Path to the Kerberos database file, if not the default.
**-p**
Allows the user to specify the pathname to the :ref:`kdb5_util(8)`
program; by default the pathname used is |sbindir|\
``/kdb5_util``.
**-D**
In this mode, kpropd will not detach itself from the current job
and run in the background. Instead, it will run in the
foreground.
**-d**
Turn on debug mode. kpropd will print out debugging messages
during the database propogation and will run in the foreground
(implies **-D**).
**-P**
Allow for an alternate port number for kpropd to listen on. This
is only useful in combination with the **-S** option.
**-a** *acl_file*
Allows the user to specify the path to the kpropd.acl file; by
default the path used is |kdcdir|\ ``/kpropd.acl``.
**--pid-file**\ =\ *pid_file*
In standalone mode, write the process ID of the daemon into
*pid_file*.
**-s** *keytab_file*
Path to a keytab to use for acquiring acceptor credentials.
**-x** *db_args*
Database-specific arguments. See :ref:`Database Options
<dboptions>` in :ref:`kadmin(1)` for supported arguments.
FILES
-----
kpropd.acl
Access file for kpropd; the default location is
``/usr/local/var/krb5kdc/kpropd.acl``. Each entry is a line
containing the principal of a host from which the local machine
will allow Kerberos database propagation via :ref:`kprop(8)`.
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kprop(8)`, :ref:`kdb5_util(8)`, :ref:`krb5kdc(8)`,
:ref:`kerberos(7)`, inetd(8)

View file

@ -0,0 +1,85 @@
.. _kproplog(8):
kproplog
========
SYNOPSIS
--------
**kproplog** [**-h**] [**-e** *num*] [-v]
**kproplog** [-R]
DESCRIPTION
-----------
The kproplog command displays the contents of the KDC database update
log to standard output. It can be used to keep track of incremental
updates to the principal database. The update log file contains the
update log maintained by the :ref:`kadmind(8)` process on the primary
KDC server and the :ref:`kpropd(8)` process on the replica KDC
servers. When updates occur, they are logged to this file.
Subsequently any KDC replica configured for incremental updates will
request the current data from the primary KDC and update their log
file with any updates returned.
The kproplog command requires read access to the update log file. It
will display update entries only for the KDC it runs on.
If no options are specified, kproplog displays a summary of the update
log. If invoked on the primary, kproplog also displays all of the
update entries. If invoked on a replica KDC server, kproplog displays
only a summary of the updates, which includes the serial number of the
last update received and the associated time stamp of the last update.
OPTIONS
-------
**-R**
Reset the update log. This forces full resynchronization. If
used on a replica then that replica will request a full resync.
If used on the primary then all replicas will request full
resyncs.
**-h**
Display a summary of the update log. This information includes
the database version number, state of the database, the number of
updates in the log, the time stamp of the first and last update,
and the version number of the first and last update entry.
**-e** *num*
Display the last *num* update entries in the log. This is useful
when debugging synchronization between KDC servers.
**-v**
Display individual attributes per update. An example of the
output generated for one entry::
Update Entry
Update serial # : 4
Update operation : Add
Update principal : test@EXAMPLE.COM
Update size : 424
Update committed : True
Update time stamp : Fri Feb 20 23:37:42 2004
Attributes changed : 6
Principal
Key data
Password last changed
Modifying principal
Modification time
TL data
ENVIRONMENT
-----------
See :ref:`kerberos(7)` for a description of Kerberos environment
variables.
SEE ALSO
--------
:ref:`kpropd(8)`, :ref:`kerberos(7)`

Some files were not shown because too many files have changed in this diff Show more