mirror of
https://sourceware.org/git/glibc.git
synced 2024-11-24 02:03:35 +08:00
fed8f7f7ae
1998-06-15 18:07 Ulrich Drepper <drepper@cygnus.com> * iconvdata/ksc5601.h: Clean up a bit. 1998-06-15 14:35 Ulrich Drepper <drepper@cygnus.com> * sunrpc/svc_tcp.c (readtcp): Also listen to all the other connections and process incoming data. * sunrpc/xdr_rec.c (set_input_fragment): Only recognize zero length fragments as invalid. Patches by Thorsten Kukuk <kukuk@weber.uni-paderborn.de>. 1998-06-13 Andreas Jaeger <aj@arthur.rhein-neckar.de> Removal of duplicates like "the the". * manual/time.texi (TZ Variable): Likewise. * manual/users.texi (Manipulating the Database): Likewise. * manual/sysinfo.texi (Filesystem handling): Likewise. * manual/signal.texi (Signaling Yourself): Likewise. * manual/message.texi (Message Translation): Likewise. (Common Usage): Likewise. * manual/llio.texi: Likewise. * manual/arith.texi (Old-style number conversion): Likewise. * manual/argp.texi (Argp): Likewise. (Argp Parsers): Likewise. (Argp Help Filtering): Likewise. * hesiod/README.hesiod: Correct typo. * manual/llio.texi: Correct typos. * manual/sysinfo.texi: Likewise. 1998-06-13 Andreas Jaeger <aj@arthur.rhein-neckar.de> * rt/aio.h: Rename __op to __operation for g++. 1998-06-13 20:17 -0400 Zack Weinberg <zack@rabi.phys.columbia.edu> * iconvdata/gen-8bit-gap.sh: Use awk not perl. Clean up sed script. * iconvdata/gen-8bit-gap-1.sh: Likewise. * iconvdata/gen-8bit.sh: Clean up sed script. * iconvdata/Makefile: Remove all PERL conditionals; build all the modules all the time. Export AWK. (gen-8bit-gap): Delete setting of PERL. (gen-8bit-gap-1): Likewise. (perl-generated-headers): Renamed awk-generated-headers. * iconvdata/gap.pl: Removed. * iconvdata/gaptab.pl: Removed. * iconvdata/gap.awk: New file. * iconvdata/gaptab.awk: New file. 1998-06-15 Ulrich Drepper <drepper@cygnus.com> * sysdeps/generic/Dist: Add errno-loc.c. * sysdeps/unix/sysv/linux/alpha/glob.c: Add prototypes for the __new_* functions. 1998-06-14 14:21 -0400 Zack Weinberg <zack@rabi.phys.columbia.edu> * stdio-common/bug12.c: New test for fread()/fseek(). * stdio-common/Makefile (tests): Add bug12.
151 lines
5.3 KiB
Plaintext
151 lines
5.3 KiB
Plaintext
The GNU C library contains an NSS module for the Hesiod name service.
|
|
Hesiod is a general name service for a variety of applications and is
|
|
based on the Berkeley Internet Name Daemon (BIND).
|
|
|
|
Introduction
|
|
============
|
|
|
|
The Hesiod NSS module implements access to all relevant standard
|
|
Hesiod types, which means that Hesiod can be used for the `group',
|
|
`passwd' and `services' databases. There is however a restriction.
|
|
In the same way that it is impossible to use `gethostent()' to iterate
|
|
over all the data provided by DNS, it is not possible to scan the
|
|
entire Hesiod database by means of `getgrent()', `getpwent()' and
|
|
`getservent()'. Besides, Hesiod only provides support for looking up
|
|
services by name and not for looking them up by port. In essence this
|
|
means that the Hesiod name service is only consulted as a result of
|
|
one of the following function calls:
|
|
|
|
* getgrname(), getgrgid()
|
|
* getpwname(), getpwuid()
|
|
* getservbyname()
|
|
|
|
and their reentrant counterparts.
|
|
|
|
|
|
Configuring your systems
|
|
========================
|
|
|
|
Configuring your systems to make use the Hesiod name service requires
|
|
one or more of the following steps, depending on whether you are
|
|
already running Hesiod in your network.
|
|
|
|
Configuring NSS
|
|
---------------
|
|
|
|
First you should modify the file `/etc/nsswitch.conf' to tell
|
|
NSS for which database you want to use the Hesiod name service. If
|
|
you want to use Hesiod for all databases it can handle your
|
|
configuration file could look like this:
|
|
|
|
# /etc/nsswitch.conf
|
|
#
|
|
# Example configuration of GNU Name Service Switch functionality.
|
|
#
|
|
|
|
passwd: db files hesiod
|
|
group: db files hesiod
|
|
shadow: db files
|
|
|
|
hosts: files dns
|
|
networks: files dns
|
|
|
|
protocols: db files
|
|
services: db files hesiod
|
|
ethers: db files
|
|
rpc: db files
|
|
|
|
For more information on NSS, please refer to the `The GNU C Library
|
|
Reference Manual'.
|
|
|
|
|
|
Configuring Hesiod
|
|
------------------
|
|
|
|
Next, you will have to configure Hesiod. If you are already running
|
|
Hesiod in your network, you probably already have a file named
|
|
`hesiod.conf' on your machines (probably as `/etc/hesiod.conf' or
|
|
`/usr/local/etc/hesiod.conf'). The Hesiod NSS module expects this
|
|
file to be found in the sysconfdir (`/usr/local/etc/hesiod.conf' by
|
|
default, see the installation notes on how to change this) or in the
|
|
location specified by the environment variable `HESIOD_CONFIG'. If
|
|
there is no configuration file you will want to create your own. It
|
|
should look something like:
|
|
|
|
rhs=.your.domain
|
|
lhs=.ns
|
|
|
|
The value of rhs can be overridden by the environment variable
|
|
HES_DOMAIN.
|
|
|
|
Configuring your name servers
|
|
-----------------------------
|
|
|
|
In addition, if you are not already running Hesiod in your network,
|
|
you need to create Hesiod information on your central name servers.
|
|
You need to run `named' from BIND 4.9 or higher on these servers, and
|
|
make them authoritative for the domain `ns.your.domain' with a line in
|
|
`/etc/named.boot' reading something like:
|
|
|
|
primary ns.your.domain named.hesiod
|
|
|
|
or if you are using the new BIND 8.1 or higher add something to
|
|
`/etc/named.conf' like:
|
|
|
|
zone "ns.your.domain" {
|
|
type master;
|
|
file "named.hesiod";
|
|
};
|
|
|
|
Then in the BIND working directory (usually `/var/named') create the
|
|
file `named.hesiod' containing data that looks something like:
|
|
|
|
; SOA and NS records.
|
|
@ IN SOA server1.your.domain admin-address.your.domain (
|
|
40000 ; serial - database version number
|
|
1800 ; refresh - sec servers
|
|
300 ; retry - for refresh
|
|
3600000 ; expire - unrefreshed data
|
|
7200 ) ; min
|
|
NS server1.your.domain
|
|
NS server2.your.domain
|
|
|
|
; Actual Hesiod data.
|
|
libc.group TXT "libc:*:123:gnu,gnat"
|
|
123.gid CNAME libc.group
|
|
gnu.passwd TXT "gnu:*:4567:123:GNU:/home/gnu:/bin/bash"
|
|
456.uid CNAME mark.passwd
|
|
nss.service TXT "nss;tcp;789;switch sw "
|
|
nss.service TXT "nss;udp;789;switch sw"
|
|
|
|
where `libc' is an example of a group, `gnu' an example of an user,
|
|
and `nss' an example of a service. Note that the format used to
|
|
describe services differs from the format used in `/etc/services'.
|
|
For more information on `named' refer to the `Name Server Operations
|
|
Guide for BIND' that is included in the BIND distribution.
|
|
|
|
|
|
Security
|
|
========
|
|
|
|
Note that the information stored in the Hesiod database in principle
|
|
is publicly available. Care should be taken with including vulnerable
|
|
information like encrypted passwords in the Hesiod database. There
|
|
are some ways to improve security by using features provided by
|
|
`named' (see the discussion about `secure zones' in the BIND
|
|
documentation), but one should keep in mind that Hesiod was never
|
|
intended to distribute passwords. In the origional design
|
|
authenticating users was the job of the Kerberos service.
|
|
|
|
|
|
More information
|
|
================
|
|
|
|
For more information on the Hesiod name service take a look at some of
|
|
the papers in ftp://athena-dist.mit.edu:/pub/ATHENA/usenix and the
|
|
documentation that accompanies the source code for the Hesiod name
|
|
service library in ftp://athena-dist.mit.edu:/pub/ATHENA/hesiod.
|
|
|
|
There is a mailing list at MIT for Hesiod users, hesiod@mit.edu. To
|
|
get yourself on or off the list, send mail to hesiod-request@mit.edu.
|