Table of Contents
rancid-run - run rancid for each of the groups
rancid-run [-m mail_rcpt] [-r device_name] [group
[group ...]]
rancid-run is a sh(1)
script to run rancid(1)
for a
set of rancid group(s).
rancid-run reads rancid.conf(5)
to configure itself,
then uses control_rancid(1)
to run rancid(1)
for each rancid group. The
set of rancid groups can either be provided as command-line arguments or
via the LIST_OF_GROUPS variable in rancid.conf(5)
, if the group argument
is omitted.
A lock file is maintained per-group to prevent simultaneous runs
for a given group by rancid-run(1)
. The lock file will be named .<group>.run.lock
and will be located in $TMPDIR (see below).
A log file is produced under
$LOGDIR/logs for each rancid group processed. The file name will be <group>.YYYYMMDD.HHMMSS
(year month day . hour minute second).
rancid-run is normally used to run
rancid from cron(8)
. For example:
0 * * * * /usr/local/rancid/bin/rancid-run
The command-line options are as follows:
- -m mail_rcpt
- Specify the recipient
of diff mail, which is normally rancid-<group>. The argument may be a single
address, multiple comma separated addresses, or -m may be specified multiple
times.
- -r device_name
- Run rancid for a single device, device_name. device_name
should be name, as it appears in a group’s router.db. The device must be
marked "up". If a group is not specified on the command-line, rancid will
be run against any group in which the device_name appears.
The -r option
alters the subject line of the diff mail. It will begin with <group name>/<device
name> rather than just the group name alone.
rancid-run utilizes
the following environment variables from rancid.conf(5)
.
- BASEDIR
- Location
of group directories, etc. This is set to the "localstatedir" by the configure
script at installation time.
- LIST_OF_GROUPS
- List of rancid groups to collect.
- PATH
- Search path for utilities.
- TMPDIR
- Directory to hold temporary and
lock files.
If rancid fails to run or collect a device’s configuration,
the particular group’s log file (mentioned above) should be consulted. Any
errors produced by the revision control system (CVS or Subversion) or any
of the rancid scripts should be included there, whether they be a botched
cvs tree, login authentication failure, rancid configuration error, etc.
If the log file produces no clues, the next debugging step should be run
the commands manually. For example, can the user who runs rancid login
to the device with ’clogin hostname’, and so on.
- $BASEDIR/etc/rancid.conf
- rancid-run configuration file.
control_rancid(1)
, rancid.conf(5)
,
router.db(5)
Table of Contents