ganeti-watcher - Ganeti cluster watcher
ganeti-watcher [–debug] [–job-age=*age* ] [–ignore-pause] [–rapi-ip=*IP*]
The ganeti-watcher is a periodically run script which is responsible for keeping the instances in the correct status. It has two separate functions, one for the master node and another one that runs on every node.
If the watcher is disabled at cluster level (via the gnt-cluster watcher pause command), it will exit without doing anything. The cluster-level pause can be overridden via the --ignore-pause option, for example if during a maintenance the watcher needs to be disabled in general, but the administrator wants to run it just once.
The --debug option will increase the verbosity of the watcher and also activate logging to the standard error.
The --rapi-ip option needs to be set if the RAPI daemon was started with a particular IP (using the -b option). The two options need to be exactly the same to ensure that the watcher can reach the RAPI interface.
Its primary function is to try to keep running all instances which are marked as up in the configuration file, by trying to start them a limited number of times.
Another function is to “repair” DRBD links by reactivating the block devices of instances which have secondaries on nodes that have been rebooted.
The watcher will also archive old jobs (older than the age given via the --job-age option, which defaults to 6 hours), in order to keep the job queue manageable.
The watcher will restart any down daemons that are appropriate for the current node.
In addition, it will execute any scripts which exist under the “watcher” directory in the Ganeti hooks directory (/etc/ganeti/hooks). This should be used for lightweight actions, like starting any extra daemons.
If the cluster parameter maintain_node_health is enabled, then the watcher will also shutdown instances and DRBD devices if the node is declared as offline by known master candidates.
The watcher does synchronous queries but will submit jobs for executing the changes. Due to locking, it could be that the jobs execute much later than the watcher submits them.
The command has a set of state files (one per group) located at /var/lib/ganeti/watcher.GROUP-UUID.data (only used on the master) and a log file at /var/log/ganeti/watcher.log. Removal of either file(s) will not affect correct operation; the removal of the state file will just cause the restart counters for the instances to reset to zero, and mark nodes as freshly rebooted (so for example DRBD minors will be re-activated).
In some cases, it’s even desirable to reset the watcher state, for example after maintenance actions, or when you want to simulate the reboot of all nodes, so in this case, you can remove all state files:
rm -f /var/lib/ganeti/watcher.*.data
rm -f /var/lib/ganeti/watcher.*.instance-status
rm -f /var/lib/ganeti/instance-status
And then re-run the watcher.