We have random timeout when waiting for mirroring health. When this happens rbd-mirror fails to connect to the ceph mon but the system does not recover from the error. Restarting the rbd-mirror daemon fixes the issue.
This change:
Adds infrastructure for watching commands with a timeout
We have random timeout when waiting for mirroring health. When this happens rbd-mirror fails to connect to the ceph mon but the system does not recover from the error. Restarting the rbd-mirror daemon fixes the issue.
This change:
Restarting the rbd-mirror may not be needed in the future when the root cause is found and fixed.
Fixes #1332