drbd-wait-promotable@RESNAME.service - Wait for DRBD to become promotable
Usually you do not want to "hardcode" an unconditional promotion attempt of DRBD resources, because you usually cannot know a-priory whether this instance of DRBD will have access to good data (yet).
If you try to promote (mount, use) DRBD before it can be certain to have access to good data, the attempt will fail in various ways, typically claiming "wrong medium type", "need access to good data".
This unit will wait for the DRBD internal state to indicate that it has access to good data. It does not monitor the state further, so this is good enough only to avoid trying to mount "too early" if you hard-code the mount attempt into your boot sequence, for example using an fstab entry or similar.
To monitor and react to DRBD state changes, use drbd-reactor or a cluster manager like pacemaker.
If you systemctl enable this unit, it will be required by drbd@.target, so anything depending on that target, for example an fstab entry, will implicitly wait until this DRBD indicated it was "healthy enough" to be used.
It can still timeout, in which case systemd sees this as a start failure, and starting of any dependencies (e.g. the mount) won’t be attempted at this point.
drbd@.service(7), drbd@.target(7), drbd-lvchange@.service(7)
LINBIT HA Solutions GmbH https://linbit.com