[ClusterLabs] SLES15 SP2: Problem deactivating Clustered RAID

[Ulrich Windl]

Hi!

I'm fixing my configuration after each unexpected fencing. Probably I missed some dependency, but these messages seem somewhat unexpected to me:

Jan 11 15:54:32 h18 Raid1(prm_lockspace_raid_md10)[30630]: INFO: running lsof to list /dev/md10 users...
Jan 11 15:54:32 h18 systemd[1]: Stopping LVM event activation on device 9:1...
Jan 11 15:54:32 h18 Raid1(prm_lockspace_raid_md10)[30645]: WARNING: command failed: lsof /dev/md10
Jan 11 15:54:32 h18 Raid1(prm_lockspace_raid_md10)[30651]: INFO: Attempting to mark array /dev/md10 readonly
Jan 11 15:54:32 h18 Raid1(prm_lockspace_raid_md10)[30659]: ERROR: Failed to set /dev/md10 readonly (rc=1)

(resource-agents-4.4.0+git57.70549516-3.12.1.x86_64)

Earlier the RA had reported:
Jan 11 15:54:31 h18 Raid1(prm_lockspace_raid_md10)[30537]: WARNING: virtlockd 4684 root   12uw  REG   9,10        0 124945 /var/lib/libvirt/lockd/lvmvolumes/6X7qr3-GXDN-zjMo-vwF2-2MeG-vftx-XzFxOb
Jan 11 15:54:31 h18 Raid1(prm_lockspace_raid_md10)[30550]: WARNING: virtlockd 4684 root   13uw  REG   9,10        0 124946 /var/lib/libvirt/lockd/lvmvolumes/VaKRkB-ap2u-flZm-1RDY-Ba5u-222P-cL3c1A

(virtlockd.socket is enabled in systemd)
So how do you implement clusterwide virtlockd?

Regards,
Ulrich