Hi,
I have a cluster which has three nodes.
Two SAN volumes (from IBM SVC) are mapping to the cluster nodes.
One is for the quorum, and the other is for CSV.
After adding this volume for CSV usage as an available storage,
this volume can be moved whether using draining the host which owns the volume or moving the volume manually,
so as the quorum volume.
However, after adding the volume as a CSV, I could not move the volume either using draining the host or moving manually.
At the meantime, if the quorum is owned by the server which just is drained, the quorum moves to another node immediately.
But the CSV will show "online pending" and wait for timeout and show as failed.
After looking into the error log, it seems that the original node owner do not release the volume after the moving command is triggered, so the new node owner could not take over the volume successfully.
The new node owner also shows the error "ReleaseDiskPRFailure".
I've tried to validate the cluster without any error.
Do you know any possibilities that CSV cannot be moved properly?
Thanks.