Online Disk migration not working after upgrade

[DoruB]

Hello,

I have a Proxmox cluster made of 4 nodes. Before the upgrade I was running Proxmox 4.4 with GlusterFS 3.5.2-2+deb8u2 and Online Disk migration from GlusterFS storage to thinpool storage was working fine.

After upgrade to 5.0-20 and Gluster 3.8.8-1 Online Disk migration is not working anymore on 3 of the nodes. The configuration on all the nodes should be the same, so I don't see why it will work on 1 node, but not on the other 3.

This is the error I'm receiving when I'm trying to move the disk:

create full clone of drive virtio0 (vms1:103/vm-103-disk-2.raw)
  Using default stripesize 64.00 KiB.
  Logical volume "vm-103-disk-1" created.
drive mirror is starting for drive-virtio0
drive-virtio0: transferred: 20316160 bytes remaining: 42929356800 bytes total: 42949672960 bytes progression: 0.05 % busy: 1 ready: 0
drive-virtio0: Cancelling block job
  Logical volume "vm-103-disk-1" successfully removed
TASK ERROR: storage migration failed: mirroring error: VM 103 not running

Please note that it works fine if the VM is shut down, but since this is a Production environment it would really be great if it would work witht he VM powered on.

 

[wolfgang]

Hi,

please use the Glusterfs packages form Glusterfs.org and update your system to the current Version.