I'm currently in the process of migrating all of our virtual servers from OVirt to new Hyper-V 2016 clusters.
Two of the VMs that have been migrated are running Centos 7.5 and they have a few docker containers on them.
They both have the hyper-v tools installed on them that come from the Centos repositories.
When I take a checkpoint of these VMs they hang permanently with the following recurring message:
task hv_vss_daemon:1397 blocked for more than 120 seconds. and the last entry in /var/log/messages is about Hyper-V doing a hot backup.
At this point the VM's are unusable and I have to force them off via task manager as the native hyper-v 'turn off' button refuses to do anything.
Things I have tried:
Upgrade to the latest version of Linux Integration tools from the MS website - this made no difference
Disable SELinux on the guest - this made no difference.
Things I did that made it work:
- Stop the docker containers and create a checkpoint (checkpoint seems to complete without issue)
- Remove the Hyper-V agents from the guests (checkpoint seems to complete without issue)
Has anyone come up against this before ?
For the time being, I'm just backing up via Veeam with the hyper-v vss agent removed from these guests as crash consistent backups should be fine for these.