Hey,
is the problem back with qemu/dynamic_ownership?
We have one DC with two clusters: one: vdsm-4.30.38 / libvirt-4.5.0-23,
two: vdsm-4.30.40 / libvirt-4.5.0-23, LibgfApi enabled, Gluster-Server
6.5.1, Gluster-Client 6.7, opVersion 60000
It went well for a log time but now, with getting vdsm like above, we
are getting old problems:
°setting disk files to root:root when VM went down
°cannot create snapshot etc.
Do we have to set dynamic_ownership=0 in qemu.conf? We were testing
anything but nothing other worked.
Is there any idea?
Christoph