I have tested an upgrade from engine 3.3.3 to 3.3.4 under centos 6. It seems that the vdsm nodes are now showing as non-operational and the only way I have been able to cure this was to:
- Upgrade the vdsm version on a node to 4.13.3-4
- install vdsm-gluster
- start glusterd service
We are only currently using ISCSI. I cannot get a node running vdsm-4.13.3-3 to show as being operational under engine 3.3.4
Has anyone been able to get 3.3.4 functioning with previous version vdsm nodes ?
This issue may hinder our progress when it comes time to upgrade our production engine / nodes when putting hosts into maintenance mode and performing migrations etc.
Thanks
Gary Lloyd----------------------------------
IT Services
Keele University
-----------------------------------