What is the best practice to upgrade a cluster from 4.2 to 4.3

[Matthias Looss]

I am still fairly new to PROMOX but I have a working 4 node cluster (not HA) running 4.2 and I won't be able to upgrade and reboot all nodes at the same time. Will this be a problem with the new interface if some nodes have not been upgraded?

Thanks!

 

[dietmar]

Will this be a problem with the new interface if some nodes have not been upgraded?

If you connect to a new with the new GUI, there will be some functions not available on older versions, so you may get errors when you try that on older nodes. But everything else should work.

 

[Matthias Looss]

I went ahead and updated each node and they started to show the new v4.3 GUI and it was working just fine. I did not restart any nodes until this evening and even so I only restarted one node, all my nodes went offline? I restarted because I was adding another node to my cluster already running 4.3 and even so I could see the node in the cluster, it kept saying waiting for quorum for a very long time and never reached it. This is the reason I decided to restart one cluster node. The node I restarted is up and running now but I have these errors in the log:

Oct 05 19:43:25 hn3 pveproxy[4080]: proxy detected vanished client connection
Oct 05 19:43:25 hn3 pveproxy[4080]: proxy detected vanished client connection
Oct 05 19:43:25 hn3 pveproxy[4080]: proxy detected vanished client connection
Oct 05 19:45:09 hn3 systemd-timesyncd[1322]: interval/delta/delay/jitter/drift 2048s/-0.008s/0.111s/0.003s/-12ppm
Oct 05 19:46:05 hn3 pvedaemon[2210]: <root@pam> successful auth for user 'root@pam'
Oct 05 19:47:33 hn3 pvedaemon[2210]: <root@pam> successful auth for user 'root@pam'
Oct 05 19:58:56 hn3 pvedaemon[5350]: starting vnc proxy UPID:hn3:000014E6:000465E9:57F593C0:vncproxy:103:root@pam:
Oct 05 19:58:56 hn3 pvedaemon[2210]: <root@pam> starting task UPID:hn3:000014E6:000465E9:57F593C0:vncproxy:103:root@pam:
Oct 05 19:58:56 hn3 qm[5353]: VM 103 qmp command failed - VM 103 not running
Oct 05 19:58:59 hn3 pvedaemon[5350]: command '/bin/nc6 -l -p 5900 -w 10 -e '/usr/sbin/qm vncproxy 103 2>/dev/null'' failed: exit code 1
​