Hi, has the PC also been updated to the latest release?

yes. It only appears to be affecting clusters that we’ve updated to 5.20.3.5. We have one newer one that shipped with 6.x, and it was ok after the update.

I experienced similar issues to this after upgrading PC, and discovered the CVM was out of disk space. Maybe run a “df - h” command if you can get to the CVMs and check the usage on the /home dir.

Nope, not seeing high disk usage on the CVMs.

Rather basic but have you tried rebooting the PC as well? If that doesn’t help I think you’ll need to reach out to support, see if it’s a known bug with the latest release of 5.20.x

PC reboot unfortunately did not clear the issue. We have an SR open, but crickets so far. Uncharacteristically long response time on this one.

What version of PC and can you share the SR#?

same issue here after upgrading the AOS to 5.20.3.5 LTS. It’s ok with other cluster with 5.20.3 LTS so it seems to be related to the prism element and not to PC pc.2022.1.0.2.

so far, the SE was thinking something is going on with Mercury on the PC cluster. We see this issue with every cluster we move to 5.20.3.5 LTS. One cluster was newer and shipped with 6, an it’s unaffected. PC v. 2022.10.2. Still waiting for his further analysis.

sorry, PC 2022.1.0.2.

You are hitting an known bug, see this KB for the solution: Unable to launch Prism Element from Prism Central due to missing "PrismUI.tar.gz" file after upgrade.

The SE went through, that, and it didn’t help. He said he was seeing some logs that pointed him to an issue with the mercury service. Waiting on an update.

Got it!

Poked around some today. The various version folders on the PC nodes were missing the ‘console’ folder. KB 11663 gives a command to run to manually extract this. Once I did that on all the PC nodes, was again able to launch PE from PC.