I've been having this issue a lot lately but it got really bad recently, just had one take longer than the 15mins my UPS had and seems to be the cause of corruption on one of my cache disks.

I've already looked at my array stop times, that's under a min so it's not any of those causes

Today has been the worst case. While in safe mode to work on recovering my corrupted drive, my array shut down very quickly and 10 mins after that with nothing in the syslog I logout of the directly connected gui, ssh in and run 'shutdown now' from terminal. Nothing in the syslogs for a few mins other than from the shutdown command. Then after a few mins, all happening in less than half a second, the syslog is inundated with the same line repeatedly 'publish curl to diagnostic tool failed', the gui session ends, my ssh session ends, and the direct monitor connection goes to the console with minimal logs (attached photo).

After all that I wait another 20mins of nothing written to console, short press power, nothing, repeat a couple times. Wait another 5mins, still nothing new on the console and just gave up and did a hard reset

Edited August 24, 2025Aug 24 by c0d3m0nk3y

I will be trying a new flash drive in case that's the cause, but I'm out of spares so need to go get a pack and wanted to get this down while it was fresh