Hi,

 

I've been on Unraid 6.12.8 and few days from know my server crashed. It was on but WebUI wasn't loading, SSH wasn't working, I haven't even seen server on my router. So I did hard reset.

 

From this moment I can't properly reboot my server. If done by WebUI it counts indefinitely but when I open WebUI in new window there is text on the bottom "Array Stopping Unmounting disks..." and one of the cores is on constatnt 100% load.

 

Reboot from SSH by reboot or shutdown command dosen't work either. Only this command works and after reboot there is unclean shutdown notification:
echo b >/proc/sysrq-trigger

 

I've updated to Unraid 6.12.10 and problem still occurs. I've tried swtich macvlan to ipvlan. Still nothing.

 

I've enabled mirroring syslog to sdcard and here is the diagnostics.

unraid-diagnostics-20240415-2003.zip

The current diags don't show the pool mounting, and that pool was also apparently the problem before, does it still mount?

Yes, after forced reboot pool mounts successfully and I can use them normally.

Post new diags.

Now tried to shutdown from WebUI and I connected video output. On screen last output was "Generating diagnostics..." and It stayed there for couple of minutes before I forced the shutdown by switch.

unraid-diagnostics-20240416-1729.zip

When I run in safe mode and tried to start array manually I saw almoast immediately this output in terminal:
kernel:PANIC at range_tree.c:436:range_tree_remove_impl()

Rebooting in safe mode with no dockers and VMs is also stuck on unmouting and gathering diagnostics.

The zfs pool is corrupt, you should backup and re-format, see if it mounts read-only without panicking, you can then backup.

 

Reboot the server, don't start the array, and type:

 

zpool import -o readonly=on cache

 

If OK, start the array, cache will show unmountable in the WebGUI, but the data will be under /mnt/cache, backup and then re-format.

 

 

Thanks, that was it. Took some time but now all is working.