Crash tonight: end kernel panic - not syncing: Fatal exception in interrupt

[adminmat]

I noticed my server was not responding. I went in via IPMI and saw this so I rebooted. The GUI is up and I'm logged in. This is freaking me out. I rebooted and when I started the array I got this message, But the NVME Cache drive seemed to mount ok.
 

 

Any Idea what caused this? Really appreciate some help.

Should my next step be a MemTest?

 

tower-syslog-20210107-0522.zip

 

 

Edited January 7, 2021 by adminmat

[JorgeB]

Please post the diagnostics: Tools -> Diagnostics

[adminmat]

8 hours ago, JorgeB said:

Please post the diagnostics: Tools -> Diagnostics

Thanks. I'll add that there was one movie stream running from plex at the time and I think it crashed right as I opened a 2nd plex stream for Live TV 

tower-diagnostics-20210107-1111.zip

[JorgeB]

Pool seems fine, at least for now, if it keeps crashing try this to see if it catches anything.

[adminmat]

6 minutes ago, JorgeB said:

Pool seems fine, at least for now, if it keeps crashing try this to see if it catches anything.

Thanks Jorge. When it crashes does it risk the data on the disks? Or does it gracefully shut them down? Is there something I can do to make sure the disks are stopped properly when I reboot? I get the option to "Power off - Orderly Shutdown" with my motherboard BMS 

 

[JorgeB]

If it crashes there's always some risk of filesystem corruption, but not much you can do about that.

[adminmat]

2 hours ago, JorgeB said:

If it crashes there's always some risk of filesystem corruption, but not much you can do about that.

Right. So I set up the rsyslog server on another server (Ubunt in ESXi) and it's currently mirroring the log. And a thank you to @Frank1940 for the help too. If anyone has additional thoughts as to why it would crash please let me know.

[adminmat]

@JorgeB is there a way to check for file corruption before I do a backup of all my appdata etc?

 

[JorgeB]

https://wiki.unraid.net/Check_Disk_Filesystems

[adminmat]

12 hours ago, JorgeB said:

https://wiki.unraid.net/Check_Disk_Filesystems

I ran the Scrub on the Cache drive. These instructions are a little outdated when using 6.9.0-rc2. I assume now you just click "Scrub" and that's it. I don't see this option on the Cache 2 drive so I assume when scrubbing Cahce it does both? The "Check" button is greyed out and says "Check is only available when array is Started in Maintenance mode" which is in conflict with the Wiki instructions for BTRFS.

Also, do I need to run the Balance? I've not done this before.

 

[JorgeB]

10 hours ago, adminmat said:

so I assume when scrubbing Cahce it does both?

Correct.

 

10 hours ago, adminmat said:

The "Check" button is greyed out and says "Check is only available when array is Started in Maintenance mode"

This is correct, only scrub can be done with the fs online, check in btrfs should only be done without the repair option.

 

11 hours ago, adminmat said:

Also, do I need to run the Balance?

Usually not needed as a regular option.

[aronwk]

Did you ever figure this out?
I'm getting these exact kernel panic infrequently for a couple of months now. but have had it happen a few times over the past week.
I'm going to setup a syslog server to hopefully catch more info about the issue, but any help is appreciated
 

[trurl]

Generally you should start your own support thread (or use the support thread for help with a specific plugin or docker). But since this is an old thread and nobody else is using it we can try to help you here.

 

Syslog server is needed since otherwise syslog will not survive reboot. But you should also attach your diagnostics to your NEXT post in this thread. That will give us more information about your hardware and configuration than we can get from just syslog.