I'm using Unraid on a custom built server: Supermicro X11SSL-F, Intel® Xeon® CPU E3-1230 v6 @ 3.50GHz, 8Gb DDR4 ECC Samsung memory. It's been running smoothly for about 2 years till Unraid version 6.8.1. I heavily use Docker (about 10-15 apps, home automation, development DB servers, Plex, game servers,...) but it's a home server with low load. It's far from overloading.

 

Few months ago, after playing a lot with Docker, adding/removing new packages... server started shutting down suddenly. Rarely, first it took couple of weeks to shutdown, then 2 or 3 times in  a week or two. Last shutdown broke docker image and I had to reinstall all dockers. I used the opportunity to upgrade to 6.9.2. and it's been ok for almost a month now. However, this morning it happened again, early in the morning... with absolutely no load on it. This time I haven't rebooted right away. I connected through 2nd NIC and KVM and saw the last error message: "Kernel panic - not syncing: Fatal exception in interrupt. Kernel Offset: disabled".

 

I have only a mobile screenshot of this, syslog doesn't survive reboot. So for now, the only thing I did is enabling Syslog, so next time I have more info.

 

Anything else I could do to catch the cause of this? Something to look at right away or wait for annother incident, this time with syslog?

You could always share your diagnostics before a crash. That will not show the crash itself but will provide more information on the system and its configuration.

Your syslog after a crash will still provide more valuable information about the crash specifically.

Well, here is the one few hours after reboot today. I guess now I can only wait for the next incident to posto more info.borg-diagnostics-20220306-2224.zip 

So a week passed, server crashed again but my syslog is empty!? I don't know what I'm doing wrong. Settinmg > SyslogServer: Local syslog server: enabled. Local syslog folder is "temp"... but folder is empty. I just don't have a clue how to reach logs before crash and every try takes a week. It's really a painfull process to get to the bottom of this.

 

But attached is error screenshot from IPMI. Does it mean error was on `do_raw_spin_lock`?

 

Edited March 13, 20224 yr by Psihoza
added screenshot

A Month passed without an issue so I think I found out what the problem was. It was related to a VM. I figured out that was the only change I did to my Unraid setup, was playing with one VM so I activated it, and VMs were disabled before. So, even if my test VM machine was turned off, it seems that just enabling VMs in Unraid settings causes this on my machine.

 

Since I don't really use VMs, I'm not really motivated to investigate further. I just turned it off again.