I recently rebuilt my array after some drives failed, or went missing.  Things were fine for a day or two then one morning this week unraid wasn't responding checking on it I saw what looks to me a link a kernel panic.

<screen shot attached>

I was forced to power cycle unraid, and twice now during a parity sync the same thing has happen by the next morning.

Anyone have any ideas what may be the problem?

Should I just start by running old school memtest or something?

phoenix-diagnostics-20231102-1258.zip

Enable the syslog server and post that after a crash.

Thanks JorgeB I was able to do that, and looks like that syslog data is included in the diag.zip I was able to grab from the next crash.

This morning I found unraid again not responding when I try to  open the webGUI via its usual IP, and found a similar kernel panic error on screen when I checked it.

After forcing another reboot I grabbed the diag.zip again here nothing is jumping out at me as I looked through the logs, hopefully someone sees something I'm not.

 

phoenix-diagnostics-20231105-0912.zip

The syslog in the diagnostics does not include the syslog server captured file which needs posting separately.

Ahh I see, here's a copy of the logs directory from the flash drive.  That documentation seems to indicate it should have been mirroring the logs to there.

But I also noticed this folder selection in the settings, but I guess that must be something else I don't see anything pertaining to system logs in the selected appdata folder.

logs.zip

5 minutes ago, snowmirage said:

But I also noticed this folder selection in the settings, but I guess that must be something else I don't see anything pertaining to system logs in the selected appdata folder.

You nee to input the server IP in the remote syslog server field.

 

6 minutes ago, snowmirage said:

here's a copy of the logs directory from the flash drive.

 

Nov  3 22:34:00 phoenix kernel: macvlan_broadcast+0x10a/0x150 [macvlan]
Nov  3 22:34:00 phoenix kernel: ? _raw_spin_unlock+0x14/0x29
Nov  3 22:34:00 phoenix kernel: macvlan_process_broadcast+0xbc/0x12f [macvlan]

Macvlan call traces will usually end up crashing the server, switching to ipvlan should fix it (Settings -> Docker Settings -> Docker custom network type -> ipvlan (advanced view must be enabled, top right)).

 

 

I went back and filled in the server IP of my unraid server.

I also adjusted the docker networking setting you suggested and changed it from macvlan to ipvlan

I don't see anything different in the logs folder on the flash drive.   attached as log2.zip

logs2.zip

But I did find a syslog file in the appdata folder, maybe that is being mirrored to the flash drive and I just didnt see it?

syslog-192.168.0.216.log

The 'syslog' file in the 'logs' folder is the one that is being mirrored to the flash. 

58 minutes ago, snowmirage said:

I also adjusted the docker networking setting you suggested and changed it from macvlan to ipvlan

And is it still crashing?

9 hours ago, JorgeB said:

And is it still crashing?

Yes it is.  Took longer this time was up for ~2 days.

Woke up to it crashed this morning, before I grabbed those last logs above.  Same crash screen on the monitor as above (kernel panic).

 

Last macvlan crash I see logged is from Nov 6th, and no toher call traces after that, did you reboot after changing to ipvlan?

5 hours ago, JorgeB said:

Last macvlan crash I see logged is from Nov 6th, and no toher call traces after that, did you reboot after changing to ipvlan?

It has crashed once since I changed to ipvlan.

Found it that way the morning of Nov 8th if I recall correctly and had to do a hard reboot (reset switch).  As of 10am Nov 9th its still up and running.

If it crashes and I have to hard reset again, I'll grab new logs and diag files.  Hopefully it was just that macvlan setting or something else points to a fix.

Thank you for the help