(SOLVED) unRAID crashes randomly, can't find cause

[CameraRick]

Hi there,

 

yesterday my unRAID crashed on me, while I was at work. I noticed because my DAV services gave an error on my phone. I thought nothing of it, things can happen, and rebooted the machine soon after. This morning, it crashed again, and now I'm curious.

It's also for no appereant reason, there was no load on the machine. I run some dockers (usually 8-9 are active, 17 are installed) and one Win 10 VM (which exclusively is there so I can TeamViewer on there to do things on remote). Only docker that usually pulls some heavier load is Plex, but no one was using it at the time.

 

Crash times were yesterday, 4th May around 16:00h and today, 5th May around 09:30h.

 

When I came home now, I tried to connect locally to see if it was accessible (could've been that just all Dockers and VMs acted up or whatever), no dice. I pinged it, you can see the result attached; 8 is the ending of the IP of unRAID, 43 from my desktop PC. I had a look into my router menu, there it was still listed as active connection (screenshot also attached)

 

I can't really see what happens, but I'm also not good with logs  I attached the syslog (mirror to flash) and the diagnostic file which I just created after the reboot.

Anyone might know what to look for? I didn't do a memtest yet (because time), I use a desktop mainboard but have ECC memory (ECC is enabled in BIOS)

 

Best,

Rick

 

 

syslog.zip lunas-diagnostics-20220505-1856.zip

Edited May 8, 2022 by CameraRick

[JorgeB]

May  5 10:26:53 LuNAS kernel: ? Sys_DumpTask+0xe9/0xf1 [corefreqk]
May  5 10:26:53 LuNAS kernel: Cycle_AMD_Family_17h+0x31b/0xb37 [corefreqk]

 

Unisntall the corefreq plugin.

[CameraRick]

Hi Jorge,

 

thank you, I did it at once. What exactly does the data from the syslog mean?

 

best,

Rick

[JorgeB]

It means the plugin is crashing.

[CyrIng]

On 5/5/2022 at 5:12 PM, CameraRick said:

Hi there,

 

yesterday my unRAID crashed on me, while I was at work. I noticed because my DAV services gave an error on my phone. I thought nothing of it, things can happen, and rebooted the machine soon after. This morning, it crashed again, and now I'm curious.

It's also for no appereant reason, there was no load on the machine. I run some dockers (usually 8-9 are active, 17 are installed) and one Win 10 VM (which exclusively is there so I can TeamViewer on there to do things on remote). Only docker that usually pulls some heavier load is Plex, but no one was using it at the time.

 

Crash times were yesterday, 4th May around 16:00h and today, 5th May around 09:30h.

 

When I came home now, I tried to connect locally to see if it was accessible (could've been that just all Dockers and VMs acted up or whatever), no dice. I pinged it, you can see the result attached; 8 is the ending of the IP of unRAID, 43 from my desktop PC. I had a look into my router menu, there it was still listed as active connection (screenshot also attached)

 

I can't really see what happens, but I'm also not good with logs  I attached the syslog (mirror to flash) and the diagnostic file which I just created after the reboot.

Anyone might know what to look for? I didn't do a memtest yet (because time), I use a desktop mainboard but have ECC memory (ECC is enabled in BIOS)

 

Best,

Rick

 

 

Hello,

 

Using an AMD Zen processor, CoreFreq is incompatible with any monitoring tool running in parallel.

 

CyrIng

[CyrIng]

EDIT: The root cause is that I'm missing the base address of the Cezanne memory controller.

 

See project issue AMD/Zen3/Cezanne UMC base address is unknown