Regular crashes without warning

[DesertCookie]

My system started randomly crashing three days ago. After 5-30 minutes it would crash. Updating my UEFI yielded me one night of continuous operation before it went back to the old behavior. I switched from an AMD X399 1st-gen to an Intel Z590 11th-gen system two weeks ago and upgraded from 6.9 to 6.10 at the same tune; it was running without issues to this point.

 

In the instanced where unRAID would still start the array, the log was saved to the flash drive. I added the last log I have from when the server was still running fine and a few of the past days (some of them are 385MB in size!). I also managed to export diagnostics when I actually once made it to the web UI.

 

Current system:

• Gigabyte Z590 Aorus Master
• i5 11600K
• GTX 1650
• 2x16GB DDR4-3200
• 10GbE SFP+ Card
• 3x12TB HDD, 2TB HDD, 4TB HDD, 1TB M.2 SSD
• Enermax PSU (unlikely the culprit)
• APC UPS

 

I already tried:

• UEFI update
• unRAID update (6.10.2 to 6.10.3)
• removing GPU, drives (except M.2), and SFP+-card
• different RAM
• reseating all power and data connections
• unRAID safe-mode
• fresh unRAID boot stick (wasn't recognized by the UEFI until I copied all original data to it again)
• running without UPS

 

I am already suspecting a hardware error somewhere but cannot believe this new system would outright fail after two weeks (though that would be exactly my luck with tech).

 

20.06.22-1_syslog-1655701321 20.06.22-2_syslog-1655727397 06.06.22_syslog-20220603-181331.txt tower-diagnostics-20220620-0922.zip

Edited June 20, 2022 by DesertCookie
added more info

[JorgeB]

Nothing obvious in the logs, suggest blacklisting the i915 driver if you don't need it, it's been known to cause issues for some.

[DesertCookie]

 I enabled it as otherwise I could see /dev/dri for use in Jellyfin; the iGPU was the whole reason for my upgrade. I will try disabling it for now to see if that fixes the issue.

 

Edit: Blacklisting the driver did not change my situation. I now am further investigating the possibility of a hardware failure.

Edited June 22, 2022 by DesertCookie

[DesertCookie]

It seems to have been a temperature issue. The 11600K (60-70°C) runs a lot hotter than the previous 1900X (50-60°C) while consuming the same amount of power. I had left the RPM of my Noctua NF-A14s at the same 600RPM I used in the old Threadripper system and had removed the second NH-D15 fan. Adding the second fan back in and increasing the fan speed to 700RPM seems to have fixed the issue.