Problems after crashes [UNRAID 6.12.11]

[german_plz]

I had the Problem that my server crashes a short time after it booted. Because of that, I had to do a few unclean shutdowns.

The Server now runs again, but has Problems. I don't know if the crashes are fix. I can access my Shares.

 

The Problems are:

• Date And Time Incorrect (can't change it, doesn't allow me to safe it.)
• Parity-Check is really slow @ 273.1 KB/sec (tried to cancel it and now is stuck) / also fund a Sync error
• Can't mount any of my 10 Pool Devices - they are Unassigned
  ("Device '/dev/xxx' failed to mount. Check the syslog for details." Syslog: "Error: Device '/dev/sdm1' mount point 'system_cache' - name is reserved, used in the array or a pool, or by an unassigned device." )

Any tips on what I should do now? Need Help!

 

 

Also did a Memtest and it passed (18h / 10 passes). Also included: biggie-diagnostics-20240920-1402.zip &Biggie-Syslog.txt
 

Some Errors I saw before the server crashes. Not always happening. I don't know if that means anything:

Edited September 20, 2024 by german_plz
spelling error

[JorgeB]

Server is showing hardware errors, so you want to fix that first, syslog shows the md driver crashing, that is almost always a hardware problem.

[german_plz]

Thx for the replay!

Does the md driver crashing indicate that a drive has faild? Or just a general hardware error?
I'm having a hard time finding any devices that are not showing up in the current stat of the server.

 

[JorgeB]

21 minutes ago, german_plz said:

Does the md driver crashing indicate that a drive has faild?

No, usually a RAM, CPU or board issue, start by running memtest, or if you have multiple sticks try just one, if the same, try a different one, that will basically rule out the RAM.

 

[german_plz]

I ran Memtest last night for a sanity check because of the hardware errors:

[JorgeB]

memtest is only definitive if it finds errors, since you have multiple sticks try using the server with just a pair, if the same try the other pair, that will basically rule out bad RAM.

 

Also make sure this has been taken care of:

https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?do=findComment&comment=819173

 

[german_plz]

I have now installed just 2 out of 4 Sticks and Memtest is running again. Will run both pairs for 24h and see if errors occur.

 

In my BIOS there are no settings for the C-States and other Power Control settings or similar.

The BIOS is on the newest version. The RAM is not over overclocked, running at its native speed 2133MHz.

[JorgeB]

32 minutes ago, german_plz said:

and Memtest is running again

Instead of memtest, try using the server normally, with just that pair, and if it still crashes, try the other pair.

[german_plz]

OK, I think RAM is not the issue.

I tried both pairs three times, but the server crashed every time (always at a different point)

Also tried completely different RAM, just one stick, and the server crashed with that RAM too.

 

Can a GPU or SAS-Card also create these crashes? That's the only other components in there, except the Drives.

 

I'll be able to try a different mainboard and CPU combo tomorrow to see if that's the problem.
Can I just switch out the components, or does unraid have a problem with that?

Edited September 21, 2024 by german_plz
spelling error

[JorgeB]

56 minutes ago, german_plz said:

Can a GPU or SAS-Card also create these crashes?

It can, but CPU or board would be my main suspects.

[german_plz]

I tested the GPU in a Windows system and it works fine.

Booted from an Ubuntu live stick with no drives connected (no SAS-Card ether) and the system still crashes.

 

So you are right that it is either the CPU or board.

I remounted the CPU and cooler because I heard the TR4 Socket can be finicky. Didn't fix it.

 

I will buy a used motherboard / CPU combo on marketplace to replace and test my current hardware.

 

Strange that server just killed itself

My PC died a few days after, not the best timing…

 

Thank you @JorgeB for the troubleshooting help!

Really didn't think it was a hardware problem at first because it ran fine before.