Hi All,

 

I'm experiencing issues with the unraid driver crashing when performing a parity check / array read check. The driver crashes right after a read error is detected. This started on 6.12.14 after I removed an empty data disk and ran tools > new config. I updated to version 7 stable, performed a 24 hour memtest, and stress tested my cpu and ran tools > new config again. The memtest and CPU stress test completed without error and I am still having the same issue. Read check finds an error and the driver crashes immediately after. I am brand new to unraid and am at a loss as what to try next to fix this issue. Any help would be greatly appreciated.

 

Jan 16 16:36:58 Tower kernel: md: recovery thread: multiple disk errors, sector=8570075568
Jan 16 16:36:58 Tower kernel: ------------[ cut here ]------------
Jan 16 16:36:58 Tower kernel: kernel BUG at drivers/md/unraid.c:1617!
Jan 16 16:36:58 Tower kernel: invalid opcode: 0000 [#1] PREEMPT SMP NOPTI
Jan 16 16:36:58 Tower kernel: CPU: 25 PID: 12438 Comm: unraidd0 Tainted: P        W  O       6.6.68-Unraid #1

 

tower-diagnostics-20250114-1636.zip tower-diagnostics-20250116-1717.zip

If it happens with both v6.12 and v7 it's almost certainly a hardware issue, and since you have a 14900K that would be my main suspect.

5 hours ago, JorgeB said:

If it happens with both v6.12 and v7 it's almost certainly a hardware issue, and since you have a 14900K that would be my main suspect.

 

I went to BestBuy last night picked up some new RAM, running a read check now and it has discovered 2 errors. No crash! I feel as if the RAM was unstable even though it was clearing an extended memtest. The 14900K is brand new and never ran on the tainted bios with voltage issues. I don't currently have a parity disk, waiting on 2 more HDDs to arrive. Is there anything I should do to correct the 2 errors that it has detected? Sorry if these seem like silly questions, still learning everything. 

 

 

tower-diagnostics-20250117-0732.zip

Edited January 17, 20251 yr by optid
Addition to post

Hmm, there can't be sync errors during a read check, only read errors:

 

Jan 16 20:20:10 Tower kernel: md: recovery thread: multiple disk errors, sector=2404718704

 

But there are no disk errors logged, not really sure where this comes from, possibly still related to the issues you are having.

12 hours ago, JorgeB said:

Hmm, there can't be sync errors during a read check, only read errors:

 

Jan 16 20:20:10 Tower kernel: md: recovery thread: multiple disk errors, sector=2404718704

 

But there are no disk errors logged, not really sure where this comes from, possibly still related to the issues you are having.

Got so fed up with it just went to Micro Center and picked up a new MB and the 12900K. Running another read check, will post back with results.