If anyone better at this kinda stuff could tell me what they think I'd be very appreciative.

 

After a reboot or even safe powerdown, when the system is started up it immediately starts up a parity check, sometimes finding sync errors and sometimes not. I don't want to keep running parity checks so I had a look at the logs and it seems my disk 2 has XFS errors with corrupted meta data...

```

Aug  3 23:07:13 Tower kernel: ata2.00: failed command: READ FPDMA QUEUED
Aug  3 23:07:13 Tower kernel: ata2.00: cmd 60/40:f8:c0:75:1e/05:00:00:00:00/40 tag 31 ncq dma 688128 in
Aug  3 23:07:13 Tower kernel:         res 40/00:f0:00:71:1e/00:00:00:00:00/40 Emask 0x50 (ATA bus error)
Aug  3 23:07:13 Tower kernel: ata2.00: status: { DRDY }
Aug  3 23:07:13 Tower kernel: ata2: hard resetting link
Aug  3 23:07:17 Tower kernel: mdcmd (37): nocheck cancel
Aug  3 23:07:19 Tower kernel: ata2: found unknown device (class 0)
Aug  3 23:07:20 Tower kernel: ata2: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
Aug  3 23:07:20 Tower kernel: ata2.00: configured for UDMA/100
Aug  3 23:07:20 Tower kernel: ata2: EH complete
Aug  3 23:07:22 Tower kernel: md: recovery thread: exit status: -4
Aug  3 23:07:34 Tower kernel: XFS (md2p1): Metadata corruption detected at xfs_dir3_data_reada_verify+0x53/0x64 [xfs], xfs_dir3_data_reada block 0x3d02fa30 
Aug  3 23:07:34 Tower kernel: XFS (md2p1): Unmount and run xfs_repair
Aug  3 23:07:34 Tower kernel: XFS (md2p1): First 128 bytes of corrupted metadata buffer:
Aug  3 23:07:34 Tower kernel: 00000000: 22 aa 78 d4 c1 33 56 90 3d dd 4d 64 24 52 56 2e  ".x..3V.=.Md$RV.
Aug  3 23:07:34 Tower kernel: 00000010: 5c 4d af 56 e8 16 83 e2 c2 a2 7b 8d 6d 48 45 99  \M.V......{.mHE.
Aug  3 23:07:34 Tower kernel: 00000020: 6f ba fa 58 d9 54 aa 75 6c af d4 c7 1e 1c 6e 8d  o..X.T.ul.....n.
Aug  3 23:07:34 Tower kernel: 00000030: 42 a7 62 2a 3c ee 4a 31 d4 ab 58 a8 5d 81 ea a3  B.b*<.J1..X.]...
Aug  3 23:07:34 Tower kernel: 00000040: 9a be b5 30 d4 47 bf 4f 16 cd a8 3b e5 93 02 94  ...0.G.O...;....
Aug  3 23:07:34 Tower kernel: 00000050: f9 6f 47 83 de 9f 9d 95 0f f5 65 f5 1f 07 13 6b  .oG.......e....k
Aug  3 23:07:34 Tower kernel: 00000060: 05 71 fc 6a 93 fc f2 61 b5 c3 78 c2 36 18 0c e2  .q.j...a..x.6...
Aug  3 23:07:34 Tower kernel: 00000070: e2 27 e2 c7 28 a4 58 13 91 e7 da 5e 61 7a fb 29  .'..(.X....^az.)
Aug  3 23:07:34 Tower kernel: XFS (md2p1): Metadata CRC error detected at xfs_dir3_block_read_verify+0x7c/0xf1 [xfs], xfs_dir3_block block 0x3d02fa30 
Aug  3 23:07:34 Tower kernel: XFS (md2p1): Unmount and run xfs_repair
Aug  3 23:07:34 Tower kernel: XFS (md2p1): First 128 bytes of corrupted metadata buffer:
Aug  3 23:07:34 Tower kernel: 00000000: 22 aa 78 d4 c1 33 56 90 3d dd 4d 64 24 52 56 2e  ".x..3V.=.Md$RV.
Aug  3 23:07:34 Tower kernel: 00000010: 5c 4d af 56 e8 16 83 e2 c2 a2 7b 8d 6d 48 45 99  \M.V......{.mHE.
Aug  3 23:07:34 Tower kernel: 00000020: 6f ba fa 58 d9 54 aa 75 6c af d4 c7 1e 1c 6e 8d  o..X.T.ul.....n.
Aug  3 23:07:34 Tower kernel: 00000030: 42 a7 62 2a 3c ee 4a 31 d4 ab 58 a8 5d 81 ea a3  B.b*<.J1..X.]...
Aug  3 23:07:34 Tower kernel: 00000040: 9a be b5 30 d4 47 bf 4f 16 cd a8 3b e5 93 02 94  ...0.G.O...;....
Aug  3 23:07:34 Tower kernel: 00000050: f9 6f 47 83 de 9f 9d 95 0f f5 65 f5 1f 07 13 6b  .oG.......e....k
Aug  3 23:07:34 Tower kernel: 00000060: 05 71 fc 6a 93 fc f2 61 b5 c3 78 c2 36 18 0c e2  .q.j...a..x.6...
Aug  3 23:07:34 Tower kernel: 00000070: e2 27 e2 c7 28 a4 58 13 91 e7 da 5e 61 7a fb 29  .'..(.X....^az.)
Aug  3 23:07:34 Tower kernel: XFS (md2p1): metadata I/O error in "xfs_da_read_buf+0x9a/0xff [xfs]" at daddr 0x3d02fa30 len 8 error 74

```

It says  to unmount the disk and run XFS repair, is there anything I should know before doing this? I have my data backed up so I'm not stressed but I'd like to not cause any further damage or complications.

Also, what would have caused this? The system is on a UPS and inverter for backup power... Could it have been the docker issues with 6.12.0? (I upgraded when it was released but now I'm on 6.12.3)

I dont think it's a faulty cable as I had replaced the ones for my main array fairly recently although still possible I guess...

Thanks in advance!

Here's the Diags just in case the above isn't enough.

tower-diagnostics-20230803-2351.zip

Your getting an unclean shutdown, start by adding 60 secs to the default timeout (Settings -> Disk Settings), if it doesn't help time how long it takes for the array to stop.