am very much a newbie and slowly learning things, but wife accidently pulled the power and instigating a parity check and errors are climbing faster than the check now sitting at 100000, 4 hours left to run

 

so i am guessing my parity is trash and needs to be corrected but is there anything i could look at to understand why or any other info i can supply, no errors showing on the drives

 

12tb parity x 2

12tb data x 2

8tb data x 6

 

Hello @Maff,

given the abrupt power cut off I guess that errors are pretty likely, the question is whether the errors are on Parity, data drives or both.

Do you know if there were writes when the power was cut (or a few minutes before) ?

 

Still, your diagnostics might provide more details. Go to Tools / Diagnostics and attach the full zip to your next post.

Hi @ChatNoir thank you for your time

 

i don't believe anything was being written when the power outage happened and not for at least an hour before

 

the log is taken whilst parity check is still running, let me know if one is required when it has finished

tower-diagnostics-20210608-2041.zip

Edited June 8, 20215 yr by Maff

1 hour ago, Maff said:

the log is taken whilst parity check is still running, let me know if one is required when it has finished

 

Don't have much experience with disk errors, luckily for me. So I'd wait for feedback from more experienced users.

 

However, I see plenty of this.

Jun  7 20:26:03 Tower kernel: pcieport 0000:00:03.0: AER: Corrected error received: 0000:04:00.0
Jun  7 20:26:03 Tower kernel: nvme 0000:04:00.0: PCIe Bus Error: severity=Corrected, type=Physical Layer, (Receiver ID)
Jun  7 20:26:03 Tower kernel: nvme 0000:04:00.0:   device [15b7:5002] error status/mask=00000001/0000e000
Jun  7 20:26:03 Tower kernel: nvme 0000:04:00.0:    [ 0] RxErr 

 

This would probably need to be taken care of too, once the primary issue is dealt with.

i am guessing that is the pci-e NVME card i have, will try and do some research, one is used for vm another dockers, and 2 in raid for a cache drive(not set up with shares yet) but i dont run a huge amount and not noticed anything wrong with them but i guess as corrected?

 

think i found the device and i can poss change this one out to a different one

[15b7:5002] 04:00.0 Non-Volatile memory controller: Sandisk Corp WD Black 2018/SN750 / PC SN720 NVMe SSD

[N:0:8215:1] disk WDC PC SN720 SDAPNTW-256G-1006__1 /dev/nvme0n1 256GB

 

 

parity check has finished newest diags

tower-diagnostics-20210608-2309.zip

Edited June 8, 20215 yr by Maff
add diags

55 minutes ago, Maff said:

parity check has finished

This was an automatic non-correcting parity check due to unclean shutdown. You will have to run a correcting parity check to actually correct sync errors.

 

How many sync errors did you get in the end? The large number of sync errors you already mentioned, plus the fact that all the sync errors I saw in the log were only for parity and not for parity2, makes me wonder if parity was valid to begin with.

 

You should get an UPS.

was 358761 errors after the parity check, have ran an error correcting check and all the writes went to parity one, so not super sure what went wrong as i built the server with both disks presant and slowly transfered over the data to it