disks disabled for write and write errors a few days apart.

[DevXen]

Sunday i woke up to an email with 2 of my disks (one of my parity drives, and disk 4) disabled for write errors. i replaced both disks with new 14tb drives. the first disk i ran through chkdsk looking for bad sectors in windows and after 1.2 days it didn't find any errors. the 2nd disk is still checking and has like 7 days to go looking for bad sectors.

 

i installed each disk one at a time then did the parity rebuild. everything finished with no problems.

 

I then yesterday updated the Unraid OS to 6.12.6. (I was on 6.11.5). and i had been copying 4TB From my Array to my newly free disk i took out of the array. which is now going to be a cold storage backup disk. it copied that for 2 days i think (but i paused it to update the OS). resumed it last night. had no issues.

 

woke up today 2 of my disks are disabled due to read errors. one of them being one of the disks I replaced a couple days ago. then a few hours later 4 more disks i think it was that had 1 read error each. so i'm not sure what to do. i'm probably going to replace the one that was disabled that hasn't been replaced yet. then check it as well in chkdsk.

 

but here's my diagnostics i'm not sure whats causing this or how to fix it. if someone there can look through them and help me out, that would be great.

 

thank you.

-Dev

mediaxen-diagnostics-20231205-1126.zip

[JorgeB]

In my experience unlikely that two disks getting disabled at the same time would be a disk problem.

 

Dec  5 06:20:03 MediaXen kernel: aacraid: Host bus reset request. SCSI hang ?
Dec  5 06:20:03 MediaXen kernel: aacraid 0000:82:00.0: outstanding cmd: midlevel-0
Dec  5 06:20:03 MediaXen kernel: aacraid 0000:82:00.0: outstanding cmd: lowlevel-0
Dec  5 06:20:03 MediaXen kernel: aacraid 0000:82:00.0: outstanding cmd: error handler-28
Dec  5 06:20:03 MediaXen kernel: aacraid 0000:82:00.0: outstanding cmd: firmware-33
Dec  5 06:20:03 MediaXen kernel: aacraid 0000:82:00.0: outstanding cmd: kernel-0
Dec  5 06:20:03 MediaXen kernel: aacraid 0000:82:00.0: Controller reset type is 3
Dec  5 06:20:03 MediaXen kernel: aacraid 0000:82:00.0: Issuing IOP reset
Dec  5 06:20:59 MediaXen kernel: aacraid 0000:82:00.0: IOP reset succeeded

 

Looks more like a controller issue, but not very familiar with these controllers, see if it's well seated or try a different PCIe slot, or if you have another you could try with, like an LSI HBA do it.

[DevXen]

9 minutes ago, JorgeB said:

In my experience unlikely that two disks getting disabled at the same time would be a disk problem.

 

Dec  5 06:20:03 MediaXen kernel: aacraid: Host bus reset request. SCSI hang ?
Dec  5 06:20:03 MediaXen kernel: aacraid 0000:82:00.0: outstanding cmd: midlevel-0
Dec  5 06:20:03 MediaXen kernel: aacraid 0000:82:00.0: outstanding cmd: lowlevel-0
Dec  5 06:20:03 MediaXen kernel: aacraid 0000:82:00.0: outstanding cmd: error handler-28
Dec  5 06:20:03 MediaXen kernel: aacraid 0000:82:00.0: outstanding cmd: firmware-33
Dec  5 06:20:03 MediaXen kernel: aacraid 0000:82:00.0: outstanding cmd: kernel-0
Dec  5 06:20:03 MediaXen kernel: aacraid 0000:82:00.0: Controller reset type is 3
Dec  5 06:20:03 MediaXen kernel: aacraid 0000:82:00.0: Issuing IOP reset
Dec  5 06:20:59 MediaXen kernel: aacraid 0000:82:00.0: IOP reset succeeded

 

Looks more like a controller issue, but not very familiar with these controllers, see if it's well seated or try a different PCIe slot, or if you have another you could try with, like an LSI HBA do it.

 I don't have another one i can try with. it's weird cause it's been working fine for the last 4 years i guess. i'll try resenting it in the server.  then do i just unassign both these disks then reassign them and rebuild parity one them? - one at a time?

 

it was weird cause like i said i replaced 2, one at a time just a few days ago and the parity rebuild ran fine on both of them which is pretty intensive for the drives i hear.