Hey everyone,

 

Looking for advice on next steps on how to deal with some disk errors I have been fighting with. About a month ago I had one disk report read errors, and so I did the usual pull bad disk, replace with known good disk and rebuild. This was followed by another that did the same, so I pulled it and rebuilt with another known good disk. Everything rebuilt fine and all was good.

 

I always do a pre-clear postmortem on bad disks to confirm they are unreliable. However, this time 3 cycles of preclear on each disk reported that these disks were still good. This worried me, as I wondered if the issue could then be with the cabling or HBA. And of course, this issue reared its ugly head again on the same disk slots, even though the disks were new and pre-tested (I always do 3 rounds of preclear before putting a disk into the array). So I swapped the HBA and got new cables. Everything rebuilt fine and I ran a parity check immediately after, which passed with zero errors.

 

But I wouldn't be posting here if the issue didn't persist. Today I got read errors again on the same two disks, both at the same time. No data loss (thank goodness for dual parity), but I'm at a loss on next steps for troubleshooting. Attached is the diagnostics zip file with the most recent round of disks errors captured (disks 13 and 14, sds and sdv).

 

Any suggestions on what to try next would be greatly appreciated.

 

jt-diagnostics-20210913-1333.zip

Looks like a connection/power problem, see if the disks that failed share something, like a miniSAS cable or power splitter, could also be the PSU.

Interesting. Thanks for the suggestions. Fairly certain its not the mini sas cable they share since I've replaced it, but I hadn't considered the PSU and/or power splitter. I'll have to investigate.