Hi all! First time posting. Have been reading and learning a lot from everyone on here, but now is the time to get involved....
I've been having a few issues with my Unraid setup over the past month. Under full load (ie all disks spinning up and being actively read or written to), I will have random hard drives fall off the controller causing read errors on that drive and then for it to be marked as disabled. This is guaranteed to happen as part of the parity rebuild, and subsequently causes the parity rebuild to fail.
Under normal usage, then the array works absolutely fine. Yes there are some bad sectors on the disks but all disk have passed extended SMART tests. Interestingly enough, if i do extended SMART tests on all HDDs at the same time, the same issue occurs. A random HDD will disconnect from the HBA controller and reconnect on a different address, causing the SMART Test to fail. I had to do extended SMART tests on 12 of the disks, wait for that to be done, and then do the other 12 disks to avoid a disk disconnecting.
I've attached the diagnostics.
Points to note
It's only ever the Toshiba disks that this happens too.
I can swap the disks around into different slots, matters not.
Replaced older LSI 9200-8i controller with brand new LSI 9300-8i and brand new cables. No difference
1000W PSU is running all this.
shinstore-diagnostics-20201207-0947.zip
