Write errors during array expansion/rebuild

 

Hey guys, first let me say that I've been an unraid user for several years, and I've had to visit the forum on numerous occasions, and have usually found solutions/similar problems to my issues (thank you all).

 

But this time i haven't found an identical issue.

 

I am running Unraid 6.0-rc3, in my Norco 24bay chassis (with some off the shelf Intel/asus cpu/motherboard)

 

Here is what has happened so far.

 

I decided to swap out disk17 in my unraid to upgrade its size, i was going from a 6tb to an 8tb drive. The 8tb drive i had pre-cleared ~6times or so without any issues.

 

I stopped the array as i normally would, i removed disk17, waited ~30secs, installed the new 8tb disk into the disk17 position, waited ~30secs for the drive to be detected.

 

Selected the new 8tb disk in the unraid UI (in the disk17 position), and selected 'Rebuild/Expand' and started the array.

 

The array began rebuilding, and the following morning (Today) the rebuild had completed, but instead of all of the disks having a 'green ball' next to it, Disk17 had a 'gray triangle'. Checking the syslog, i found a bunch of 'Write Errors'.

 

I stopped the array, reseated disk17 in the array, and rebooted unraid.

 

Unraid started, but disk17 still has a 'gray triangle' on it - and in my dashboard view, i have a 'Data is Invalid' warning at the bottom.

 

What should i try next?

 

Attached is my syslog and a couple screenshots

syslog.zip

I prefer more complete info when analyzing issues, but even from that piece of the syslog I can say a few things.  It starts with a Call Trace that appears to be associated with the removal of a drive from a SAS card.  Apparently the card didn't like that!  While hot-plugging is *supposed* to work, if you have the right hardware, I don't always trust it, and in your case, there was an issue, of unknown consequences.  I can't say for sure that that set the stage for the failure later, but it's suspicious.

 

You plugged the new 8TB drive in, and it initialized OK.  You assigned it to Disk 17 and started the array, rebuilding Disk 17, no issues so far.  But at May 18 04:33:11 in the morning, the SAS card appears to fail, dropping all drives.  The drive being accessed was the new one, being written to, so it is the one that reports write failures and is dropped from the array.  The kernel wastes no time giving up on the drive, disabling it in less than 30 seconds.  In this situation, the drive(s) are usually completely fine, just an issue with the controller.

 

This is something we used to see more often, haven't seen recently.  You are on 6.0-rc3!  After rebooting, I would head to the Plugins page first(!), and update unRAID to the latest, 6.1.9.  Reboot and try rebuilding Disk 17 again.  You may also want to check for a firmware update for the SAS card, before rebuilding.

Thanks for the advice, i upgraded to the latest version, and i've been able to swap out 3 disks so far with larger disks, and i received no errors on rebuild. Maybe it was a funny driver version for my sas cards in that specific version of unraid.

 

Thanks again!