I came home last night and found I had a red balled disk. This is my first one in 3 years with Unraid, so I'm hoping to get some advice.

 

Unfortunately, I didn't grab a syslog before shutting down last night. But the drive showed up as uninstalled (via the web with the array stopped) and smartctl failed to run on the drive stating, "Smartctl: Device Read Identity Failed (not an ATA/ATAPI device)". This morning I checked the SATA and power cables and they seemed fine, but I unplugged/plugged to be sure. Fired up the server and now the drive shows up and passes a SMART check.

 

The one new (relatively) variable is that I switched SATA cards to a used AOC-SASLP-MV8 a month ago. This drive is on that card, although it shares the SAS port with 3 other drives, which have not reported any errors.

 

I'm attaching the syslog that I grabbed this morning and the output of smartctl for the drive in question.

 

I don't have a backup drive on hand and would need to buy one and pre-clear it in order to replace this drive. I think I need to do that anyway just to have one available in case of future failures, but I can't quickly swap drives right now. I'm also not convinced it's a failed drive.

 

I should also mention that I've got plenty of space to spare on other drives and this is one of the smallest & oldest drives in the array.

 

Based on the info here, what would you recommend? Help is GREATLY appreciated!

 

(I'm on 5.0-rc11 and the red balled drive is sdk)

syslog.txt

smartctl_output.txt

Smart report looks fine.

 

I expect that with vibration or something the drive lost connection and dropped off of the array.

 

The safe thing would be to replace the disk with a new disk and allow it to rebuild. But in this situation many people will just rebuild onto the existing disk.

 

Note that a drive is kicked when a write to the disk fails, and whatever that write was is likely corrupted even in parity. So any recently copied files should be verified with an md5 check if possible.

 

 

Thanks for the response. No new data has been written to this disk since Feb, so I'm not really sure what caused it to be kicked. But there shouldn't be any corrupt data to check.

 

I'm rebuilding the drive onto itself now. I also grabbed a new drive today and am pre-clearing it. I'll try this first, but if this fails then I'll definitely replace it.