We had a power outage in the middle of the night last night and today I notice that my array was started and running but one of my drives was showing a red indicator.  I have the server connected to a APS UPS with apcupsd and the powerdown script running to allow a proper shutdown when the battery gets low.  It's been quite sometime since I tested it, but I assumed it worked as expected since the array was back up and running without my intervention and no parity check was performed. 

 

I now have a drive showing a red indicator.  I checked the log and nothing jumps out at me, but I can't claim to understand most of it.  I have stopped/restarted and rebooted the array mulitple times and this drive always shows red.  I've attached a copy of my log.  Is my drive bad or is there something I can do?

syslog.txt

Attached is the SMART report for that drive.  Again, I don't understand much of it, but it seems fine, no?  Is it safe for me to perform the "Trust My Array" procedure?

smart.txt

The safest course is to rebuild the disabled drive. If you have a precleared spare ready that would be best to use. Then you can perform some more tests on the suspect drive before making it the spare or RMAing it. The SMART report looks OK. None of the values are near threshold. And there are no reallocated or pending sectors.

 

Rebuilding:

1. Stop the array.

2. Un-assign the disabled drive.

3. Start the array.  (This will make unRAID forget about the drive and treat it as new.)

4. Stop the array.

5. Replace with spare (if applicable).

6. Assign the drive.

7. Start the array and rebuild.

 

P.S. get a UPS ASAP. An APC SMART UPS is most compatible.

I don't have a spare drive right now.  This server is for my movies and TV shows so if it's down for a few it won't bother me that much.  What further test would you perform on that drive and I can I perform them with it still in the server?

I don't have a spare drive right now.  This server is for my movies and TV shows so if it's down for a few it won't bother me that much.  What further test would you perform on that drive and I can I perform them with it still in the server?

 

I would preclear it and see if it passes.

What would a preclear benefit it?  I ask because I don't know.

What would a preclear benefit it?  I ask because I don't know.

 

Preclear is designed to stress a drive. Approximately 1 out of 5 new drives fail the preclear test. The clearing and partition formatting is a side effect of the stress testing by design. I use it on all new drives regardless of whether their for unRAID. If a drive passes you can have confidence that it will not have an early failure.

I can understand that for a new drive, but hasn't my drive already been stressed (formatted once and written to many, many times)?  Could my current SMART report possibly be less indicative than one returned after a preclear on said drive?

I can understand that for a new drive, but hasn't my drive already been stressed (formatted once and written to many, many times)?  Could my current SMART report possibly be less indicative than one returned after a preclear on said drive?

 

It has been used but not stressed in terms of what preclear does. Since it has given a problem I would stress it and see how it handles it. 

OK.  So I should stop the array, unassign the drive, reboot, preclear the drive, add it back to the array, restart and rebuild the data? 

OK.  So I should stop the array, unassign the drive, reboot, preclear the drive, add it back to the array, restart and rebuild the data? 

 

1. Stop the array.

2. Un-assign the disabled drive.

3. Start the array.  (This will make unRAID forget about the drive and treat it as new.)

4. Stop the array.

5. Preclear the drive.

6. Assign the drive.

7. Start the array and rebuild.

Gotcha!  I'll do that.  Thanks!