I am getting a red ball next to disk 4 of my 16 drive array.

 

I have checked the smart reports, but i can't see any obvious errors. I have attached the smart report here.

 

I have not written to that drive for a long time, and I suspect something happened when I just installed a new drive in the array.

 

I thought the obvious thing to do was to follow the http://lime-technology.com/wiki/index.php/Make_unRAID_Trust_the_Parity_Drive,_Avoid_Rebuilding_Parity_Unnecessarily procedure, but that doesn't seem to work on this version of unraid.

 

Any suggestions?

 

smart_sdn.txt

When you installed the new drive, you may have bumped the cables for disk4 and they may not be firmly seated.    I'd reseat (unplug - replug) both ends of the cable for that drive ... or if it's in a hot-swap cage remove it and replace it.

 

However, since it's been marked as disabled, it'll then need to be rebuilt.  Assuming your parity is good (When's the last time you checked it?  ... do NOT do a check now with a red-balled drive).  With good parity, here's what you need to do:

 

Stop the array; and un-assign that drive (disk4).  Then Start the array -- it will now show a missing drive.  Then Stop the array and assign the drive back to that slot; and finally Start the array again and it will rebuild the drive.

 

Two other thoughts ...

 

=>  As for what happened ... what most likely happened (since you said you weren't writing anything new to the drive)  is that you were reading from the drive and a read failed.  This won't "red ball" the disk;  but when a read fails UnRAID recreates the data from the other disks, and rewrites it to the disk where it failed.  If THAT write fails, the disk is red-balled.

 

=>  You're running a very old version.  Once you get this fixed, update to v5.0.4 

Thanks for the advice.

 

I tried this "Stop the array; and un-assign that drive (disk4).  Then Start the array -- it will now show a missing drive.  Then Stop the array and assign the drive back to that slot; and finally Start the array again and it will rebuild the drive.", but it doesn't work. I stop the array, unassign the drive, start the array again (drive is now unassigned). I stop the array, assign the drive and start again. It is briefly blue, but it goes red and doesn't attempt to rebuild.

 

The server is 24 bay norco build, and the other drives in the same row as this one are working ok. I tried sliding it out and back in again.

 

It's been a while since I ran a full parity check, but I now this failure must be relatively recent as I installed a new drive about a month ago and didn't have any of these issues.

 

I have attached the system log if that sheds any further light.

tower2_log.zip

I'd repeat the process, but this time use a different drive when you do the re-assignment (i.e. a new or spare drive).    It should definitely do the rebuild then.