So I woke up this morning to a failed drive, which obviously I'll have to replace right away.  What's the best thing to do regarding preclear?  Normally I run a couple of preclear cycles to shake out a new disk before adding it to the array. With a failed drive, though, I'm tempted to just stick in a new drive and get the array rebuilt before anything else has a chance to fail.

 

What's best practice here?

 

(Note to self for next time - start looking for a cheap second system I can use to preclear drives apart from my "real" unraid. )

So I woke up this morning to a failed drive, which obviously I'll have to replace right away.  What's the best thing to do regarding preclear?  Normally I run a couple of preclear cycles to shake out a new disk before adding it to the array. With a failed drive, though, I'm tempted to just stick in a new drive and get the array rebuilt before anything else has a chance to fail.

 

What's best practice here?

 

(Note to self for next time - start looking for a cheap second system I can use to preclear drives apart from my "real" unraid. )

You should really still run the pre-clear commands if the replacements are new disks, although you can skip it if they are disks you have previously used and are confident that they are not faulty.  It is not a pre-requisite that disks to be rebuilt are pre-cleared but it is a sensible confidence check.

 

You mention a failed drive - do you mean it has been 'red-balled' by unRAID or that it really has failed.  Just checking as a red-balled disk just means a write failed and this can often be for external reasons (loose connection, power glitch etc).  The recovery action is the same but if it is simply the red-ball case then after completing any recovery you might want to put the problem disk through a pre-clear cycle to see what the results are before you decide to RMA the drive.

Thanks for the quick reply.  Yes, it's just a red ball (with quite a few write failures in the log).  No noise from the drive suggesting a head crash, no obvious loose connections, no power glitch (says the UPS).  You're right, there's hope for the drive itself, but I'll figure that out when the immediate crisis is passed.

 

I guess what I can do is disconnect all my drives so I'm left with little more than a Linux box, use it to preclear a new drive, then reconnect it all back with the new drive in the proper place.  A little paranoid, maybe, but I just don't want to take any chance that a second drive might fail while the replacement is going through preclear.

 

Thanks!

I guess what I can do is disconnect all my drives so I'm left with little more than a Linux box, use it to preclear a new drive, then reconnect it all back with the new drive in the proper place.  A little paranoid, maybe, but I just don't want to take any chance that a second drive might fail while the replacement is going through preclear.

That is definitely the safest thing to do.  You might want to put a clean unRAID install on another USB stick while doing this and boot off that as it will then not have any drives configured.

 

Keep the 'red-balled' disk somewhere safe while doing recovery.  If by any chance anything goes wrong during recovery then as long as the 'red-balled' drive has not physically failed it is highly likely you could recover virtually all of the data off it.