Smart error on parity drive and data disk

[paperclip]

While looking in MyMain, my parity drive shows current_pending_sector=1 and one of my data drives shows multi_zone_error_rate=40

 

I've attached both smart reports to this topic. They're both 2GB WD20EARS. I'm not sure what to do. Do I need to replace the data drive, and if I do how could parity rebuild it if it's also problematic?

 

I first saw the current_pending_sector=1 a month ago and forgot about it. Since then, a monthly parity check was ran and it didn't report a problem with the parity drive, even with the current_pending_sector=1

 

I'm using unraid version 4.6. Thanks in advance for anyone's help.

parity_disk.txt

data_disk.txt

[dlandon]

A current pending sector of 1 on a 2Tb drive is not that big a problem.  i have a drive with 456 pending that will be replaced in a few days.  That many is indicative a drive is going to fail.  I'd watch the drive with the multi-zone errors.  Those are write errors on the drive.

 

If the errors increase quickly, you need to consider a replacement drive.

[paperclip]

A current pending sector of 1 on a 2Tb drive is not that big a problem.  i have a drive with 456 pending that will be replaced in a few days.  That many is indicative a drive is going to fail.  I'd watch the drive with the multi-zone errors.  Those are write errors on the drive.

 

If the errors increase quickly, you need to consider a replacement drive.

 

Thanks for the reply dlandon. Even though the current pending sector is only 1 my concern is that it's been reported on the parity drive. That is the one drive out of the array that I wouldn't want to be deteriorated. What happens if one my data drives goes bad and the parity drive has to rebuild it? Could that current pending sector value of 1 result in a failed or contaminated restoration?

[dlandon]

Probably not.  Just keep an eye on it.  The drive will generally remap a sector when it detects a bad area on the disk surface.  A few doesn't indicate a failing drive.  If remapped sectors increase rapidly, it could indicate pending head or drive electronic failure though.

[Rajahal]

Even though the current pending sector is only 1 my concern is that it's been reported on the parity drive. That is the one drive out of the array that I wouldn't want to be deteriorated.

 

Just so you know, the parity drive is actually the least important of any of the array drives since it doesn't contain any data that you actually care about.  If a drive fails (any drive - data or parity) every other drive in the array needs to be able to run long enough to rebuild the data on the failed drive.  If a data drive fails, the data on that drive is at risk until the data rebuild finishes.  If it were the parity drive that failed, then no data is at risk (but you also have no protection from another drive failure).  Does that make sense?  I'm not sure I explained that in the best way...

[SSD]

The OP is correct that there is some risk in having a pending sector (even just one) on the parity drive. The risk is quite small, but it the pending sector is actually bad, and a disk fails, the recovery will be imperfect.

 

If you run a parity check several times, the pending sector should either clear or get reallocated.

 

My advice is users start to see pending or reallocated sectors is to run daily parity checks. If you can run three in acrow without new sectors giving trouble, the drive is probably fine. Just monitor it after every parity check to make sure new sectors are not being reallocated. Normally, once the count gets beyond 1 or 2, the counts keep getting worse on every parity check.