I just got home and noticed my 100TB unRAID server has been doing a parity check and one of my drives has a red X and unRAID is reporting 5 sync errors so far in the parity check. It has been running for 15 hours and says it has another 11 to go.

 

1. Do I let it finish and then address the drive that is offline?

2. Do I stop the parity check now and address the drive that is offline?

 

I am running the 6.2b with two parity drives, so I am ok in terms of potential data loss.

 

Suggestions?

Since you have dual parity it's up to you, if you're running a correcting parity check and have full checksums do a check after the rebuild as there could be some corrupt files.

How do I do a full checksum?

If you don't already have checksums, you can't do anything to check them.

 

If you want to create checksums for all your files, do this AFTER you've resolved your failed drive.

 

First, let the parity check finish.    If this was not a correcting check, you'll need to run it again to fix the errors before you can rebuild your drive.

 

Once you have good parity; then you can replace the failed drive.

 

After that, you can create checksums for all your files.  I use Corz from a Windows box to create mine; a lot of folks like to use the Dynamix File Integrity plugin.  Note that either of these will take a VERY long time to initially build checksums for all the data on a 100TB server !!    Checking the full server will also take a long time; but in general you'll only need to check the files on a drive that's been problematic [e.g. in this case you'd want to check the rebuilt drive just to be sure none of the sync errors were in the actual data).