So got my first parity check errors last night. I was working with VM stuff and the computer froze which caused me to hard shutdown the computer. Upon reboot, unRAID started doing a parity check (which i stopped and set to non correcting over night) and within the first 5% I had 74 errors. I think I remember reading that if the parity check gives you error messages early on then it is most likely the parity disk that is out of sync and not the data drives... Is that correct?

 

My plan of attack is as follows:

1. Run MD5 checks on all my data to confirm data is still good.

2. Run a short Smart test on all disks to confirm not a bad drive.

3. Check cabling.

4. Assuming all those tests pass -> run 2 correcting parity checks.

 

Let me know if anyone agrees or disagrees with my plan!

Parity check errors early in the parity check are characteristic of a dirty shutdown. RFS does a better job of handling and recovering from power outages/hard shutdowns than parity.

 

I think you're overthinking this. Run a correcting parity check and let it repair the 74 sync errors. Then run your md5 check. That will get you back into a recoverable environment the quickest and allow you too detect if a data integrity issue had been created on any source files (very unlikely).

 

Check that any recently copied files are not missing.

Hmm. that does make more sense. I just started the correcting parity checks and all 74 errors happened in the first 0.2%. Thanks for the advice and helping me take off the tin foil hat  .