Hi !
Yesterday one of my drive has been disabled without me having the time to notice problems (it had been returning errors for about a week, according to the log ; unRAID disabled the disk avec 10k errors).
The first thing I did was to stop the machine, check the cables and restart the server. unRAID then tried to check the parity, found parity sync errors, and after a few hours the disk was disabled again (after exactly 1733 errors).
Here are the errors :
First, parity sync errors
Then "ata" errors :
And finally :
(hundreds of occurrences)
(a few occurrences)
I'm getting a new disk tomorrow, but I'm worried about those parity sync errors. From what I could understand, there are two possible scenarios:
1) It's the disk that has those errors, in that case when rebuilding on a new disk I won't loose data
2) It's the parity disk that has errors, in that case if I rebuild on a new disk, data corruption will ensue.
Is that correct ?
I'm pretty sure that the faulty disk isn't completely dead, and could be plugged into a computer to recover some files. In the event of scenario #2 happening in a few days (after rebuilding), how can I compare data from the "old" disk with data from the "new, rebuilt" disk and replace corrupt files if needed?
(I'm thinking "recursive MD5 check" for instance, but don't know where to start).
Thank you!
Edit : adding attachment "syslog"
syslog.txt