Where to start with parity read errors?

[lovingHDTV]

Yesterday my server lost power and it had an unclean shutdown.

 

I the past I have always run pairty check with auto correct turned on. Recenlty I disabled that as I figured I wasn't able to see if I ever had errors.  I never saw a report that says xxx errors corrected.

 

The parity run that automatically kicked off when I restarted the server has reported a ton of errors.

Current operation started on Sat 01 Jan 2022 02:03:48 PM EST (yesterday)
Finding 68180805 errors
 Elapsed time: 21 hours, 48 minutes
 Estimated finish: 5 hours, 31 minutes

 

I had successfully ran a parity check a week or so ago without error, so something happened when the server lost power.

 

Ideas on where I can start looking to try and figure out what is happening?

 

My disk smart data is fairly clean I only have a single drive with 8 allocated sectors.  I do see messaging in syslog:

Jan  2 08:42:13 tower kernel: eth0: renamed from veth756310a
Jan  2 08:50:09 tower kernel: veth756310a: renamed from eth0
Jan  2 08:50:23 tower kernel: eth0: renamed from veth71571d0
Jan  2 09:35:12 tower emhttpd: spinning down /dev/sdj
Jan  2 09:51:27 tower kernel: md: recovery thread: P incorrect, sector=19532874280
Jan  2 09:51:27 tower kernel: md: recovery thread: P incorrect, sector=19532874288
Jan  2 09:51:27 tower kernel: md: recovery thread: P incorrect, sector=19532874296
Jan  2 09:51:27 tower kernel: md: recovery thread: P incorrect, sector=19532874304
Jan  2 09:51:27 tower kernel: md: recovery thread: P incorrect, sector=19532874312
Jan  2 09:51:27 tower kernel: md: recovery thread: P incorrect, sector=19532874320
....
Jan  2 09:51:27 tower kernel: md: recovery thread: P incorrect, sector=19532876584
Jan  2 09:51:27 tower kernel: md: recovery thread: P incorrect, sector=19532877352
Jan  2 09:51:27 tower kernel: md: recovery thread: stopped logging
Jan  2 10:18:48 tower emhttpd: spinning down /dev/sdi
Jan  2 10:18:58 tower emhttpd: spinning down /dev/sdh
Jan  2 10:18:58 tower emhttpd: spinning down /dev/sdf
Jan  2 10:19:01 tower emhttpd: read SMART /dev/sdh
J

 

thanks,

david

 

tower-diagnostics-20220102-1144.zip

[JorgeB]

Sync errors after an unclean shutdown are normal, just run a correcting check.

[itimpi]

19 hours ago, lovingHDTV said:

I the past I have always run pairty check with auto correct turned on.

It is recommended that the automatic parity checks are set to be non-correcting so a drive that is misbehaving (that you have not yet noticed) will not end up corrupting parity and thus prejudicing your chance of recovering from drive failures without data loss.   You then only use correcting parity checks that are started manually and when you do not believe you have any problem drives but may have some parity errors that need correcting (e.g. after an unclean shutdown).

[lovingHDTV]

Thanks, kicked of a correcting check.