Parity is valid, errors found though?

[wickedathletes]

My last 2 parity checks completed but I noticed it:

 

Finding 24047 errors Duration: 1 day, 19 hours, 21 minutes, 52 seconds. Average speed: 89.7 MB/sec

 

Same error count, both checks. Any thoughts on what could be going on?

hades-diagnostics-20210121-1803.zip

[itimpi]

19 minutes ago, wickedathletes said:

My last 2 parity checks completed but I noticed it:

 

Finding 24047 errors Duration: 1 day, 19 hours, 21 minutes, 52 seconds. Average speed: 89.7 MB/sec

 

Same error count, both checks. Any thoughts on what could be going on?

hades-diagnostics-20210121-1803.zip 157.03 kB · 0 downloads

Sounds as if you have some parity errors and are running a non-correcting check so each run is reporting the same errors.

[wickedathletes]

54 minutes ago, itimpi said:

Sounds as if you have some parity errors and are running a non-correcting check so each run is reporting the same errors.

Write corrections to parity is checked. Do I need to check something else?

[trurl]

Were those scheduled parity checks? Whether scheduled checks correct parity depends on the setting in Scheduler. Usually you will want scheduled checks to not correct in case you recognize you had some problem that would have made it wrong. 

 

How long have you been running with parity out of sync? 

 

Run a correcting parity check and when it finishes correcting parity, then run a noncorrecting parity check to verify you have no sync errors. Exactly zero is the only acceptable result and until you get zero you aren't finished fixing your problem. 

[wickedathletes]

1 minute ago, trurl said:

Were those scheduled parity checks? Whether scheduled checks correct parity depends on the setting in Scheduler. Usually you will want scheduled checks to not correct in case you recognize you had some problem that would have made it wrong. 

 

How long have you been running with parity out of sync? 

 

Run a correcting parity check and when it finishes correcting parity, then run a noncorrecting parity check to verify you have no sync errors. Exactly zero is the only acceptable result and until you get zero you aren't finished fixing your problem. 

one was scheduled, the second I manually ran with "Write corrections to parity" checked. Both came back with the same number. All monthly checks have been 0 prior.

[trurl]

On mobile now so haven't looked at Diagnostics. On Main are there any I/O errors in the Errors column for any disk?

[wickedathletes]

4 minutes ago, trurl said:

How long have you been running with parity out of sync? 

 

1/2 was my last 0 errors run. 1/21 was my manually run check.

 

[wickedathletes]

2 minutes ago, trurl said:

On mobile now so haven't looked at Diagnostics. On Main are there any I/O errors in the Errors column for any disk?

nothing... that is what is confusing me. Nothing in my logs that I can see either... I am running out of space, could a space issue cause something like this? I had 2 drives down to KBs left while something was trying to move to them. I am trying to replace a drive to add some more TB's but didn't want to do anything with my parity with errors in it.

[wickedathletes]

10 minutes ago, trurl said:

On mobile now so haven't looked at Diagnostics. On Main are there any I/O errors in the Errors column for any disk?

I did pull my disk 4 some time around the 13th. Maybe its seated bad? I plan to replace it because its causing a bug in unraid, so rather than wait for a fix I figured I would swap it with another 8TB, rebuild parity, then take that 8tb drive a clear it (hopefully fix the bug) and replace a 4TB drive with that. I didn't start this because I had some work done to my home that required power outages, now that thats over of course I have parity showing issues so now I don't want to do it yet haha.

[trurl]

The check on 1/13 was a scheduled noncorrecting parity check that found sync errors (this was before reboot so can't see it in syslog). Then a manual correcting parity check on 1/21 found and corrected those same sync errors (this according to syslog in those diagnostics).

 

So, run a noncorrecting check to verify that you now have no sync errors.

[wickedathletes]

1 hour ago, trurl said:

The check on 1/13 was a scheduled noncorrecting parity check that found sync errors (this was before reboot so can't see it in syslog). Then a manual correcting parity check on 1/21 found and corrected those same sync errors (this according to syslog in those diagnostics).

 

So, run a noncorrecting check to verify that you now have no sync errors.

Running it now. How do you know if the one run on 1/13 did not have write turned on? My scheduler is set to run on the first day of the month but write corrections is on. So this must have been a crash, do those not write corrections by default?

[JorgeB]

Last check was correct, log doesn't cover the previous one, but next one should return zero errors.