I have had my first parity sync errors which I have noticed and not sure how to tackle it, using mostly 18tb drives so as you can imagine it takes a few days to run one parity sync.

 

Installed a new HDD on Saturday - I did not turn the machine of whilst doing this (silly I know was in a hurry), shutdown the array which in hindsight would have been better to do BEFORE installing the new HDD, formated the new drive and added it to the array, I then let the disk get cleared and started a new parity sync without the checkbox ticked to write corrections.

 

Its now 40 odd percent complete and I am showing 80 sync errors. None of my drives are showing errors and disks are passing smart tests (most of them a couple get stuck on 90% but have read this a bug with the drives in questions)

 

Depending on how large this error count grows I am not sure on what the best course of action to take is and hoping you guys can guide me to it.

 

System log is attached.

 

The drives in my system are as below:

 

 

If the number didnt grow to large I initially thought best thing to do is do another run and write the corrections this time, but upon reading further mostly on here it seems this is not the best idea since I could then be writing trash to my parity drives.

 

Not sure if better course is to shut the machine down once this run completes and do another run to see if they clear themselves.

 

It did have a hard lock up a week or so ago and I am wondering if this is where the issues began.

r3l-unraid-syslog-20240924-1422.zip

Any unclean shutdown since the last check?

The last check finished the same day I installed the disk - at the end of it the report showed 4 errors:

 

Parity-Check2024-09-21, 13:35:43 (Saturday)18 TB5 day, 4 hr, 23 min, 3 sec40.2 MB/sOK4

 

The above check was the run after its hard lock up a week prior.

 

So unsure if a disk is failing or if when adding a new drive I have hit a sata connector by accident when installing.

So there were already error in the previous check? Start by running memtest.

Should I stop the parity check I am currently running and just start memtest now?

 

Update * Cancelled it at 55% with 90 errors, going to run memtest now, after that I am going to see if my HBA card is in the x1 slot or x4 slot of my MB and ensure it has good cooling.

 

Will then run another parity check on fresh boot.

 

If any of this sounds out of order please let me know.

 

Have also checked all array disks with the disk speed plugin and no dips in the performance.

Edited September 24, 20241 yr by vindictive-voyeur2295
Spelling

Small update memtest passed its first cycle fine and I have been running it again over a few times and still 0 errors yet.

 

Not sure wheather to continue this or move onto something else.

memtest is only definitive if it finds errors, but if there aren't any obvious issues, reboot to clear the logs, run a correcting check, then a non correcting one, all without rebooting, if the second one still finds errors post new diags.

Thank you - just so I am clear:

 

1 - reboot

2 - run the parity check with the corrections box ticked

3 - run another check without the box ticked

 

All without rebooting.

 

Sorry if I sound dumb - just dont want to do the wrong thing  

and thank you for all the help.

Correct

Just a small update.

 

First correcting checked finished at about 2am and corrected 72 errors. 

 

Have been running the second non correcting sync since then, fingers crossed this comes back as 0.

 

I also must say the HBA controller card has certinaly upped the speed now I have it in a x8 slot  now getting beetwen 100 and 200 MB/sec, I thought it was originally in a x4 slot but my mobo does not seem to have an x4 slot so I removed one GPU which was in there and does not get used.

 

Also added an extra fan to the machine blowing in the HBA's direction.

 

Will update with logs once the second run finishes.

Happy weekend everyone.

 

The non correcting parity run has now finished as well and it found 0 errors claiming parity is valid.

 

I have attached the logs, please let me know what you think and what the next steps are if any.

 

I have not started any dockers or VMS since troubleshooting any of this hopefully this was the correct thing to do.

r3l-unraid-syslog-20240928-0949.zip

Since the 2nd check didn't find any more errors all should be OK now, but keep monitoring the next checks, if you get new errors without a reason, there may still be a problem.

Thank you for all your help on this.