Random Parity checks

[korey_sed]

Please help!!  Unraid is driving me crazy and it was working so well until a few months ago. 

 

It is running parity checks out of schedule and no matter what I do, it does not seem to stop doing it.  At this rate, my drives will go bad before their time because of this.

 

I have extended the shutdown timer, had clean reboots (no parity starting right after reboot), have completed parity checks successfully, and yet Unraid will start a parity check randomly.  If I cancel it, it will just start it again in a few days.  if I pause it, it will start it again in a few days. if I disable it, then i keep getting notifications that I do not have a parity check scheduled.

 

This is not a scheduling issue, or should not be, since it is telling me when the next one is scheduled for, and yet it will start one before that in the middle of the day and all the services that I am running will become unresponsive.

 

Where is the config file to see the schedule?

What in the logs will tell me why it is starting a random parity check?

 

I do randomly get a mcelog event, and I am trying to isolate if one of my RAM modules is bad but my understanding is if I shut it down and start it up cleanly, it should stick to schedule.  is that not true?  I did share my diagnostics a little while ago in a different post and the best i got was to check my RAM.

 

I am on 6.12.10, and here is my drive config (attached).

 

Thanks in advance.

[bmartino1]

Please post diag file...

if you are getting mcelog events please reboot unraid and run memtest.

You may have a bad ram chip.

[korey_sed]

Thanks for the response.    The diag files are in the post I linked above.  I am switching out RAM modules back and forth to see if I can figure out which module may be the issue.  Was not sure if mem test would tell me which module was bad or just that i have an issue.

 

So there is no way to see why parity check was started?

[Frank1940]

You really need to post up your diagnostics in THIS Thread  in your next post in this thread!  (Going back and editing a earlier post to add info is confusing to the folks who are trying to help you.)  And starting a second thread is not the way to get a problem solved--- Although, it does not seem that these two problems are related...  The diagnostics should cover the time period when the Parity check started.  You can look at the time span for the syslog in the diagnostics file to determine that the start of the check is there.  It would also be helpful if you could provide the approximate date and time.

[korey_sed]

I have just switched my RAM modules again so waiting for it to do its thing and I will post diagnostics on the next mce log event or parity check.

[korey_sed]

I had both the disk check and MCE log event happen yesterday again.  diagnostics attached.

 

Any help is appreciated.

unraid-diagnostics-20240810-1313 2.zip

[JorgeB]

Aug 10 13:02:06 unraid emhttpd: unclean shutdown detected

 

After an unclean shutdown there's always a parity check, I see macvlan call traces in your log, those may be crashing the server, update to 6.12.11

[korey_sed]

Update to 6.12.11 complete.  Macvlan workaround in place (not ipvlan).  Memory tested with Live Memory Tester with no errors. 

Lets see what happens in next couple of days.

[korey_sed]

Looks like something is crashing my unraid.  I woke up to another parity check and now the system shows unclean shutdown again and the logs are from 11 PM last night post whatever caused the crash.

 

I've set the syslog to mirror to flash for now in the hopes i can catch this the next time.  Any other way to do that?

[JorgeB]

7 minutes ago, korey_sed said:

I woke up to another parity check

That suggests the server is rebooting on its own, and that is almost always a hardware problem.

[korey_sed]

Thanks.  anyway I can narrow this down to what could be doing it.  Unless all of my RAM modules are bad, I doubt it is that.  I have switched both sets out and this still happens.

 

I started pushing the system logs to flash to see if i can capture the error before the crash but not seeing anything there.  Just a constant log message that it cannot join my AD

Edited August 21 by korey_sed

[korey_sed]

Posting an update here for anyone else who runs into this.  

 

I contacted UNraid support and they mentioned that it was the CPU that was the issue.  So, I checked for a motherboard bios update and after that I have not seen the issue happen again. Phew.

 

Thank you Unraid Support.