6.12 mover crashing

[dopeytree]

Have set shares to backup cache to array. There's plenty of space. Mover works then crashes. Nothing in syslog.

If I reboot it causes a parity check.

 

What should I do?

moulin-rouge-diagnostics-20230616-2023.zip

 

Mover 'running'

 

No large disk activity (indicating mover has stopped or crashed)

Edited June 16, 2023 by dopeytree

[w0lv3r1n w0rm]

I'm having similar issues here. Nothing appears to be reported in the syslog of mover failure.

 

I've noticed if I start the mover in terminal it will run until it reaches a specific file and then seems to crash. If I stop the mover and manually delete that file it will run for a few files then crash again.

 

Any thoughts?

gyarados-diagnostics-20230624-1207.zip

[dopeytree]

I've not attempted this since posting.

Have updated to 6.12.1

System will attempt a weekly move tomorrow.

 

Someone mentioned using the command 'move stop' or similar if it appears to stall again.

Edited June 24, 2023 by dopeytree

[Craig Dennis]

My server has been crashing consistently in the middle of the night (full reboot) and Mover is the only thing scheduled to run at that time. The last few log entries are related to Mover and Parity Check Tuning. I have removed that plugin and will see if the crashes persist. No errors in syslog. Also ordered a new PSU to eliminate that. I don't really want to run Memtest but I guess I'll have to if none of these work.

[dopeytree]

This stopped crashing with the new updates. Am currently on 6.12.2

[Craig Dennis]

Good to know. I'm also on 6.12.2 so may not be the same issue.

I have noticed a new parity check accumulation feature (I don't know when this was added) which may have been conflicting with the Parity Check Tuning plugin.

 

I only recently enabled Mover logging so we'll see. I've manually initiated Mover now and have also changed the scheduled time to see if the crashes also change. It could also be hardware. With nothing in logs it's difficult to tell. Constant hard reboots since 6.12 i915 issues could have damaged something.

[itimpi]

Just now, Craig Dennis said:

Good to know. I'm also on 6.12.2 so may not be the same issue.

I have noticed a new parity check accumulation feature (I don't know when this was added) which may have been conflicting with the Parity Check Tuning plugin.

 

I only recently enabled Mover logging so we'll see. I've manually initiated Mover now and have also changed the scheduled time to see if the crashes also change. It could also be hardware. With nothing in logs it's difficult to tell. Constant hard reboots since 6.12 i915 issues could have damaged something.

The cumulative check was added some time ago in the 6.11.x series of releases

 

The parity check tuning plugin will warn you when you go the settings page that you should use either it’s version of running in increments or the one now built into Unraid.   The plugin’s version is more comprehensive and flexible so I would think going with it is the better approach  

[Admin9705]

Hey Team, I am having the same issue. I disabled mover and all my server stopped freezing. I setup a syslog, but the syslog does not capture anything went it freezes. This problem was in 6.12.0/1/2(latest).  When it happens, cannot connect to server and everything is unresponsive. I am not using parity at all, so know this cannot be the issue.

Edited July 7, 2023 by Admin9705

[dopeytree]

I take back my words.

 

Something else happened today which is mover didn't stop during a normal shutdown event. I pressed the power button that instigates the shutdown sequence and mover didn't stop properly causes it to want to do a parity check on power on.

 

I assumed 'mover stop' would be written into the shutdown sequence?

 

If my 4x cache drives wasn't already as ZFS I would downgrade to 6.11

 

Hope you find a fix soon.

[bucky2076]

 same thing is happening to me... Mover is causing the server to crash... even check parity is not working properly. I am going to put the server in maintnenace mode and check to see if there is courruption in the file system. I am on 6.12.3.

It took me a while to get syslog working to my flash drive. Once I did this, I could see mover was what was causing the problem.

Not sure how to proceed... 

 

Edited August 21, 2023 by bucky2076

[bucky2076]

further to this... I noticed system would crash when I attempted to rebuild parity... So I have shelved parity for the time being.

Conclusion is that the power supply is not able to handle the load at startup of parity check...  based on further readings around this forum.

My system is well past its use by date, and so I've ordered MB, CPU  Memory, and new Power Supply. I am going to rebuild my system to something I can be a little proud of.

I originally assembled my server from used PC parts... which > 10 years ago were at the forefront of their time. Cest la vie.

 

[Craig Dennis]

Mine was related to overheating. Mover caused NVMe and HDDs to run hot and overheat the CPU causing a shutdown. Bios meant it would start up again. Solved for me. I used NetData to track all temps; some were _much_ higher than Unraid Temp addon was reporting.