Diagnostics attached, but I haven't scoured through them. I logged in today and unraid gave me the error that an MCE was triggered and I should post here. I'm not seeing any problems with my system for the moment, but a little more than 10 days ago my system was just constantly rebooting after I started the array. System would run fine without array started, but as soon as I started it I'd have something happen and the system would hard reboot. I disabled everything I could find to see what might be happening and finally disabled the UPS integration and the system appears to be stable again. I disabled VMs and Docker after that and allowed my parity drive size increase to finish and then I started re-enabling services and have everything but the UPS integration turned back on. I was planning on adding to my array with a couple more drives this weekend and logged in to see this error.

 

Can anyone take a look or help me know how to read the diagnostics so I can look myself and figure out why I'm getting this MCE error and know what hardware I might need to replace (that may be the original reason for the reboots)? I will note that I'm running out of disk space on my array which is why I upgraded my parity to start adding larger drives into the mix but I'm not sure if that's a reason for an MCE to trigger.

 

Any assistance or direction on this would be helpful. Thanks in advance and apologies if I'm not sharing something I might need to help diagnose. Please ask if there is more information needed and I'll do my best to provide.

 

Unraid Pro Version 6.12.8

Curernt Uptime 10 days 3 hours 25 minutes

614GB Free of 13.4TB Array (3 Data, 1 Parity)

 

tower-diagnostics-20240512-2014.zip

Edited May 13, 20242 yr by nobrakes

do you have amd board?
90% of the time its was a ecc memory correction.

It recommend running a memory test.

otherwise in unraid terminal type mcelog

mcelog error on amd its a known issue and will be fixed in unraid 6.13

 

It's an intel board - supermicro X8DT3 with 2 Intel Xeon X5675 Processors - I have ECC mem, but I will run a memtest next as soon as I figure out how (lol). Typing mcelog on the terminal gives me nothing - should I be seeing an output or does it just save a file somewhere I should go find?

then you reboot the machine since that error? Then mcelog would be empty on intell if there was a machine check event by the processor then it would show up under mcelog.

Since its blank there I no issues as yet.

to run mem test at unraid boot at the grub menu chose the last option. tnhis will run a memeroy test.

 

I would also recommend installing the plugin FCP Fix common problems. as it will scan your logs and give you useful info in event of error or misconfigurations.

Fix common problems is what told me I had MCE errors, but I have not rebooted in 10 days so I will ignore the error for now, run the memtest on next reboot (probably later tonight) if I get nothing in the test I'll add my drives and go from there. Thank you for the help and direction, I really appreciate it.

Still haven't figured anything out. The system is suddenly stable. I disabled my UPS integration, finished my array rebuild for a larger Parity and added 2 drives, then turned the UPS back on. It's been running fine ever since, no reason why it failed, and it hasn't failed again since. I'll continue to monitor, but thanks @bmartino1 for your suggestions. I appreciate the help.