System crashes

[Die_piggy]

Hi Guys,

 

I have been using unraid for 5-10 years (since 4.something), and have been collecting hdd and the like over that time. My array has expanded, I have completely replaced the hardware for new equipment and I have rebuilt my array from the ground up a number of times (reinstalling all my plugins and more recently dockers from the ground up.)

 

Over the last month I have been getting more and more system crashes where I have had to restart the server from mains power. These crashes usually come from periods of high usage such as watching a plex stream, or running a parity check. Its to the point where when I get the system running I usually cancel the parity check, as it causes a system crash itself usually about 5-10% in.

 

I have done a little bit of investigating, but I know I am out of my depth. One hard drive (disk 7) is throwing up an amber alert, (UDMA CRC error count: 3) but I am loathe to try and replace it until I have system stability so I can rebuild. It's a 4TB drive, so I would upgrade it at the same time, most likely to an 8TB one.

 

My unraid is currently a 42TB server with 97% utilization. it however has 1.3TB of free space still on it. System currently has 3x2TB HDDs, 5x4TB drives and 2x8Tb drives. I have currently been replacing the 2TB drives as I needed space, and that's why I have an 8TB drive currently available.

 

The fullest drive currently has 115GB free, so I don't think the issues would be space related?

 

CPU: Intel i3 3220T @2.8GHz

Ram: 12GB

HDD's: 10+parity+cache

Total Array Size: 42TB

 

I haven't been able to get a diagnostic from when it has crashed...as I am not even able to telnet into the system at that point. I am wondering if my actual hardware has started to wear out and I should look at upgrading the mobo/cpu and ram?

 

I have attached a fresh diagnostics file as well as screenshots of what is currently running

 

Does anyone have any suggestions for me?

 

Die Piggy

wotbox-diagnostics-20181125-1543.zip

[Die_piggy]

The server just stayed up for about 2 hours. I got this diagnostic abwotbox-diagnostics-20181125-1543.zipout 5 minutes before it crashed

[Die_piggy]

Noone has any idea's?

[JorgeB]

This type of issues are very difficult to diagnose remotely, diags are clean, if crashes are mostly during high server usage I would start by checking temps and if you have one trying with a different power supply.

[Die_piggy]

Thanks mate. Is it likely to be power if its a recent thing?

 

[JorgeB]

It can be if there's a problem with the PSU, and that can happen at any time.