Server is restarting every 1-2 days

[mainframe.admin]

Hi,

 

I’m currently experiencing crashes on my unraid server every 1-2 days. This is a system running a 9700k set to 4.8GHz on all cores that’s been 100% stable as my main PC for the last 4 years that I have now setup as my unraid server as it was more powerful then my previous setup (Dell 7060 MFF).

 

I’m not able to physical access the system atm so i’m trying to troubleshoot via VPN. I have noticed that unlike on Windows, the CPU frequency is not locked at 4.8GHz which in the past the up and down changes can and have caused instability. The governor is set to performance so i’m not sure why it’s scaling up and down rather than staying at 4.8GHz.

 

I can’t see any obvious errors in the syslog which i’ve set to save to a folder on the array. I’m at a loss as to why this system is not stable when running Unraid but was completely stable on Windows 10 and 11 for the past few years.

 

Any suggestions?

mainframe-diagnostics-20240512-1227.zip

[JorgeB]

17 hours ago, mainframe.admin said:

The governor is set to performance so i’m not sure why it’s scaling up and down rather than staying at 4.8GHz.

This is normal, it will just scale up earlier than the other options.

 

When you say the server is restarting do you mean it's rebooting on its own?

 

 

[mainframe.admin]

2 hours ago, JorgeB said:

This is normal, it will just scale up earlier than the other options.

 

When you say the server is restarting do you mean it's rebooting on its own?

 

 

I think crashing would be a better word. Roughly every 2 days there’s an unclean shutdown and it’s restarting itself, resulting in a parity check etc. It comes back up just fine but I can’t see any issues in the syslog. 

 

I’ve disabled Turbo boost to see if it’s due to CPU instability. I’m wondering if there’s more AVX instructions going on compared to Windows/gaming and those tend to be harder on an overclocked system leading to instability. Frustrating if that’s the case as I’m away from the system for a few weeks. 

[JorgeB]

39 minutes ago, mainframe.admin said:

Roughly every 2 days there’s an unclean shutdown and it’s restarting itself,

This is still the server rebooting by itself, and a server rebooting on its own is almost always a hardware issue, but can be caused by different components, start by running memtest, but note that memtest is only definitive if it finds errors, if you have multiple sticks you can also try with just one, if the same try with a different one, that will basically rule out bad RAM, next suspects after that would be PSU, board, CPU.

[mainframe.admin]

20 minutes ago, JorgeB said:

This is still the server rebooting by itself, and a server rebooting on its own is almost always a hardware issue, but can be caused by different components, start by running memtest, but note that memtest is only definitive if it finds errors, if you have multiple sticks you can also try with just one, if the same try with a different one, that will basically rule out bad RAM, next suspects after that would be PSU, board, CPU.

This system has been running without issue so i'm inclined to think that this might just be CPU instability which is showing up on Unraid that didn't show up while being used as a gaming system running Windows. I'll try and run some tests remotely and then do some more in-depth testing when I can get to the system.

[JorgeB]

You can also try playing with C-States in the BIOS, it has been know to make a difference with some boards.

[mainframe.admin]

Well since disabling Turbo Boost the system seems stable now, been up for over 3 days without any crashes. When I get back to the system physically I will make some BIOS changes to make sure it's stable at 4.8GHz etc.

 

Thanks for your help!