Proxmox Server

Moved everything to one node, and resource values look good. I have not overcommitted anything on the RAM or CPU side of things.

The poor little Celeron CPU is grafting HARD today

1755509538175.png
 
  • Like
Reactions: OCP
PhireSide said:
Anyone have random VM reboots of guests running on their Proxmox hosts?

There is nothing in the Proxmox VE logs so it looks like a VM failure. But I am certain it’s something in particular with my setup as it happens across both my nodes (identical machines, Optiplex Thin Client 3000’s), and moving the VM from one node to another yields no difference.

Nothing in the VM logs either, which is making troubleshooting harder than it should be.
Click to expand...
Not that I've noticed. Both HA and FreePBX have been running without reboot since I upgraded to PVE 9.0.3. No stability issues on my end.
 
Mzezman said:
so could it be config on PVE2? sleep states or something else?
Click to expand...
It's happening on both nodes. Sorry if I am not making sense - I caought the flu so feeling a bit scatterbrained today.

I had them all on one node, then the rebooting issue started, so I moved one VM to another node to see if it was related to the VM, or the node. But the rebooting happens on the other node too, so it's something that is following the VM.

OR, whatever it is is just common to my setup as both nodes are identical machines. So it could be a software bug as well, but I'm coming up empty-handed as the journaltcl is not giving me any pointers and from Proxmox's perspective, all is well.
 
PhireSide said:
It's happening on both nodes. Sorry if I am not making sense - I caought the flu so feeling a bit scatterbrained today.

I had them all on one node, then the rebooting issue started, so I moved one VM to another node to see if it was related to the VM, or the node. But the rebooting happens on the other node too, so it's something that is following the VM.

OR, whatever it is is just common to my setup as both nodes are identical machines. So it could be a software bug as well, but I'm coming up empty-handed as the journaltcl is not giving me any pointers and from Proxmox's perspective, all is well.
Click to expand...
What CPU did you choose for the VM's?
 
PhireSide said:
Currently running Host type and I left the additional parameters as-is.

Each VM has 2x cores available.
Click to expand...
Changing to host type has resolved issues in past for me but clearly not your issue :-(

Do you have spare hardware you can install ProxMox on to move these 2 VM's to eliminate your current hosts?
 
OCP said:
Changing to host type has resolved issues in past for me but clearly not your issue :-(

Do you have spare hardware you can install ProxMox on to move these 2 VM's to eliminate your current hosts?
Click to expand...
Drat. I guess I could change it to another type to see if that changes anything. I've always left it as host as I reckon that is probably best for compatibility reasons.

No other hardware I'm afraid, unless I get get it running on a Pi 2B
 
PhireSide said:
Drat. I guess I could change it to another type to see if that changes anything. I've always left it as host as I reckon that is probably best for compatibility reasons.

No other hardware I'm afraid, unless I get get it running on a Pi 2B
Click to expand...
My OMV is using this as CPU type - x86-64-v2-AES but my HA is also using Host
 
Mzezman said:
My OMV is using this as CPU type - x86-64-v2-AES but my HA is also using Host
Click to expand...
Gonna give that a go as well and see what the outcome is - moving CPU type to x86-64-v2-AES-10111-802.11b-911-Y2K or whatever that type is called.

Odd that the LXC's are fine, it just seems like the VM's themselves are affected...
 
PhireSide said:
Anyone have random VM reboots of guests running on their Proxmox hosts?

There is nothing in the Proxmox VE logs so it looks like a VM failure. But I am certain it’s something in particular with my setup as it happens across both my nodes (identical machines, Optiplex Thin Client 3000’s), and moving the VM from one node to another yields no difference.

Nothing in the VM logs either, which is making troubleshooting harder than it should be.
Click to expand...
Overcommiting memory of the host? Does the VM itself die (timer in Prox reset)?
 
Are you passing through any pcie devices to the VMs?
 
Mzezman said:
Mine is setup the same as well but it’s been running fine. Shuu this is a tough one
Click to expand...
Yeah, you know it's bad if even ChatGPT gives up on you :ROFL:

The lack of log files is a weird one, as if it is doing a hard reset within the VM. And the fact that it is happening across two different machines of identical spec makes me think there is something not playing well with the Optiplex 3000's hardware.

One post suggested turning off C-states in the BIOS but that seemed to be related to PVE crashing, not the guest VMs
 
You must log in or register to reply here.
Back
Top
Sign up to the MyBroadband newsletter
X