proxmox 4.2 and cpu units

[dfgas]

On proxmox 4.2 how do we find out cpu units for our cpu? vzcpucheck is not on 4.2

 

[wolfgang]

Hi,
There is no openvz in PVE4.
What you are locking for is pct.
pct config <vmid> shows you the cpu units.
With pct set you can set it.

 

[dfgas]

Hi,
There is no openvz in PVE4.
What you are locking for is pct.
pct config <vmid> shows you the cpu units.
With pct set you can set it.

That just shows me the units I have set for that container. I am trying to find out how many units my system is capable of doing.

 

[Erza86]

I have the same question because even the wiki doesnt show a valid documentation about it - it only says use vzcpucheck - but as already said - its not available anymore and there is no alternativ documented solution.

i hope you guys can help us

regards

 

[dcsapak]

please refer to the pct documentation:

man pct

the relevant part:

-cpuunits integer (0 - 500000) (default=1024)
    CPU weight for a VM. Argument is used in the kernel fair scheduler. The larger the number is, the more CPU time this VM
    gets. Number is relative to the weights of all the other running VMs.

        Note
        You can disable fair-scheduler configuration by setting this to 0.

 

[sigxcpu]

@dfgas units are relative, as @dcsapak pasted in docs. You cannot find "how many units my system is capable of doing" because there is no such thing.

 

[Erza86]

So for example(only that im understand it right):

If i only have 1 KVM i can set 500000 or 0 (0 because as i understand it right i dont need to share with anyone)

if i have 2 i set both to 250000

and so on... (when i want that each kvm have the max possible time without effecting other kvm's)

or does the Dedicated also count as a "vm" and i need to let space?

regards

 

[sigxcpu]

I don't think you need to set anything to have a fair scheduling between tasks. You set when you need one thing to get more CPU time than the others.

For example, you have 3 VMs. First one has 1000, the others have 500. So there is a total of 2000, 50% is going to the first VM, 25% to 2nd VM, 25% to 3rd VM. This is when the system is loaded, otherwise things go as normal.