[one-users] Some questions about overcommitment in OpenNebula 4.2 using KVM

[Gerry O'Brien]

On 16/10/2013 22:32, Ruben S. Montero wrote:
>>   libvirt uses cgroups, so yes, it does.
>> cgroups do not seem to be enabled with kvm. Am I correct in assuming this?
>
> Yes it is configured by default. You have to enable it in the hypervisors,
> though
>
>
> http://opennebula.org/documentation:rel4.2:kvmg#working_with_cgroups_optional

Ok, as neither /etc/cgconfig.conf nor /etc/cgrules.conf exist on the 
hosts, the CPU parameter is not used to limit execution but only 
deployment. My understanding that in this scenario setting a small CPU 
value allows for cpu overcommitment without having to modify kvm.rb. 
This also allows for creating cpu hungry machines by setting CPU and 
VCPU equal.
>
>
>>
>> Hi,
>>
>> This didn't work for me. I'm running 4.2. The scheduler has a capacity
>> test (see below). Any idea where the code for "get_requirements" lives?
>>
>>
> Note that you have to do onehost sync to propagate the changes (or copy
> manually the new probes). And then wait for the hosts to be monitored. This
> approach should work without modifying the scheduler.
>


-- 
Gerry O'Brien

Systems Manager
School of Computer Science and Statistics
Trinity College Dublin
Dublin 2
IRELAND

00 353 1 896 1341