In a vSphere 6.5 cluster, the administrator notices that it is possible to set a percentage in the DRS option,
“Control CPU over-commitment in the cluster.”
Which is the purpose of this setting?
A.
It sets and enforces a maximum percentage of virtual machine CPU usage in GHz in comparison to
physical CPU availability in GHz.
B.
It sets and enforces a maximum ratio in percentage between powered on virtual machine CPUs and
available physical CPUs on hosts.
C.
It prevents the creation of any additional virtual machines in the cluster.
D.
It calculates the total shares assigned to all virtual machines in comparison to GHz of available physical
CPU.
Explanation:
https://www.starwindsoftware.com/blog/vsphere-6-5-ha-drs-improvements-part-2
CPU over-commitment: This is an option to enforce a maximum vCPU:pCPU ratios in the cluster. Once the cluster reaches this defined value, no additional VMs will be allowed to power on.
https://blogs.vmware.com/vsphere/2016/10/whats-new-in-vsphere-6-5-host-resource-management-and-operations.html
0
0
B,as per above link and below.
http://vmnomad.blogspot.com.br/2017/05/testing-CPU-overcommitment-feature.html
1
0
http://vmusketeers.com/2016/11/17/testing-vsphere-6-5-drs-cpu-over-commitment-ratio/
Well what is 0% good for?
0
0
B
I’m guessing 0% could be to say don’t bother enforcing anything… Just let me keep going. But I cannot find documentation to support that.
0
0
0% would mean no VMs would be allowed to power on, though I am not sure what would be the use case for such situation
0
0
In my opinion 0% means no overcommitment allowed or max vCPU = pCPU
0
0
Any value between 0-99 percent is under-committing (where vcpu:pcpu is less than one).
100 percent is a 1vcpu:1pcpu-core ratio
500 percent is a 5vcpu:1pcpu-core ratio
This is from student book of what’s new course.
4
0
B
0
0
B
it is a vCPU to pCPU ratio and not the usage.
http://vmusketeers.com/2016/11/17/testing-vsphere-6-5-drs-cpu-over-commitment-ratio/
0
0
B is not correct as there might be a lot more vCPUs on running VMs than pCPUs but they are not doing much, i.e. CPU utilization being low.
A is the correct answer because the described method actually measures utilization caused by vCPUs vs. the potential offered by pCPUs.
0
0
Upon further reading on this topic, apparently my reasoning above is not the meaning of this setting used by VMware. So indeed – ‘B’ is the one to pick.
0
0