Hi Folks,
i am studying towards VCP-DCV 550 and have to following questions where I am not entirely sure:
1,
I had a question regarding CPU reservations for a VM:
When I set for example a CPU reservation of 1GHz on a multi-vCPU VM, does that mean a 1GHz (1 billion per second) CPU cycles reservations for a total to the VM
or does that mean 1GHz per vCPU?
2,
I had another question regarding ressource pools and reservations:
Does the reservation setting on the ressource pool only control the total ammount of reservations the VMs/ and direct child ressource pools can have.
OR does the remaining, not used up CPU reservation on the ressource pool grant CPU ressources on top of the reservations for VMs if they need it?
Example:
ESXi (8Cores @ 2GHz) with two ressource pools:
Pool1: reservation: 10GHz, Shares: Custom: 1 share (for simplicity lets calculate with 0 shares, but the minimum I could set is 1 Share), NON-exapndable
Pool1_VM1: 8x vCPU, reservation 1GHz
Pool 2: reservation: 0, Shares: high (8000), NON-expandable
Pool2_VM1: 8x vCPU, reservation 0GHz
(i know that its not best practice to have 8vCPU VMs on an octa core host, its for the sake of the example)
Now what happens if booth VMs are (trying to be) busy?
How much CPU cycles will Pool1_VM1 get from the total?
3,
Ist there a link to an ISO image which contains some CPU busy script like we used at the ICM/fast track course? I would like to experiment with contention in my home lab..
Thanks for the answers in advance!
Best Regards
Janos
