Admission control policy is all about giving you kind of guarantee that if host failure happens then there will be enough resources to perform successful failover.
in simple terms, it's about reserving failover resources.
I would recommend going through concept of vSphere HA, Adminission control policy on following document.
https://pubs.vmware.com/vsphere-60/topic/com.vmware.ICbase/PDF/vsphere-esxi-vcenter-server-60-availa...
Starting from Page number 22, details which you need to look into is given. Specially look into Slot Size calculations part of it.
in your case you have vSphere HA with number of host failure to tolerate as admission control policy set.
that means system is going to calculate CPU and Memory slots out of available resources in your cluster, and from those total number of slots, it will keep some resources as reserved for failover and remaining are going to be available to be used. All depends upon how many hosts failure that you want to tolerate.
now in your case, if you are trying to increase reservation of one of the VM, that will then affect the number of slots in your environment, and if reserved capacity for failover is being violated than system won't let you do what you want to do. Like power on a VM, or increasing reservation of a powered on VM etc.
if you find that you have enough resources in your cluster and due to some VMs having very big cpu or memory reservation, number of slots are less than you can still manage them by using certain advance configuration parameters in vSphere HA, but I would strongly recommend, try to take help from someone who has done it before.
also try to go through other Admission control policies which are explained in the document for more inputs.
Narendra Padmani
VCIX6-DCV | VCIX7-CMA | VCI | TOGAF 9 Certified
