Why is it that when one of the compute resources is reached, all resources will go 0?
For example, we cannot deploy additional VMs because there is no additional CPU but still with enough memory.
Can it also calculate how many VMs based on each resources, for example for memory how many VMs, for CPU how many VMs etc.
It seems that if one is reached, it will zero out on other resources as well.
Also, may I know how much of the total resources vREalize will calculate as reserved for additional VMs.
I've encountered one that even though all resources are 50% used, it says that no VMs can be actually deployed....
If i am reading this correctly you are using an allocation based capacity model and if so it is working as expected.
Lets use your example you have no CPU left as all is allocated but you have memory left why would you want to deploy more VMs if you have allocated all your cpu regardless of how much memory is left.
The allocation is based on what is require to be allocated to a VM not just one compute resource. The allowed you co specify a hard limit for over allocation. If you feel you could get more VMs on the hosts then change the ratios in the policy for CPU.
to your second question
The only reservation you can apply is the HA buffers. This is either a hard % of resources or using the cluster HA settings. With out seeing a screenshot it is hard to say what the problem is