A. If you've been reading my last six Q&As, you've probably been getting a bit excited about VMware HA's admission control settings. These setting are ignored by far too many admins, and configuring admission control incorrectly can have a deleterious effect on cluster operations.
There's one final question that I discussed back in this Q&A \\[link to Q4\\] that I still haven't done a good job in explaining. How much of my resources get set aside when admission control is enabled.
The quantitative answer lies in the number of CPUs (and, thus, their contributions to processing) and the amount of RAM that's installed in all the servers in your ESX cluster. The cluster actually calculates real values for the number of CPUs, their contribution in GHz, and the amount of RAM in MB. The actual numbers calculated are essentially the mathematical sum of each server's resource contribution.
In English, two four-processor computers contribute eight processors total. Two servers with 16GB of RAM contribute 32GB of RAM total. By knowing these values, you can add up how much resources your servers supply to virtual machines (VMs). You can also know the deep details about how admission control calculates how many resources are available (and needed by VMs).
The calculations are different depending on which admission control policy you choose. For the specific details, I'll direct you to Duncan Eppling's fantastically-deep explanation. It's a very engaging read!
Catch up with @ConcentratdGreg on Twitter!
