I' use Alan Renouf's vCheck to receive a daily cluster health report.
It reports how much overcommit I have for datastores and memory.
What is the best way to know how much datastore storage is really being used??
Multiply the datastore's LUN allocation by 1.80 if the datastore is overallocated by 80%??
Most of our VMs are thin provisioned.
Or should I just total up the actual disk sizes regardless of thin or thick provisioned??
Same idea for our hosts reporting memory overcommit. Just add up the actual specified memory allocations??
I ask because I'm trying to understand overcommit vis-a-vis sizing new storage or new servers' RAM.
I do understand that overcommit is acceptable, I'm trying to understand how much is acceptable before performance becomes an issue.
BTW very little ballooning and swap usage reported...
Thank you, Tom
