My question is similar to this thread:
http://communities.vmware.com/thread/184186?start=0&tstart=0
I thought after 3 years there might be updates so trying to ask again.
My scenario is this: I am working on a final year project in my university on QoS in the cloud environment. I want to implement a resource optimization with overload avoidance algorithm. I wanted something like no additional host needs to be powered up if the current hosts still have enough capacity. So I would like to know how much is the advised maximum safe CPU and memory capacity for an ESX server? If this threshold is reached VMs will be migrated out to another host.
And one of the responders to that last thread mentioned something like a 10GB RAM host can only host 9 1GB VMs. A 10th 1GB RAM VM will not be hosted as it guarantees that all the VMs maxing out their memory will not overload the ESX host.
Any documentation for further reading will be appreciated.
