I am attempting to help in a situation where the frankencluster design specification outlined by depping here was chosen to increase complexity, reduce consistency and risk availability. Multiple differing sizes of servers with differing RAM and CPU capacity have been chosen. Admission Control has been disabled in order to maintain the highest risk possible when provisioning is untracked. As a result, I need to manually figure out if all Vms can restart on all hosts given the failure of a single host, without using Admission Control, in order to try to salvage the situation. How can I do this?
↧