I want to find out which VMs are creating excess IO on the CSV datastores that result in high latency. I see that latency on 10K datastores goes higher then 80 ms at times.
VM's reside on common 10K disk group datastores, so when the disk group is impacted, many VM experience problems. The problem with monitoring individual VMs with perfmon is that there are many 10K datastores with VMs that are impacted at the same time
Which tools should I use to looks at statistics at VM level ?
What metrics should I look at?