by pufferdude » Wed Dec 12, 2012 9:36 pm people like this post
I have Monitor set up and running great. Once in a while I get an alert that the read or white latency on a specific datastore and/or host has exceeded the default 100ms threshold (our rickety SAN is being overtaxed and will soon been replaced with a newer model)... but I can't seem to figure out how to use that info to find out what *VMs* on the identified host are actually causing the excessive I/Os and increasing my latency.
I realize it's possible that "a bunch of VMs all did something at the same time" and thus the entire I/O capacity of the storage is being exceeded without a specific VM being the culprit, but is there a way to know for sure? Can someone point me in the right direction on how to investigate "what" is causing the latency beyond which host and datastore I get alerted-to?