Looking for some advice on areas to look @. I have the following:
If I look into the spike, I don't see much write or read latency back to disk from this vm:
latency is relatively low
but if I look @ I/O wait for that same period seems high, just trying to understand what else I might want to look into? seems that IO wait could become an issue. Should mention this is a 2008R2 windows box, running SQL server. LSI logic for storage controller
IO wait is basically cpu.wait - cpu.idle - cpu.swapwait which means something in the VM world is causing it. if its not storage, or networking, then you need to look at other things that could be attached to the VM, pci devices, usb devises and believe it or not one of the big offender sis if you leave an ISO attached to the virtual cd drive can cause high wait times.
If everything looks fine and the wait times are impacting performance then you can cause it to crash to get dump data but i recommend talking to VMware support if this becomes an option.
I'm a fan on CPU READY % as it's universal metric typically. Can get in to WAIT, etc after you do an initial problem determination using READY %.