Honestly, that's what I'm pretty sure it is. Not VMWare resources, but this all started happening around the time we started not being able to get LUN utilization below 90% and SAN utilization crossed the 80% boundary, so disc resources.
Nothing but issues when you don't have room for sequential writes anywhere; just wish I could prove it, or do something about it, other than slap another bandage on the thing to keep it limping along until the replacement goes live.