Don't know if anyone has seen anything like this before but...
The issues we're having seem to be occurring during periods of peak usage (over 200+ virtual desktops in use). Seemingly random desktops as they are refreshed and powered on become invalid and eventually come back with a virtual machine question posed (see attached screen shots from vSphere). If I answer the question (with either "moved" or "copied") the desktops soon show up in View Administrator as a Problem Desktop with Agent Unreachable. Usually this happens on a dozen or so desktops during the day....the only way I've found to fix them is to delete them and let them reprovision.
What could be causing this? Could it be storage related file locking issues or VAAI related errors? We've done all the obvious that were recommended by VMware and our storage vendor, VMware initially recommended Fibre Channel host firmware updates and we did those, and the storage vendor updated the firmware on the hybrid (spinning disk with SSD caching) SAN and reconfigured it with Fixed Fibre Channel pathing instead of Most Recently Used. That seemed to help performance some (occasional slow desktop logging in seemed to get much quicker), but the erroring out of desktops continues under high usage and our system should be capable of handling a lot more than 250 desktops. We have 6 hosts with a combined 208 cores and 1.44TB of RAM.
When there are less than 150 users of the system, everything behaves completely normal.
We are using CBRC on the pool.
Thanks for any insight anyone might have.
