Host: MacPro6,1, latest efi and smc firmwares. 12cores, 24logical processsors, 32gb of ram, 1x1tb SSD.
ESXi: 6.5u1, October updates.
Guests: Total 4: All macOS 10.12.6 or 10.13 w/ 2-4cores and 5gb of memory. 100GB thick provisioned, lazy zeroed vmdks.
If I take a snapshot of any of the vm's, when the system with the snapshot is in use, the rest of the systems constantly(every few seconds for multiple seconds) become unresponsive. The host becomes unresponsive at the terminal via SSH and in the web interface as well.
This system worked fine performance wise for the last three years using ESXi6.0. We moved to 6.5 last week to support the latest macOS release, after which this performance issue appeared.
Does this issue ring any bells with anyone? What am I missing?