I am adding some new hosts to an existing 5.5 u3 cluster. After building and adding I have them sitting in maintenance mode. To validate builds I place cluster DRS to manual and exit maintenance mode on the new hosts. I used 8 vms to test. One at a time I vmotioned from existing hosts to each new hosts. Orig host 1 to new host 1, new host 1 to new host 2, new host 2 to new host 3, etc ( I have 6 new hosts. At the end all 8 vms where running on new host 6. I then vmotioned all 8 vms to new host 5...again worked fine. I then vmotioned all 8 vms to new host 4 and then all vms lost network pings for an extended period...I did not time outage but at least a minute. The vmotion showed complete while ping was not responding. Eventually all VMs start responding at same time. Since I didn't like what occurred I wanted to vmotion these VMs back to my original hosts. So one at a time I vmotion the VMs back. Each one again had an extended ping not responding after the vmotion had completed.
I am ruling out vlans missing from hosts since individual vmotions worked and vm eventually comes back online. (Also each of the 8 vms tested were on different vlans.)
I am ruling out rarp issues since again individual vmotions worked and update the physical switches.
Any other thoughts as to what I can look into? Thanks