This is probably a somewhat nebulous question. . .
We are seeing some network latency related to our virtual machines in a four-host ESX 3 cluster. Pings from physical systems to VMs are slower than we would expect and pings between virtual machines are also slower than we would think. Ultimately, we noticed this as part of troubleshooting some application/database performance issues we are seeing with a couple VMs.
Configuration of the physical esx hosts looks like this:
vswitch0 --> 2 fibre, gigabit NICs configured as a team
The number of VMs per switch varies depending on the esx host:
host #1: 13 VMs
host #2: 33 VMs
host #3: 29 VMs
host #4: 42 VMs
Has anyone seen latency like this?
One configuraiton question I have: on the port group, the load balancing "strategy" is set for IP hash, on the vSwitch itself, it is set to Port ID. Does that mismatch matter?
Any help is greatly appreciated
Jeff