While the poor network performance on the virtual nic was happening after a P2V, the same symptoms could also be seen after a vmotion of another virtual machine between hosts. It turns out that the problem was caused by the physical switch not handling MAC addresses moving between ports well. The arp cache on the switch did not like having many virtual MAC addresses on one port let alone virtual MAC addresses moving between ports, and they only way to clear out the ARP cache was to reboot the switch.

If you are getting poor performance on your virtual machines network interface, and you have tried troubleshooting things like the list below, then consider organising an outage and rebooting the physical switch it is connected to. It could be the problem.

 Other things to troubleshoot for poor network performance include :-

• Check the patch level of the ESX host
• Check that VMware tools is installed on the VM and is the latest version
• Check the speed and duplex settings of the nics attached to your virtual switch
• Check for high CPU usage on the ESX host using the esxtop command at the service console. Virtual switch network traffic is CPU cycles on the ESX host
• Try forcing the speed and duplex settings on the switch and on the ESX host the nic is in.

Hope that solves some of your networking woes.