We seem to be having some network issues and whle investigating i notices that some servers are showing that the NIC's are connected at 10GB and our switches aren't capable of that. The NIC on the Op system is set to AUTO NEGOTIATE and when I check teh status it shows connected at 10GB. First, is this possible or is it a flaw. All the other server VM's that have a E1000 NIC installed show the correct status and are also set to AUTO NEGOTIATE. I am curious if this mght be part of the reason that we are experiencing some issues. If anyopne has any thoughts or comments i would appreciate it. Thanks.
It is possible as the virtual switch is indeed 10gb. When traffic leaves your virtual switch out to a physical port, then the traffic would slow to whatever is available at that point.
It's good to have your guests running as 10gb nics even though your host may only have 1gb nics because they can achieve higher guest to guest transfer rates. Also, the vmxnet3 driver performs better under load than the intel compatability driver.
Thanks for teh explination. That would explain why all teh VMXNET3 nics are at 10GB. That being said, there is not a link compatibilty issue when it comes to the networK? I know it's best to have same settings on both ends of teh connection and wasn't sure if this was an issue.
