I've got a standalone vSphere ESXi 4 host, with 2 VMs - they both talk to other devices out in my organization. The switches that the host uplinks to are set at1000-Full, and I've configured my ESXi host's physical NICs to run at 100-FULL per this article (http://kb.vmware.com/selfservice/microsites/search.do?cmd=displayKC&docType=kc&externalId=1004089&sliceId=1&docTypeID=DT_KB_1_1&dialogID=84030079&stateId=0%200%2097267496). Do I also need to configure the NIC drivers inside my VM's (Windows 2003/2003R2, x32) to run at 100-FULL?
Do I also need to configure the NIC drivers inside my VM's (Windows 2003/2003R2, x32) to run at 100-FULL?
The NIC speed inside the VM's are just informational, not the TRUE speed. Even if they show 10 MEG they will get whatever the speed of the Switch shows.
The standalone host is actually part of a blade environment, and unfortunately the pass-thru card on the back of the enclosure is hard-coded at 1000-full (hence, my need to hard-code my physical NICs at 1000-Full). I know that Vmware recommends "auto" whenever possible for gigabit Ethernet, over 100-full, so I'm wondering if that might be the causing some of the packet loss I've been seeing.