VMware Cloud Community
ajnrock
Enthusiast
Enthusiast

Odd Problem with Virtual NIC / Virtual Switch

I have a VM that has been stable for about two months and the stragest thing just happened. It became unreachable from the network, however from the console I could still ping out. The only clue I have at the moment is a useless error in the event log.

Event ID 3019 MRxSmb The redirector failed to determine the connection type

The only thing I had to do to recover is reset the NIC using the Windows NIC "Repair" Option to restet the NIC and all returned to normal. No other VMs on the same virtual switch were impacted.

Host ESX 3.0.1

VM W2K3 64 Bit w VMWare Tools running.

Has anyone encountered this?

Reply
0 Kudos
5 Replies
kitarra
Enthusiast
Enthusiast

I have sorta seen this. It is possible that the MAC that the server was using got recycled. Thus it detected a MAC address conflict and the server dropped the connection. I suspect that is the most likely reason. Especially since a repair fixed it.

Reply
0 Kudos
bggb29
Expert
Expert

Not all that uncommon with a windows box to just stop sending and receiving packets.

I do not think it was esx realated at all

Reply
0 Kudos
ajnrock
Enthusiast
Enthusiast

Can't really disagree, but seemed like it could be virtual switch, nic driver, or maybe VMWTools related somehow.

Reply
0 Kudos
jeffslapp
Enthusiast
Enthusiast

I am experiencing this issue with a couple of my 64-bit servers. The only solution I have right now is to disable the NIC from within the VM OS and then reenabled it. I am submitting a support request for this today. I will post back any findings.

Reply
0 Kudos
ajnrock
Enthusiast
Enthusiast

Just curious, How much RAM do your 64 Bit VMs have and how much available kernal memory was available when the NIC stopped responding?

I caught it in the act two nights ago and kernal memory and available memory were both very depleted so I increased from 1 to 2 GB RAM to see what happens.

Reply
0 Kudos