I finally took the plunge today and upgraded my first ESX 3.5 cluster node to vSphere 4.1. Installation went pretty much without a hitch as is my usual experience with VMware products, however once I got the new 4.1 node up and back into my cluster I noticed a warning telling me that “Host X currently has no management network redundancy”. I thought that was odd, since I have three physical NICs assigned to the vSwitch connected to my service console. Turns out the solution is quite simple.

When you first install ESX 4.1 you are asked which NIC you want to use for your management from the ones available in your server. This is the initial NIC that will be connected to your network to allow you initial access the server from vCenter and continue configuring, adding to the cluster, configuring networking etc.

When you add the new host to the cluster in vCenter it automatically sets it up for HA if your cluster has that enabled, and it does this will the current networking configuration, ie, one pNIC. If you then add the other pNICs in your server to the vSwitch containing the service console, the HA configuration doesn’t initially take this into account.

All you have to do at this point, is right-click your host node in the cluster and select the “Reconfigure for VMware HA” menu option. This will initiate a reconfiguration of the current HA settings taking into account the new networking configuration you just added. Once this step completes the warning about HA redundancy will disappear.