We’ve rolled out a few instances of the Cisco 1000V virtual switch in our environment and the standard monitoring looks great. The issue I’m coming up against is that once the VSM is added into NPM, NPM doesn’t enumerate interfaces when it periodically rediscovers the device. As VEMs and VMs are added, I have to manually list the resources to start monitoring the interfaces. I also imagine deleting VMs and VEMs will also leave orphaned interfaces until manually pruned. Any idea how NPM can be setup to match the dynamic nature of the 1000V?
Open to ideas on this one from the community, but right now I think you will have to update them manually in regards to discovering new ones. In another post regarding ESX4, not sure if you have seen it or not, we used to use SNMP to gather VM and Guest stats, but with 4, they no longer support populating their MIB's, so we are working on re-enabling that integration by using their API instead. I would have to research, but we may be able to get this type of dynamic data via their API