NSX needs more than the ability to be innovative in managing networks and “just” adding value similar to how x86 virtualization has done for the server team. One of the key moments of VMworld was when the VMware COO called VMWorld the largest gathering of Network Administrators. There are more virtual ports than physical and virtualization administrators administer most of those virtual ports. However, the physical network and the therefore the keys to the network and the destiny of SDN lay in the hands of the network team.
It has been my experience that the network manager buys the next box that their network hardware vendor recommends. I’ve rarely seen wholesale shifts from one hardware provider to another in the enterprise space. Networks are complex and when they are up and running there is very little desire to want to change something that isn’t perceived to be broken.
When you talk to server administrators they perceive the network as broken. It doesn’t fit the service model that they want to achieve. When you talk to network managers there’s really nothing wrong with the network. It gets packets from Point A to Point B reliably and within the latency tolerance. So, why introduce a whole new management and service delivery to a complex system such as the network because your customer (server admins) say that the model is broken?
I believe it’s up to VMware and their partners to convince the Network Managers that the network is broken and the next “box” they need is not a box at all but a new service model. I think this is a hard sell. But, we’ll see as VMware does have a substantial partner list with NSX. Cisco is going the route of leveraging their relationship with the network decision maker directly. The future of the technology my weigh on who’s the better salesman.