I have been getting into discussions on web and see people claiming or rather predicting how SDN is going to take over the current Network Industry . While SDN to me is not necessary a bad thing but something couldn't just be called as SDN just because it solves an old problem in a different way (Programming the network is just one of those).
Also IMHO its not necessarily to do some new things just because that's possible. For example I have been recently to a NSX session where trainer was telling about how VMware teams won't depend upon Network Engineers any longer to setup a network to grant access to new VM or Server since they usually takes weeks and now NSX solves this problem with overlay networks.
But is that really a Network Engineering Problem to begin with ? or Rather it's a Business Model / Process Design Issue ? :) ... So while SDN or NSX tries to solve this problem with overlay networks, they seem to be rather focusing on wrong areas. :)
What Do You Think ?