I don’t believe VMware’s NSX or network virtualization for that matter is the innovation being brought by the release of the product. One of the advantages to VMware releasing NSX is that we get to see the true value of network virtualization via the hands of end users. Like most new technologies that progress from whiteboard to production, we really never know the true use of the technology until it’s put in the end user’s hands.
Twitter is a great example of a platform that was taken over by end users and morphed into what it is today. I see the maturation of network virtualization in the same light. To date, with the exception of some limited large production solutions under the Nicira brand, NSX hadn’t many implementations and limited opportunity to be stretched to and beyond its capability. It’s when you stretch the pre-defined use of technologies that you begin to see innovation.
To date the advantages of network virtualization have mainly been academic. We’ve gotten Cisco’s take on the need for network virtualization to be a hardware first approach from a technology perspective. I’m sure there are merits to Cisco’s argument but this is an example of a whiteboard discussion. The great thing about the release of a large product such as NSX is that we get to move from academia to the real challenges of implementation and desired features.
Related post – Dispelling the FUD around Network Virtualization
VMware has said the value proposition for network virtualization product such as NSX it that they allow you to do things faster. I understand the need to create a simple value proposition however, I don’t know that this is network virtualization’s ultimate value proposition. Just like the hardware first agrument of Cisco’s ACI, it has been an academic discussion. I think the community will flesh it out and we’ll see use cases that spring up that weren’t possible until the product we put in the hands of network engineers that will use it in ways never considered.
It’s the way of Computer Science
When we provide abstraction, we get results that couldn’t be conceived without the decoupling of the logical constructs from the physical components. This is the general course of computer science. Give developers and end users the freedom that comes with abstraction and innovation soon follows.
So, now that we have a highly visible network virtualization solution in production, I’m excited to see the real innovation.