Page 12 - COMPASS, Q2 2017

feature > the nsx bet
THE BATTLE FOR SDN IS GOING TO BE WON NOT BY THE COMPANY WITH THE BEST TECHNOLOGY, BUT BY THE COMPANY THAT MAKES THE BEST PARTNERSHIPS.
buzzword cred. (Figure 2 provides an overview of NSX-T capabilities.)
The big gap is that, for most organizations, APIs and programmatic infrastructure are a science experiment. A versioned, desired state configuration environment is a pipe dream, DevOps is still more fad than fact, and most virtualization administrators lean heavily on their management tools.
This is the reason NSX isn’t flooding the commercial midmarket and small to midsize business (SMB) space. It isn’t because NSX has no value here: For a 100-employee SMB shop with 15 sites, NSX would be absolutely transformative. The issue, as always, is ease of use.
Individually, VMware has all the pieces to make a hybrid self-service cloud environment where workloads move between sites on the company premises, services provider hosts and select public cloud environments, all while preserving a company’s network configura- tions. Tying it together is where the difficulty lies.
HETEROGENEITY
Beyond the ease-of-use argument lies the cold reality that datacen- ters are heterogeneous. If they weren’t, we’d all be using Microsoft on Microsoft with some added Microsoft, and Microsoft in the cloud. It is the only one that owns all the pieces in a convincing enough state of readiness to deliver a great-big, vertically integrated stack.
In most of the IT industry, when people say “virtualization,” they mean VMware ... oh, and Microsoft sometimes, too. Similarly, when people say “public cloud,” they mean Amazon Web Services ... oh, and Microsoft sometimes, too. That vertically integrated play has created for Microsoft a block of absolutely loyal customers willing to consume the whole stack. The rest of the world mixes and matches.
The battle for SDN is going to be won not by the company with the best technology, but by the company that makes the best partner- ships. Arguments about VMware’s primacy are similar to those about Microsoft and the desktop.
Yes, Microsoft owns most of the desktop market share. But when you look at endpoints as a whole, Android crushes Microsoft like a bug. Yes, VMware dominates x86 virtualization. But when you look at workload
10 vmug > compass Q2 2017
encapsulation as a whole in 2020, what will you see?
This is before you even touch integration of serverless or Software-as-a-Service (SaaS) workloads into an SDN microsegmentation and security regime.
To defeat its enemies and emerge the unchallenged victor of network virtualization, achieving a dominance it hasn’t enjoyed since the golden heyday of x86 virtual- ization, VMware must convince direct competitors to join it. These range from hyperconverged and cloud-in-a-can vendors, to services providers, public cloud providers and networking vendors. VMware’s recent ban on third-party virtual switches compli- cates the issue a great deal and, in fact, may have started at least one no-holds-barred corporate war.
VMware’s future as a tech titan rests on its ability to turn the amazing technologies it has assembled into a single management suite that can address a diverse and con- stantly evolving heterogeneous environment where all the bits under management aren’t even owned by the customer buying the tools. A daunting challenge, indeed, but one VMware is better positioned to tackle than any other player in the industry today.
Trevor Pott is a full-time nerd from Edmonton, Alberta, Canada. A columnist for Virtualization & Cloud Review, he splits his time between systems administration, technology writing and consulting. As a consultant, he helps Silicon Valley startups better understand systems administrators and how to sell to them.