Monday, September 09, 2013

Wrapping up VMworld 2013 in SF

Belated I know. But last week was spent balancing some high priority work with R&R and I wanted to give my first reactions a little time to settle anyway. At least that's my excuse and I'm sticking with it. So, without further ado, here were some things that either struck me or caught my eye at my seventh VMworld.

Disclaimer: I am Cloud Evangelist for Red Hat. Read what I write in that context. However, also please note that these opinions are mine alone and do not necessarily represent the position of my employer. I also note that I usually avoid getting into a lot of competitor specifics when I write. However, VMworld is a major event and I thought some may find my observations useful.


It was sober.

By which I mean it was more about product/initiative rationalization, integration, and extension than it was about bold claims and major new directions. Even that which was arguably most new, VMware NSX (network virtualization), was a largely anticipated consequence of VMware's Nicira acquisition. (Furthermore, based on a number of conversations I had at the show, it isn't clear to me that the typical attendee understands at this point what software-defined-networking means for them either technologically or organizationally. To be clear, I firmly believe that software-defined-everything is an important trend; I just think it will take time to develop.)

Sober isn't necessarily bad, by the way. In fact, the industry analyst who described this year's event to me thusly weigh meant it as something of a complement. The opposite of hype-to-excess. But it made for a general sense of VMware securing its core virtualization business and framing every big trend in the context of that existing business even though cloud computing, for example, is hardly just virtualization with a bit of management tacked on. Note, for example, how deliberately the VMware NSX moniker evokes VMware ESX. 

Exploiting existing beachheads can be a good business and technology strategy. It's also a strategy that can fail miserably. (See, for example, Microsoft's determination to bridge from its Windows dominance to mobile.)

Speaking of which. "Defy Convention." Really? 

I get that someone in VMware corporate marketing or some executive thought it would be a good idea to theme the show around the polar opposite of stodgy old VMware. The problem is that this brainstorm apparently began and ended with a slogan and some grunge graphics. All the actual content of the conference was about incrementally extending virtualization. A comforting message for the thousands of virtualization admins in attendance? Sure. A message about upending the status quo? Not so much.

(OK. VMware NSX--or at least what it represents--is threatening to network admins and to Cisco. But even most anything that whiffed too strongly of software-defined-storage was pretty low profile given how VMware parent EMC is still deeply in the business of selling expensive specialized storage hardware.)

The non-hybrid hybrid.

Just one example of not-defying-convention is VMware's approach to hybrid cloud. You can have any kind of hybrid cloud you want, private or public, so long as it's based on VMware technology. It has a certain logic from VMware's perspective. They've had no real success competing with non-VMware-based public cloud providers and are unlikely to change that using their proprietary software. What they can arguably do is provide a simple bridge for their installed base to a tightly circumscribed set of public cloud resources. (That these public cloud resources will be a combination of VMware data centers and those partners would seem to be a channel conflict issue but that's another topic.)

A useful incremental feature for a subset of their installed base perhaps. But hardly hybrid in any true meaning of the word. Chalk up another one for convention.

 Pivotal MIA.

Finally I note that Pivotal, the Paul Maritz-led EMC initiative which includes CloudFoundry and other VMware pieces involved with developing cloud applications was largely absent from the show. It had a booth--plastered with various vague slogans around "making business faster" and such--and put out a short press release with VMware about expanding their "strategic partnership" but wasn't otherwise very visible. 

I get the deliberate arms-length relationship that helps VMware focus on its core business while Pivotal works on cool new stuff. But one consequence is that VMware looks to be refocused on infrastructure and its historical comfort zone (and the comfort zone of many in attendance at VMworld) rather than the broader possibilities opened up by a complete cloud portfolio.

And that's about convention, not defying it.


No comments: