I make no apologies whatsoever.
I am an unabashed fan of Vblocks: a converged IT infrastructure, produced and delivered by a converged entity (VCE), and consumed by converged IT organizations.
I was one of the earliest -- and perhaps the most vocal -- proponents of this new model. My posts on the subject drew more than a reasonable amount of criticism and outright scorn at the time.
That's to be expected, really. I was challenging conventional wisdom of how IT infrastructure should be organized and delivered.
But time has a way of helping people catch up. For many, the discussion has now evolved from "why would we do that?" to "how quickly can we get there?".
People tended to focus on deconstructing the Vblock itself, rather than the stepping back and understanding the context it was envisioned for. For example, if you run your IT infrastructure group as traditional, isolated silos, you look at something like a Vblock and say "that's not how we do things". And you'd be quite right.
However, if you've invested in delivering services vs. silos (ITaaS), the logic and appeal of a Vblock is almost irresistible.
As more IT groups are evolving their model, Vblocks continue to succeed in the market. That success enables additional investments, forming the basis of today's big VCE announcement. For me, there's two components of the announcement: the sort of thing you might expect, and a rather controversial surprise worth discussing in more detail.
So let's dive in.
The Model Shift
I entered the workforce in 1982 -- the tail end of the mainframe era, and the early days of the minicomputer / UNIX / desktop era.
Even at that tender age, I could see an established converged and homogeneous service delivery model start to fracture and splinter into all the "traditional" IT disciplines we see today, ostensibly in the name of cost savings.
It's been a long 30 years :)
There is now ample evidence that a converged IT team -- one that's organized around service delivery vs. technology silos, and given the right platform -- can deliver far better services, much faster and at far lower unit cost than IT teams not organized around this model.
This is not a new idea: you see it in IT service providers, well-run mainframe shops, etc. -- it's just relatively new for a majority of traditional enterprise IT teams.
As more IT leadership teams move to implement a new services delivery model, they go looking for a new platform to support IT service delivery at enterprise scale.
And -- very quickly -- the VCE Vblock leaps to the head of the pack. More groups are trying their first one; early adopters are now buying multiples.
With that all-important customer success now clearly achieved comes the ability to better understand those customer's true needs, to go off and invest in new capabilities and to create the beginning of a virtuous cycle -- one that's hard for competitors to catch up with.
The Four Elements Of The VCE Game Plan
If you'd like to understand VCE's strategy in a nutshell, here it is.
The fundamental "product" is the Vblock itself -- which is constantly evolving in multiple directions.
VCE is starting to create specialized and tailored versions of Vblocks -- one class for well-known application landscapes, the other for popular vertical and industry-specific use cases.
With this announcement, VCE introduces its first home-grown software management product: VCE Vision. That alone is worthy of extended discussion which we'll get into shortly.
And finally, converged infrastructure demands converged service delivery from VCE -- everything from requirements planning to operational support.
These guys routinely do amazing things for customers and partners because -- unlike their competitors -- they're organized for success.
VCE is product company, and its product is the Vblock.
Built from the best components of three recognized industry leaders (VMware, Cisco, EMC), it's a fast-moving treadmill because all three are innovating like crazy.
VCE has been very successful in creating a unique "productization" model where they can continually take the very best from all three, and still deliver a single product platform without breaking the attractive operational and support model that's an integral part of the Vblock value proposition.
This is no small feat -- consider the torrent of new product capabilities from EMC alone. Now fold in a similarly-sized fast-moving river from VMware, and yet another one from Cisco. Bringing all this parent-supplied value forward -- and masking the inherent complexity -- hasn't been done this successfully before in the industry.
And customers appreciate it.
One of the frequent observations of Vblocks is that -- to many people -- they're sorta big. Not everyone has the luxury of starting with >500 VMs in one go.
While the world is chock-full of large lumps of IT infrastructure requirements, the world is also full of more modest use cases -- ones where you still want to keep the consistency of the model, but just want the value proposition in a smaller package.
Today, VCE officially unveils two smaller versions of the Vblock -- the 100 and the 200.
While some of these will undoubtedly find their way into smaller IT shops, the real appeal will be larger, geographically-dispersed organizations who have to deploy IT infrastructure in multiple locations. I think they'll find the consistency of the deployment and operational model very attractive indeed. I know I would.
The original premise of a Vblock was as a standardized and converged consolidation platform for all sorts of wildly divergent enterprise IT workloads: mostly virtual, but a few physical ones as well if circumstanced dictated.
While the appeal of that model hasn't diminished, there's now an opportunity to create bespoke and pre-integrated configurations for popular enterprise IT use cases that are usually implemented at reasonable scale, and in one go.
Today's announcement unveils two built for SAP HANA implementations (very popular indeed!) in a 4 node (3+1) and an 8 node (7+1) configuration. All pre-built, pre-tested, pre-configured and pre-loaded with software. Fire it up, and get to work.
Unlike other flavors of purpose-built appliances, these are Vblocks. They have all the functionality of Vblocks, they grow and scale like Vblocks, they are managed and supported like Vblocks, and they play nice with other Vblocks that might be around.
The Management Conundrum
Perhaps the most interesting part of today's announcement is around VCE's new Vision management software. I find it a refreshing, realistic and progressive look at the current state of infrastructure management, and proposes a better solution than most.
A bit of background?
The whole topic of IT infrastructure management is both painfully complex and incredibly important.
The relative importance is easy to understand: in any form of IT service delivery, you're only as good as your processes, which are inevitably on a treadmill of continual improvement. And, from what I've seen, the treadmill only runs faster over time.
That means that any quest for the "perfect management tool" is usually a quixotic endeavor at best.
From a northbound (user-facing) perspective, there's inevitably continual and rapid evolution.
Processes change, management models change, roles change -- so the ideal tools to implement the new models inevitably change.
There is no magic automation pill.
From a southbound (infrastructure-facing) perspective, there's equally turbulent change: new versions of products, new capabilities within products, new integration points between products, etc. -- all of which can be highly visible and impactful to upper-level operational workflows.
So how does one go about masking the complexity?
One intriguing answer is the new VCE Vision. I tend to think of it as a BIOS for a Vblock -- masking the underlying complexity, and exposing the services and information bases needed for any variety of upper-level management frameworks: legacy, transitional and emergent.
It doesn't manage your Vblock device -- but it does make your device much easier to manage.
Unpacking VCE Vision
Today, most of the functionality in VCE Vision is passive monitoring, but -- whoa -- it does an awful lot of useful bits there.
For starters, it's continually discovering itself, so the infrastructure hardware and software inventory -- and associated configuration topology -- can be always be assumed to be accurate and up-to-date.
As it is checking itself, it's comparing current state against a constantly updated rule base of supported versions and associated configurations.
Exceptions and remediation points are flagged and raised in near real-time -- before there's a problem. Does your infrastructure do that today?
Logs are standardized, normalized and consolidated -- which then can be used by any number of management or security applications. Framework integration is off-the-shelf with VMware's cloud management suite (naturally), and there's a well-provisioned set of APIs, SDKs and coding examples available off-the-shelf, including an intriguing "Vblock System Simulator".
My guess is that it won't be too long before we'll see dozens of integrations with upper-level management frameworks: traditional, transitional and emerging.
The Right Answer?
The world is a diverse place, there's rarely categorically "right" answers and plenty of "it depends".
How you present and operationalize a Vblock's capabilities and management information might evolve quickly over time -- and should -- but the underlying enabler shouldn't change.
I wonder how long before we see this sort of enlightened management approach in other competing converged infrastructure stacks?
Don't Always Believe What You Read
Every so often, there's a spate of "Is This The End of VCE?" articles in the usual clickbait online publications. Anyone who's reasonably close to the action: vendors, partners, customers -- know the exact opposite is true.
It's an established winning model, and it's highly differentiated. The barriers to entry for competitors are steep, and quickly getting steeper. Customers are apparently voting with increasing amounts of their hard-won IT infrastructure budgets, limited only by their ability to adopt a new, converged operational model for IT infrastructure operations.
Today, what you're seeing is validation and expansion of that model. New models of the Vblock. Specialized versions for attractive use cases. And -- perhaps most importantly -- a fresh new approach to the vexing challenge of infrastructure management.
And, for one, I couldn't be more pleased ...