Infrastructure is Boring: The Next Phase

The VMware keynote at EMC World 2011 focused on defining the move into a “post-PC” world. It’s been a fairly common refrain for VMware for the last couple years, but in this particular keynote I got the impression that we seem to be preparing for a world where the hypervisor isn’t front and 6a0128779ad48c970c014e88ab21f6970dcenter as well. That’s not to say that the hypervisor isn’t important or critical as the foundation of the platform, but well, it’s not very sexy any more. Personally, I think VMware has done a great job of continuing to innovate at the hypervisor level, far more so than any other provider in the market, and it shows in the depth and breadth of their offering; you don’t get an 85% market-share by accident.

One of the byproducts of innovation is that by doing so you move the ecosystem forward, and I think that’s what we are seeing in this case. VMware has opened up enough new doors that the excitement has moved past just virtualization to the higher level functions closer to the applications/users. For those of us who have been intimately involved in the back-end of the current virtualization, this can come with a feeling of shock or disappointment, but since all of that infrastructure exists to better serve the end-users and their applications, I choose instead to be excited about the possibilities.  The obvious follow-up question is what this means for the converged infrastructure movement, and I would suggest that not only is the move to converged stacks going to continue to accelerate, it’s one of the core enablers that is allowing this transition to happen.

Let’s be honest: the world doesn’t revolve around infrastructure; it revolves around people and the applications they use.  Sure, I work for an 6a0128779ad48c970c01538eb7a69a970binfrastructure vendor, so maybe that’s not a cool thing to say, but I believe it’s true.  If you think about the “journey to the private cloud” concept that EMC and VCE uses to track where customers are with their virtualization project, everything after phase one is designed to move the focus of the IT groups from the physical assets to the end users.  In the business production phase the effort is around taking the virtualized infrastructure and using it to change how the internal IT groups relate to the application LOB owners and end-users.  In the business production or “Operational” phase, the focus is on improving the business processes, increasing the uptime and availability of the applications and putting the framework in place that can support growth and allow flexibility.  Here, users start really digging into the tier 1 applications and are starting to reach out to the business and define SLAs and match the infrastructure to the day-to-day business requirements and strategies.  It’s at this point where (IMO) the real magic starts to happen.

Up to this point, an enterprise really has two “teams”.  The first is the IT team, and for the sake of this discussion I’m going to lump all of the traditional silos together, including storage, compute, network, etc.  This team is typically a cost center, and they are usually caught in a place where they are reactive to requests from the business.  On the other side of the wall are the consumers.  These may be application teams, these may be LOB owners or these may be regular users who spend all day in a group of particular applications.  They may have some idea about how their applications relate to the IT team, but not always.  They probably don’t know about storage requirements, have an opinion on NFS vs. block, care about network convergence or know what sort of regulatory burden the IT staff is under.  They just want to be able to use the applications that are part of their job without any fuss.  The third step of this journey is less about technology than it is about reimagining the relationship between these two teams.  Break down the wall, align 6a0128779ad48c970c014e88ab2213970dthe resources and make it so that the IT groups can enable the LOB owners and applications teams to help service their own users.  Get in front of the process and be proactive (or allow the infrastructure to be proactive) in how it addresses the needs of those applications and their users.  Give the end users the insight that they need into the infrastructure in order to make informed decisions on how to utilize resources.  THAT is what IT-as-a-Service means to me.

Add to this the Platform-as-a-Service, rampant API development and all of the traditional SaaS use-cases that are out there, and the focus has clearly (and rightfully) shifted to the user/application.  This doesn’t mean the infrastructure isn’t as critical as it’s always been, but I think that it should cause customers to possibly re-evaluate their criteria when purchasing.  It’s not enough, anymore, to settle for the same old model when you are looking to integrate more closely into the other side of the house.  Can your infrastructure drive operational efficiency into your organization?  Can your infrastructure provide you the tools needed to manage the compute, network and storage components together, and provide enough cross-component visibility to allow the support groups to be able to work together well?  Can your infrastructure integrate with the toolset that you want to use to allow more control to your customers?  Can you get one-call support on any part of the infrastructure quickly and easily?  Does your infrastructure provider give you the pre-tested and pre-validated solution designs that you can reference to make the job of deploying/architecting easier and reduce the amount of non-recurring engineering time required?  If the answer to any of those was “no”, maybe it’s time to take a closer look at how your business is going to move IT closer to the people who matter.