Some think a converged infrastructure is nerd-vana. Some have never tried to build one. In this series of six articles, we’ll look at what a converged infrastructure is, and what it isn’t.
You’ve read the stories about a dynamic data center environment consisting of pools of high-performing computing resources that can be centrally managed, readily automated and efficiently maintained. It’s the future of computing, they say.
In their different ways, the major systems vendors are each positioning themselves to participate fully in this IT environment of the future. Each one has its own overarching architecture. Call it “converged infrastructure,” “dynamic infrastructure,” or “unified computing.” (We’ll use the term converged infrastructure in this series.)
The converged infrastructure integrates all the aspects of your IT infrastructure — servers, storage, networks and management — into a unified environment that can be addressed and managed as a single entity. It’s the ultimate realization of utility computing.
A converged infrastructure is a process, not a panacea. It’s not going to make your all your problems go away instantly. The sophisticated IT environments required for today’s successful organizations, as you know, are not that simple since you’re the one standing in a data center that occupies the same space that once housed the company’s mainframes before they were replaced by minis, only to be replaced in turn by Wintel servers, which are now being replaced by blades.
Between you and the lovely image of a converged infrastructure are legacy applications that refuse to get along with anyone, high-security systems that you don’t particularly want getting along with just anyone, and the usual assortment of run-away applications, end-of-life technology and, of course, short staff and tight budgets.
Wouldn’t it be nice?
The good news is that truly converged infrastructure for your data center is now technologically feasible. All the technologies — servers, storage, networks and management — have evolved to the point where they can be converged into a single entity. In fact, if you had an unlimited budget, you could build the future of the computing today right in your own data center. And it would work great, too.
Instead of having to deploy individual resources piecemeal, you could deploy compute capacity and provision applications as they are brought online. You’d realize the key benefits of cloud computing — rapid self-provisioning, automation and capacity-on-demand — in your own data center. The next time you have a rush project with a high level mandate, you could provision a new system from available capacity on demand, and de-provision it and use the capacity somewhere else when the project is complete.
It would be possible to provide end users with a range of choices to provision their own computing capacity for specific projects. Your department could also accurately allocate costs to departments for the capacity they use, turning the IT department into a service provider instead of a cost provider.
A realistic process
If you don’t have an unlimited budget, however, what you need is a realistic process that will enable you to incrementally transform your current heterogeneous, somewhat jerry-rigged IT environment into the well-mannered converged architecture that you’ve read about on your vendor’s website. It can be done.
The next five articles in this series will help you think through and design such a process that fits your unique situation.
Look for these titles in the coming weeks:
Should You Give Change a Chance? – Don’t be sold on moving to some new infrastructure architecture in your enterprise unless you understand the risks … and rewards.
How to Break Through Your Infrastructure Silos – Before you can link your infrastructure, you’re going to have to break through some silos.
Don’t Let Your Legacy Become Your Legacy – There are often a few white elephants lurking in the corners of the corporate data center; they need to be addressed.
It’s Time to Take Out the Trash – Migration Day is a great excuse for getting rid of those things that just seem to be hanging on and taking up space. Not your co-workers, but the old hardware, apps and data that no longer serve a purpose.
Don’t Have a Party Just Yet – Moving to a converged infrastructure environment for your organization and actually making it work are two different things.
The converged infrastructure is not a destination. It’s a journey with all the twists and turns associated with any adventure. It actually does lead to the future of computing, but you get to decide how you get there.
Brandon Harris is VP of HP Solutions at Logicalis, an international provider of integrated information and communications technology solutions and services, where he is responsible for the relationship between HP and Logicalis and the overall support and growth of the HP business within Logicalis. Mr. Harris has been with Logicalis seven years. Before joining Logicalis he held positions in technical management and sales with divine/marchFIRST and Arrow Electronics where he managed the operations side of the reseller business, built a technical presale team, and was a presale technician himself.