Cloud Computing: What Data Center Hardware Should You Keep?

Enterprises today have so much to think about when designing their overall data center strategies.  It’s critical that a holistic approach be taken when laying out a data center design.  The challenge for many CIOs and their IT organizations today is working with their partner to determine the best approach.  First, what to do with the existing infrastructure in place? 

With virtualization, many organizations rid themselves of physical servers, despite the significant dollar investment, and went to virtualization.  Although many continue to do virtualization in a more phased approach, the understanding was that long-term, virtualization not only provided organizations with cost savings, but also better performance, agility, and flexibility, to name a few benefits.  The point is, virtualization was not just about the cost savings, but about the value adds that it brought to the overall IT environment and ultimately, the business.  

Similar challenges are being faced today.  While enterprises are making the move to deploying cloud computing models in their data centers and leveraging external public clouds, it becomes difficult to determine what to keep and what to let go, particularly in terms of physical infrastructure. 

True, driving cost out of IT is the goal, but what about the investment already made?  It comes down to the old “rip and replace” story that has plagued so many emerging technologies in the past.  Should an organization really consider getting rid of its infrastructure?  What happens to all that investment made?  Worse, what happens to the IT talent that managed it?  It has always been a taboo for any IT executive to even consider this option.

However, in the past few months, I’ve heard more and more stories about CIOs and IT executives making the decision to take out particular on-premise solutions  and instead go with an as-a-service solution, whether it be hosted or managed services.  In interviewing a number of executives, it started to make sense.  The cost of maintaining and managing old infrastructure and the human resources that go along with doing that, in the long run, actually becomes more expensive, than just pulling it all out and moving to a more cost efficient deployment model.  Their only challenge was making the case to the CEO. 

However, CEOs today expect their CIOs and COOs to work together and make the necessary IT decisions that will enable further growth for the business.  In several interviews, the executive indicated that the numbers worked out, so the CEO approved moving forward and starting to take out much of that on-premise infrastructure. 

Each case is different; in many cases, it’s still a difficult decision to make.  Ultimately, most infrastructure costs are major investments made within organizations and CEOs still are looking to extract as much value as possible. 

The challenge is being able to determine this:  what is the POTENTIAL value of moving from on-premise to off-premise? Does the cost of managing and maintaining on premise (including human/capital resources) outweigh the cost of simply taking it out and moving to cloud?  In most instances, particularly with respect to legacy infrastructure, it’s surprising to see that it makes sense. 

Interested in sharing your thoughts?  Would it be practical for you to consider ripping out legacy infrastructure and moving to more of an as-a-service model?  Have you calculated what the cost is in maintaining and managing legacy infrastructure? 

Vanessa Alvarez is an Industry Analyst with Frost & Sullivan focusing on monitoring and analyzing emerging trends, technologies and market dynamics in the area of enterprise infrastructure in North America. Follow her on Twitter @VanessaAlvarez1.