Big Data

Does data center co-location make sense for Big Data?

Mary Shacklett discusses whether you should consider a co-lo for big data analytics and high performance computing.

Big data projects are exploding and not so coincidentally, so is data center co-location. The common IT case for data center co-location is that you can scale out your data center quickly, and at 20 percent of the cost that it would take to construct your own data facility. You might also have the option to only pay for what you use. This is an attractive proposition for IT departments that must be frugal at the same time that they must be able to rapidly scale IT up or down, depending on the needs of the business.

But there is another data center co-location use case that is emerging for IT: do you consider co-lo for big data analytics and high performance computing (HPC)?

If your big data needs can be solved with lower-end business analytics that process against relational databases or data marts that you probably already own, the answer is likely no. However, if you find yourself in a position where the business is demanding near-real-time analytics and HPC-strength processing that can yield answers to highly complex questions-you could find that you lack the compute resources and the IT skillsets required to run these applications internally.

This is where a data center co-location vendor with HPC cluster computing and in-house expertise on how to manage big data workloads can come into play. A co-lo vendor with on-staff big data expertise can provide fast times to market for IT as it meets new and emerging information demands from the end business. There is also the possibility of finding a solution to another IT quandary: the boatfuls of unstructured and (worse yet) uncategorized big data that potentially could be brought under control at the co-location data center with data deduplication, archiving and storage.

Do these strategies make sense?

For the short term, they present a compelling option-because startup costs can be less and time to results can be shorter.

Additional risk factors that IT must vet are the financial stability of the big data co-location facility vendor, the vendor's ability to produce value where IT can't, and the vendor's ability to meet IT security, governance, regulatory and intellectual property standards. Especially if you are a small or medium sized business (SMB), data center co-location for big data might be the only way that you will ever be in a position analytically to compete head-on with larger business competitors.

If you are a large enterprise, the long view of big data analytics is that you should probably consider in-housing it. The reason? Big data analytics in the future will be too vital and mission-critical to fully consign to an outside provider. To meet current big data demands and also to plan for the future, some enterprises are opting initially to team with a value-added data center co-location vendor in the big data and analytics space-with a longer-term strategy that brings these capabilities in-house.

Whatever your big data strategy is, here are three things you should include in your strategic execution:

  • Find a data center co-location vendor that is sympathetic to your business direction, especially if your desire is ultimately move your big data applications in-house.
  • Make sure that the co-lo vendor's technology direction aligns with your own-now and in the future.
  • Regularly communicate with your co-lo vendor. There should be an active dialogue between the two of you concerning big data projects that builds a healthy business partnership that you can both rely on.

About

Mary E. Shacklett is president of Transworld Data, a technology research and market development firm. Prior to founding the company, Mary was Senior Vice President of Marketing and Technology at TCCU, Inc., a financial services firm; Vice President o...

Editor's Picks

Free Newsletters, In your Inbox