Data Center Inbreeding

Remember that movie Deliverance, where four regular guys go canoeing and run into some hard times with a few in-bred locals? Although that one guy could really play the banjo, it was pretty obvious that the natives had formed what one might euphemistically describe as a “close-knit community”, and they weren’t to keen on the idea of any outsiders joining the party. As Twain once said, “familiarity breeds contempt and children”, and in the movie it obviously bred both. Although unconsciously, the movie actually pointed out the dangers of what can happen when societies, industries and organizations become too insular. I was reminded of this the other day while I was speaking with an industry “insider” about patterns of behavior in the data center business.

While we spoke on a variety of topics I was struck by his incessant need to push back against any data that contrasted with his “world view”. Maybe it’s just me, but I see more and more of this lately. If any alternatives to beliefs such as “the cloud is a transformative technology”, “widespread use of DCiM is a certainty” or my personal favorite, “there are two kinds of modular data centers: shell with incremental build out or pre-fabricated units” are offered, the knee jerk response is dogmatic and devoid of hard data. Believe me, I understand that no one wants to hear that their baby is ugly, but I must admit that such intransience emanating from the people who are paid to be “visionaries” is curious.

Responses such as “That’s not what we see” or “That’s not what my client’s are telling me” aren’t exactly illuminating. I find the latter comment most disconcerting since these universes tend to be small, and I wonder how often these folks actually disagree with the customers that are paying them for their services. How often are actual end users consulted and in what volume? Does a Fortune 500 customer share the same thoughts on cloud computing as an emerging enterprise? I don’t think that we really know. While speaking with key industry participants is certainly one way of gaining insight into industry trends, this closed shop approach can serve as an echo chamber that only reinforces opinions rather than challenge them.

With each trade show I attend, or blog, article or research report I read, I can’t help but wonder if we don’t devote too much of our time preaching to the converted. Dissent does not seem to be a part of our discourse. Wouldn’t our industry, and our customers, be better off if more of us actually called a spade a spade? Not every idea or innovation is a good one, and we should feel free to say so. Instead it seems that most of us hide behind terminology and concepts that appear to be safe, and we then spend the majority of time analyzing the most trivial nuances of their adoption and implementation. The most absurd aspect of this mindset is that when someone does attempt to make a “big prediction” they tend to make them devoid of, or completely counter to, actual data, and the prognostication only serves to elevate prevailing views. In light of this lemming-like mindset, please allow me to be the first to say that the cloud will not diminish data center adoption. Demand data doesn’t support that conclusion, and all that cloud-based information requires computing and storage capacity that need a home, in other words, data centers. That said, the cloud will diminish the server-closet, as small and medium business continue to elect to purchase their IT services over an internet connection in lieu of buying boxes and software (as an FYI, at Compass we don’t own a server).

Sometimes I wonder if the opinion makers of the data center industry really even understand our business. Oh, they love the technology enough, but do they actually understand why customers add data center capacity, and what they really base their evaluations on? If they do, I certainly don’t see it. Consensus opinion and a focus on the trivial seem to have replaced end-user based analysis on actual demand drivers, and what the majority of customers are using to make their data center decisions.

For our industry to continue to grow and mature we need to better understand the needs, not the buzzword wants, of our customers. At a time when we should be more open as an industry, it seems that we are only becoming more insular. While this may make us better banjo players, it doesn’t help us improve the product that we offer to our customers in any meaningful way.