February 18, 2013
Author’s Note: For any of you who have ever done a large-scale data center project, you are guaranteed to have experienced a least one-near death experience before it was approved. Elizabeth Kubler Ross was the quintessential expert on death and grief. Without her wisdom this blog would not have been possible.
Sometimes the things that you hated as a kid turn out to be pretty good when you get older. For example, whenever I wasn’t allowed to do something in my youth that was justified by the arbitrary finality of “Because I said so” I would stomp off to my room to brood about how unfair my parents were. Now I have two children of my own and I have learned to love “Because I said so” (BISS). Sure we’d all like to be like Ward Cleaver and sit down and rationally explain our reasoning when we tell our kids no–and I do try to do this as often as I can—but sometimes, like in the middle of the Super Bowl, you just need a quick go to response that neatly ties everything up. Is it fair? No. Do my kids stomp off to their rooms? Sure. Did I avoid missing the biggest play of the game? You betcha. But seriously, isn’t using BISS just a crutch because we don’t really want to think about the subject? Why do the mental gymnastics when we’ve got the old stand by to fall back on? In a lot of ways we have the equivalent of BISS in the data center industry. It’s called, “It’s in the standard” (IITS).
We all understand that the foundation for selecting a data center, whether built or purchased, is premised on looking inward and understanding how the facility is to be used over time, what the reliability requirements are, etc. Without doing the homework upfront, the resultant data center typically doesn’t perform as it should to support the required applications. However, slavish devotion to the resulting specification isn’t always the best way to go either.
It’s human nature to take pride in putting together a comprehensive specification, and nothing punctures our balloon faster than someone asking us why we made one decision over an alternative. It’s just too hard to admit you conceded that point to Bob in exchange for two other things you really wanted. Although this is exactly the strategy I used to use when making “fair” trades in Monopoly with my sister. Typically our response is not unlike Elizabeth Kubler Ross’ five stages of grief.
“It’s in the spec” is our reflexive response to the “why question” and approximates Ross’ first stage—Denial. “How dare you question the spec? We love and worship the spec since it was crafted by the finest minds of our organization, including, well, okay, me.”
“IITS” is also the central focus of stage two—Anger. “Do you realize how long it took to develop this spec? Three years for God’s sake. Of course, we need a month’s worth of fuel. No, I don’t care if the storage tanks will be bigger than King Tut’s tomb.” For too many of us this is as far as we get in the process. The end result is that we pay far too much for a data center based on requirements that we don’t really need due to our blind fidelity to the spec.
Stage Three is bargaining. It is the beginning of breaking down the IITS barrier. “Oh, we’d need to do all that polishing and testing just to make sure that the fuel would actually be available if we needed it? Well, how about 15 days worth?” Ross acknowledged the pain associated with this first brush with turning our backs on fealty to the spec, but noted that was the mature response.
Depression comprises Step Four. Our vision of the data center is now housed in a dark, remote place. Our comprising on the spec leaves us unmoored, with feelings of great uncertainty. Our faith in “IITS” has been stripped away, and we are in search of meaning.
Finally, we reach Stage Five, Acceptance. During this culmination of the process we are free to mourn the spec. Our grief, however, is offset through the understanding that the resulting data center will be more cost effective, and that its design actually meets our needs. “Two days of fuel with a local and national delivery contract to back it is more than enough to ensure the level of reliability that is needed. Oh, and it costs less and reduces the risk of the fuel going bad and gunking up (technical term) our engines.” We are reborn in our realization that specifications can be altered without damaging our personal feelings of well-being.
We all need our crutch phrases. They give us comfort and protect us. “It’s in the spec” is no different. At times it may even be sufficient. But more often than not it reflects a rigidness of belief that stifles our ability to obtain the data center that we need, not want. Just as she showed us the way to understand and deal with grief, Ross’ five-stage process provides us with a guide that enables us to put aside dogmatic adherence to misguided specifications and embrace cost effectiveness and operational suitability via enlightened self-awareness.