Data Centers Guide

A Data Center Operations Guide for Maximum Reliability

March 19, 2016
With so much focus on the critical systems infrastructure design of a data center it leaves the impression that this is the key to predicting the reliability of a data center - but may not be.

There is significant time spent in the data center industry discussing data center infrastructure design or the Tier rating of a data center. With so much focus on the critical systems infrastructure design of a data center (specifically the electrical and mechanical systems) it leaves the impression that this is the key to predicting the reliability of a data center. This way of thinking seems like a “break – fix” philosophy for data center operations. The reason to have data centers is that utilities are generally “break – fix” operations. The Tier rating of a data center is an indicator of the distribution paths, capacities, redundancies, and approximately how many planned maintenance windows the IT-end user can expect, assuming none of the maintenance is deferred. However, as the primary factor or predictor of reliability, the Tier rating alone falls short. Data center reliability is the combination of many factors of which infrastructure design is only one of many.

People, processes, operations, maintenance, lifecycle, and risk mitigation strategies are also necessary in creating reliability. The strategies within this eBook work with any Tier rating and decrease the likelihood of unplanned downtime or outages in a data center.

Recently FORTRUST surpassed its 12th year of continuous critical systems uptime without a single instance of unplanned downtime in our Denver data center. Over the past several years we have been asked on many occasions, how do we do it? I do not believe there is anything secretive or special about what we do. However, I do believe there is something to how we do things, and the mindset behind why we do it that way. As data center professionals, we are keenly aware of the math that provides the inevitable conclusion. Nothing is 100%. Because the data center industry is charged with the quest for continuous uptime, data center professionals are somewhat on-edge, or even superstitious. As a result, speaking about how long something has been operating without an outage or unplanned downtime is not always a comfortable conversation.