Q: Why is preparing a data center for the summer months different than the rest of the year?
A: Summer brings many expected and unexpected elements. Heat waves for example, are inevitable, but can be a huge hazard to operations if it causes overheating. On the other hand, summer marks the beginning of the Atlantic hurricane season, so knowing how to act and how your infrastructure will respond is crucial, especially as outages can be costly — averaging $29k and lasting an average of 8 hours, according to research we recently conducted.
Q: What can data center operators do to prepare?
A: Understanding all assets within the data center is the first step. The best way to do this is with real-time telemetry data collected through an automated process, at a server and rack level that provides a more holistic, 360-degree view on how each device is running. Ironically, many data centers still rely on manual processes for planning and forecasting in their data centers. Although a proven method, in an era of automation, it is not as efficient as it could be and automated methods, including the collecting and leveraging real-time telemetry data, grant data, and insights not otherwise provided that can be leveraged for a more efficient operation.
Q: What are some things that can be done at a tactical level?
A: Using real-time data, one of our customers realized that they could run their data center at a higher temperature than is the industry standard. In doing so, they found they could save 2% per degree on each power bill. These analytics will help in creating, implementing and maintaining a cooling distribution plan. Analytics can also help identify underutilized serves, which is a major drain on data center efficiency.
Q: Why are analytics important in the summer months?
A: Analytics are important throughout the year, but especially in the months that can potentially have the biggest impact on your data center. Analytics truly allow you to fully understand how your data center is operating and what can be improved. For example, by using analytics to constantly gauge over and under cooled areas, you can potentially detect flaws in a cooling plan — something you don’t want in the hottest months of the year.