I recently had an article published in Data Center Journal about data center infrastructure management (DCIM) that I thought was worth sharing on this blog. In a nutshell, the article attempts to clear up some of the confusion in the market around exactly what DCIM is, along with some thoughts on what it takes to generate significant return on a DCIM investment.
If you talk to seven different people about what DCIM is, you’re likely to get seven different answers, none of them exactly wrong – but none entirely correct either. The problem is too many people – vendors and customers alike – look at DCIM through too narrow of a lens, focusing only on the one or two elements that matter most to them.
To steal an example from the Data Center Journal article, when you go to a doctor for a checkup, it takes more than a blood pressure test to assess your overall health. Similarly, you can’t measure the overall health and well being of a data center by checking only on the IT equipment or the cooling plant.
Yet these days vendors who happen to be good at managing the IT layer often refer to their management tool as a DCIM platform. The suite may not even integrate easily and cost-effectively with systems that monitor power, cooling and other building infrastructure management tools. So you may get a warm and fuzzy feeling because it appears your IT equipment is in fine shape, but only because you can’t see the temperature of your chilled water is rising to dangerous levels. You may not always know the correlation between the compute utilization going up and the impact that may have on the HVAC and power system. Conversely, if you can correlate these data points it may indicate things are explainable and OK, rather than have you thinking there’s a big issue.
A true DCIM platform has to integrate multiple systems to provide a complete picture of the health of your data center. It starts from the utility transformer, through the plug strip in the rack, to the server plugged into that plug strip, to the network providing communications to the cooling system and more.
In practice, a DCIM platform will integrate with a number of management tools that may already be in place in the data center. These include:
- The building management system (BMS) that monitors and controls a facility’s mechanical and electrical equipment
- A computerized maintenance management system (CMMS) that helps data center managers keep on top of maintenance schedules for various equipment
- A branch circuit monitoring system (BCMS) that helps ensure a steady flow of power to critical systems
- An IT asset management system that helps track placement, use, power draw and more for IT equipment
Some vendors will tell you that to implement DCIM you need to install a whole slew of new management tools so they’ll all work together. But if you choose an “open” DCIM tool, such as Schneider Electric’s StruxureWare, that is simply not the case.
StruxureWare is built on a Web services platform, so it can work with any other tool that produces Web services type output. With some work, it can also integrate with tools that don’t adhere to a Web services architecture.
However you accomplish it, it’s crucial to integrate the multiple tools that all play a role in painting a complete picture of the health of your data center. But you shouldn’t have to rip out those tools that are already working for you. Rather, look for a DCIM platform that will allow you to retain the investment, and experience, in the tools you already have. That’s how you can achieve rapid ROI and wind up with a DCIM tool that will give you sound insights into the health of your data center.