Where DCIM Software and Big Data Meet
Integration of ITSM and DCIM is just scratching the surface of true analytics-driven IT operations of the future
![Where DCIM Software and Big Data Meet Where DCIM Software and Big Data Meet](https://eu-images.contentstack.com/v3/assets/blt8eb3cdfc1fce5194/blt455a8084905b80f7/6621221e17903806645469d9/Comcast-distro-center-getty.jpg?width=1280&auto=webp&quality=95&format=jpg&disable=upscale)
After a couple of years of DCIM buzz, the hype has died down, and while analysts are forecasting healthy growth for the DCIM software market overall, they expect most of that growth to happen among a handful of the biggest players in the space.
What may make DCIM software increasingly relevant is progress in Big Data analytics. As companies deploy more sophisticated analytics systems to help with operations (and beyond), DCIM can provide a set of operational data about infrastructure that can be very useful to those systems.
It’s already happening today to a certain extent, but such scenarios aren’t yet widespread. The current trend in IT management is toward automating infrastructure to support any particular application, and this application-aware IT automation concept may soon spread to the underlying data center infrastructure – the domain of DCIM software – too.
The increasing focus by DCIM vendors on integrating with IT service management solutions – some people even say DCIM is becoming a subset of ITSM – is a step in that direction. DCIM tools today can tell you things like rack density, temperature, power consumption, where a server is, or where to put it to use the capacity more efficiently, but they don’t go much beyond that.
“This is where DCIM really stops, and there is a gap between the business and the data center; even IT and the data center,” Richard Jenkins, VP of worldwide marketing at RF Code, says. RF Code sells sensors for data center instrumentation as well as software solutions for data center management.
The future, according to him, is DCIM feeding data into Big Data analytics systems that companies use to make business decisions. “DCIM is a small piece of the overall picture,” he says. “The next wave of true Big Data will be an analytics platform that will stick between the data center and the business itself.”
Nlyte Software, one of the major DCIM software vendors, has put a lot of effort into integrating with ITSM platforms and has also always made sure to have an open API, so that the data its software collects can be pushed to whatever systems need it. In that future of DCIM working as part of a holistic analytics-based management system, open APIs will be crucial.
Robert Neave, Nlyte CTO and co-founder, says one of the company’s biggest customers pushes asset management data Nlyte’s software collects from its massive data center infrastructure, energy usage data collected by its facilities management system, network monitoring data, and other types of infrastructure data into Vertica, HP’s Big Data analytics platform, to understand and manage infrastructure demand of its cloud-based applications.
Neave declined to say who the company was, citing confidentiality agreements with the big customer, but said it was a high-tech company with about 120,000 data center racks under management.
There are applications for Big Data within DCIM itself too, particularly when DCIM software does predictive analytics. This is a capability the team behind Emerson’s Trellis DCIM team is working on, Steve Geffin, VP of strategic initiatives at Emerson’s Network Power unit, said.
Predictive analytics is a Big Data problem, Geffin says. Trellis uses Big Data to create and refine operational models for some customers. Infrastructure data gets fed into a Hadoop cluster and analyzed to help the customer make operational decisions, he says.
One example is predicting when a UPS battery going to fail. Data center operators usually replace batteries on a defined cycle, and when a battery gets replaced it doesn’t necessarily mean it has reached its end of life. Very often a battery may remain perfectly fine for another year, but it gets replaced as a precautionary measure. By analyzing patterns over many batteries’ lifecycles, replacement can be deferred until it’s actually necessary, which can save the operator a lot of money. “Problems like this you can only solve using Big Data techniques,” Geffin says.
About the Author
You May Also Like