How UPSs have Transformed from Simple Power Devices to Predictors of Operational Efficiency Facebook LinkedIn Twitter Email Pankaj SharmaJanuary 5, 2018January 5, 2018 LinkedIn Viewed: 2652 TAGSend userUPSuptimeDMaaSedge data centersdata centerback up powerserverCertainty in a Connected Worlddata management as a service Most small and medium-sized businesses over the years have protected their small data centers, server rooms and wiring closets with traditional rack-based or under-the-desk “tower” Uninterruptible Power Supplies (UPS). These systems ensure that a power backup is in place, in case utility power is cut off due to electrical storms or other unanticipated power supply interruptions. The batteries inside the UPS afford the computer users a window of time to properly save their work before the battery power runs out. A UPS system, for the most part, acts as an insurance policy which protects and secures data during power outages, and also filters out more common electrical anomalies like power swells and sags. Some of the more advanced UPSs offer built-in management software that allows for graceful remote system shutdown during times of prolonged power outages when valuable data is at risk. Users find this capability useful for restarting their servers from remote locations. Systems operators trying to catch a good night’s sleep at home don’t need to start their cars and drive to their workplace in the middle of the night in order to reboot servers. Today, as the more successful businesses become highly data-centric and more connected, users require that their server room technology evolve to support efficiency and business growth. As the business engages more projects, programs and people, meeting the future needs of the organization becomes more complex. Within the small data center or server room, it becomes more difficult to predict how efficiently the data center is operating because the environment is continually changing. The elusive goal of most server room operators today is to gain a high level of reliability and Certainty in a Connected World. The new revolution in data center predictive analysis Many data center devices are now capable of creating a tidal wave of big data – ranging from device status, to alarms, to setting information, to battery performance tracking and so much more. The new challenge for business owners is to decide what to do with all of the data that these devices are generating. Owners also are challenged with finding ways to convert that gathered data into meaningful business information that helps improve their business’s performance and bottom line. Even devices such as UPSs, when connected to analytics solutions in the cloud, play a role in contributing to the new data pool of data center component performance intelligence. For example, connected UPSs now communicate data about their own performance and the performance of the server room devices that are connected to them (i.e., servers, storage, networking devices). This innovation is important because these new UPSs create business value above and beyond their traditional role of power protection and high reliability. These digitization-driven approaches change the way partners and technicians can offer end user support, particularly in edge computing environments. Consider the following example. All UPSs and connected devices within multiple remote data closets can now be remotely monitored from one interface (and that interface can even be a mobile phone) utilizing cloud-based software. Besides providing global visibility from anywhere at any time, an additional benefit of this DMaaS (Data Center Management as a Service) platform approach is active intelligence gathering that enhances both systems efficiency and edge data center uptime. As the equipment is monitored, a large “data lake”, driven by both algorithms and machine learning, is created and then analyzed by software. Performance data such as the amount of power that the devices connected to the UPS are consuming, or the number of times the UPSs have had to go to battery over the last month is utilized to predict when a particular device might need replacement. For example, the data can show that internal temperatures have been running hot or that a battery, based on its performance history, will die out sooner than what is normal. For the server room operator, life becomes much simpler and guesswork is eliminated. By now gaining access on cloud-based analytics outputs that interpret equipment performance data, and prompting actions based on that data, the operator and/or the affiliated partner drives a return for the business – either in increased productivity or by avoiding costly downtime and business interruption. Click to learn more about how these new predictive analysis techniques and services can simplify both on-premise and edge data center management.