The Internet of Things and Data Center Automation

Internet-connected devices first outnumbered the human population in 2008. By 2020, there will be 26 billion installed Internet of Things devices, according to Gartner.

But even beyond its rapid growth, the Internet of Things (IoT) is becoming increasingly complex as IoT devices become automated. Together, the rapid growth and increasing complexity of the IoT demand a change in the way that enterprises consume data center services, and the way that providers build and manage their data centers.

Automation of the Internet of Things affects the data center in another way, too: It creates the opportunity to automate the data center itself.

Automation in the Data Center Today – An Evolution from Monitoring to Control

The first step toward automation of the data center begins with monitoring. Sensors on the hardware (there are 700-1,000 sensors in the IO.Anywhere data center module) enable the IO operating system, IO.OS, to monitor the performance of the data center. The next step is to introduce control capability – the ability to actually control hardware functions from the operating system.

The evolution from monitoring to control is similar with other Internet of Things devices. We see that evolution in the automobile, for example. Now many cars have front and rear cameras that help a driver park with an on-dash display of what the camera sees; many will beep when the car is too close to another. Some cars even park themselves. The self-driving Google car takes those kinds of sensor-enabled monitoring capabilities to the next step – to true automation – with a completely driverless car.

That evolution is in progress in IO’s data centers. Today, there are many ways we have taken monitoring of the data center, enabled by those approximately 1,000 sensors in each data center module, to the next level of control. For example, the IO operating system can control cooling infrastructure including fans and chiller valves. It can control application workloads. So if a server begins to get too hot, the IO.OS can adjust fan speeds and chiller valves to lower the temperature in a module or even below a specific cabinet. If the door of a module is opened more times than a certain threshold (potentially indicating a threat, for example), the IO.OS could shift VM workloads from one module to another.

Our customers are automating aspects of their data center operations, too, using IO’s operating system, IO.OS. Through its API and translator technology, IO.OS can control a device as long as that device is built for automated control, with appropriate sensors and mechanical functions, and as long as that device “speaks” one of the 150+ communication protocols, including BACNet and Modbus, that IO.Translator can translate into OPC, the native language of IO.OS.

io.translator1

In the case where an IO customer is receiving the IO.OS (either independently or part of an IO.Anywhere module installation) but also has legacy data center infrastructure in other locations, the IO.OS can integrate monitoring of both the new modular data center infrastructure and the legacy data center. Data center devices that can be monitored, and potentially controlled, include the chiller, generator, switchboard, power distribution units (PDUs), access controls, and fire protection – among others.

In the case where an IO customer is receiving an IO.Anywhere modular data center installation and has an existing building management system (BMS) or human machine interface (HMI), the IO.OS can be integrated into that existing system to enable integrated control of the data center modules as well as the other assets. For example, for a large defense contractor IO successfully integrated IO.OS and the EDGE modules into the customer’s Johnson Controls BMS. For a large financial institution, IO successfully integrated IO.OS and the CORE modules into the customer’s Siemens BMS.

The benefit of integrating all data center assets under one monitoring and control system is that it provides a single view of data center operations across multiple devices and even vendors.

Data Center Automation Tomorrow

The automation of the data center is really just beginning. We envision a future in which the data center is fully automated – no human intervention necessary. IO.OS already has the intelligence and, in some areas, the control capability. Google is getting humans out of the driver’s seat. We’re getting humans out of the data center.