No Jitter is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Data Center IoT

The reality is that data centers and network closets are not uniformly controlled; there are hot spots and humidity differences that can be more easily monitored and managed with things like temperature sensors. Security sensors can also be used to alert when cabinets are accessed. With the use of IoT, the data center and network closets can become more secure and attain higher reliability while also reducing costs.

Sensors are data producers. The more traditional definition of "sensor" is a device that collects information about the physical environment. In some cases that information might be logged, read and/or transmitted in real-time through a network.

A sensor is less about a specific technology deployed and more about the data produced -- anything and everything that can be measured. Sensors can measure the state of things, where "state" might mean temperature, humidity, shock, vibration, tilt, altitude, weight, specific gravity, barometric pressure, etc.

If you want to change something like increase or decrease airflow or temperature, set off an alarm, or lock a cabinet/door, then you need an actuator. The actuator controls something in response to a command, usually as a result of what the sensor data processor determines.

The data from sensors can be used to prevent overcooling, undercooling, electrostatic discharge, corrosion, and even short circuits. The sensor processing will help reduce operational costs, defer or decrease CAPEX, increase uptime, and may even improve the capacity to support future growth.

Environmental monitoring with alerts can reduce potential problems such as the presence of water, smoke, and open cabinet doors that could impair cooling.

Data centers and network closets commonly have room security access controls; most cabinets do not. Consider installing contact closure sensors on the cabinets.

Besides your employees, you will have others like contractors visit your site. When a sensor detects that a cabinet door has been opened, a webcam can be used to snap a picture of the person. Sensors have provisions for third-party devices like smoke detectors. You can ensure secure equipment changes. Sensors can have a keep-alive function or have a scheduled data transfer. In either case, the lack of security sensor data can mean that a security breach has occurred.

Airflow sensors monitor cooling airflow and the returned hot air, ensuring the cooling system is properly functioning and producing airflow at the correct level for the entire cabinet.

Temperature and humidity sensors are designed to follow ASHRAE sensor placement guidelines for accurate and complete readings. (A good reference to learn ASHRAE's guidelines and recommendations is "Thermal Guidelines for Data Processing Environments, Third Edition."

Differential air pressure sensors can identify air pressure differences that can locate hot/cold aisle partition leaks. When cooling air is under the floor, air pressure sensors provide feedback for the cooling system and building management system. Fan speed can be controlled as necessary to satisfy the underfloor pressure.

The cooling of a data center consumes about 43% of the electrical bill. A direct financial value of sensors is that you can save up to 4% in energy costs for every degree of increase in the baseline temperature, an estimate made by Gartner. Cabinet temperature sensors provide far more accurate real-time data center temperatures when compared to cooling system or room thermometer readings.

IT workloads affect the temperature. The more work that is performed, the higher the temperature produced by the equipment. Temperature information can be used to optimize cooling system operation. The shutdown one or more cooling units will reduce costs in real time.

Reliability is measured as the Mean Time Between Failures (MTBF). A second important factor is the Mean Time To Repair (MTTR). Data center availability is important, usually expressed as 99.9+% uptime. The reliability of IT equipment drops 50% when the temperature of the equipment is 88 degrees. This is how hot spots can significantly impact data center and closet reliability and availability. You want the hotter areas of the equipment to operate close to 70 degrees.

Humidity sensors can be used to maintain proper humidity levels and avoid electrostatic discharge problems when humidity is low or condensation problems when the humidity is excessive. Water sensors detect water leaks from external sources or from pipes in a water-cooled rack. Thresholds and alerts can be set when monitoring onsite, remote, or lights-out facilities to ensure equipment is operating in safe and reliable conditions.

Smart cooling systems incorporate a number of sensors to track rack power use and heat generation. Sensors should be checked and calibrated on a regular maintenance schedule. Failed sensors and false readings can waste a considerable amount of energy. An energy management system can be utilized to ensure that certain equipment at the facility will not peak and/or operate at the same time as other equipment.

Setting clear and measurable goals is critical for understanding intended results, developing effective strategies, and achieving energy reductions. Well-stated goals guide daily decision-making and are the basis for tracking and measuring progress.

Other blogs on this subject include Data Center Outages; The Costs, Measuring Sustainability: Certification Value, Part 1, and Keeping the Closet Cool.