U-level iot combined with CFD to fine monitor the temperature of the equipment room

Data center operation and maintenance, like cold and afraid of heat. In the jungle of the server, hot spots are everywhere, how to adopt a secure RFID new technology solution, starting from the real physical temperature monitoring of each server, and accurately obtain the hot spot cloud map of the machine room is a new topic that crosses the traditional technical solution, makes the operation and maintenance personnel easily cope with the daily work, and effectively protects the security of user assets.

First, the method of finding hot spots in the computer room

At present, the "thermal management" of most data centers is still in the stage of passive operation and maintenance, that is, the hot spot is found first, and then processed, the consequences of this thinking mode may be "hot after cold", and the slightest mistake will lead to a big mistake.

From a technical point of view, this can't be left to IT management or operations personnel. In the current technical environment, there are usually three ways for personnel to find hot spots in the equipment room:

1. Observe and measure the temperature of the server or cabinet in the equipment room.

Advantages: economical, simple, the most effective for identifying hot spots with high temperature;

Disadvantages: poor accuracy, time consuming, energy consuming, large radiation.

2. Measure the temperature of the cabinet or server using a handheld temperature measuring instrument.

Advantages: economical, effective, high temperature detection accuracy;

Disadvantages: physical exertion, radiation.

3. Automatic detection method, including DCIM automatic monitoring device or CFD software to predict hot spots.

Advantages: Automatic detection, save time and labor, can grasp the overall hot spot distribution of the machine room;

Disadvantages: The hardware and software cost is large, and because of the large particle size, the micro hot spot layout is not accurate.

Second, the pain points of hot spot analysis

From the development trend of intelligent operation and maintenance of data centers, automatic detection methods will become the direction of hot spot detection in the future.

However, at present, the DCIM detects the temperature through the sensors at a few important detection points or the IPMI port of the server.

1. If the monitoring is stable through the sensor, the particle size will be relatively large, and many hot spots cannot be found in time;

2. If the IPMI port is used to obtain temperature data, there may be security compliance risks, complicated deployment, and only the CPU temperature can be monitored.

Third, U-level iot combined with CFD technical solution

U-level iot positioning technology is a safe and efficient technical solution to solve the server positioning and data reading accuracy. Through the first MC-RFID technology in the digital human industry, the digital management of cabinet U-bit resources, capacity, location, status, information change, temperature and humidity is realized, and IT managers can control the real-time information of U-bit assets globally through the back-end visualization platform.

6igmaRoom, CFD software launched by UK Future Facilities Company, is a data center thermal environment analysis tool. It can predict whether hot spots will occur within the whole data center and during long-term operation, and analyze the causes of hot spots from the perspective of air flow organization. This guide guides O&M personnel to optimize device deployment positions and adjust air conditioning cooling capabilities.

Future Facilities' 6igmaRoom product, combined with Digitalor U-level Asset iot system, can provide users with real-time and accurate asset deployment data - IT equipment deployment information can be synchronized in real time without manual input, and U-level asset reading data is 100% accurate. The temperature and humidity in the cabinet are measured in real time and reported.

The method of detecting the hot spot in the cabinet of the joint solution is to detect the temperature of the server and its surroundings in real time through the intelligent control module of the Digitalor. When the location of the server changes, the U-level iot system will update the information in time and re-detect the temperature information in the cabinet. Then, 6igmaRoom will use the latest data to conduct real-time model analysis, so as to predict possible hot spots, help operation and maintenance personnel to discover risks in advance and prevent them in time.

Fourth, the decline in operation and maintenance costs is a trend

According to Gartner, the cost composition of data center asset operation and maintenance management is mainly divided into three parts:

Process maintenance costs: including daily maintenance, change management, testing costs, etc.

Operation and maintenance personnel costs: including recruitment, training, education, personnel turnover costs, etc.;

Technology and product costs: including hardware, software, network costs, etc.

Among them, personnel costs are close to half of operating expenses. Therefore, how to realize the replacement of human work by machines and reduce the loss caused by human error through technology and product innovation has become an important method for data center operations to save costs.

The combination of U-level iot system and CFD real-time simulation system is an effective practice to solve the shortage of personnel and high cost of data center operation and maintenance, and has great practical significance to improve the automated thermal management capability of data center and reduce the complexity of operation and maintenance.