High-frequency line energy consumption data acquisition method for computing power server

Abstract

The invention discloses a high-frequency line energy consumption data acquisition method for a computing power server. The acquisition method comprises the following steps: S1, determining a measurement target; S2, selecting a sensor; S3, setting data acquisition equipment; before data acquisition is carried out on the high-frequency line, the data acquisition equipment is detected, the problem that normal data acquisition is affected due to the fact that the acquisition equipment cannot be used is avoided, then calibration work is carried out on the data acquisition equipment, sensor data are remotely monitored through the edge server and uploaded to the cloud when needed, and the data acquisition efficiency is improved. According to the method, data transmission safety and privacy protection are ensured, unauthorized data access is prevented, then a data analysis result is presented to a user in a chart mode through a report producer, and the user generates a customized energy consumption analysis report according to requirements, so that the collected data can be quickly analyzed and the result is generated, and the user experience is improved. And the working efficiency of collection work is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of high-frequency line energy consumption data collection, and specifically to a high-frequency line energy consumption data collection method for a computing power server. Background Art

[0002] Computing servers are servers specially used to process large-scale data calculations. They have powerful computing power and high-speed data processing capabilities and are widely used in cloud computing, big data, artificial intelligence, and high-performance computing. When the computing server is working, it is necessary to collect data on high-frequency line energy consumption to facilitate data analysis and optimization strategies. When collecting energy consumption data, each data needs to be collected separately and then merged, and then the data is manually sorted and analyzed by the staff, resulting in slow data analysis and cumbersome operations, which reduces the efficiency of data collection. Therefore, a high-frequency line energy consumption data collection method for computing servers is proposed. Therefore, "the original data collection work is cumbersome, and the data energy consumption collection is not comprehensive", which is specifically manifested in the following aspects:

[0003] (1) When collecting energy consumption data, each data item needs to be collected separately and then merged, and then the data is manually sorted and analyzed by staff, which results in slow data analysis and cumbersome operations, reducing the efficiency of data collection; (2) When collecting data on high-frequency line energy consumption, because the data is collected during the normal working environment, data for different working environments cannot be collected, which makes the data collection one-sided and reduces the accuracy of data collection; (3) When collecting and simulating data, when abnormal data occurs, the staff cannot discover it in time, and cannot handle the fault urgently when a fault occurs, thereby reducing the safety of the operation process. Summary of the invention

[0004] The purpose of the present invention is to provide a high-frequency line energy consumption data collection method for a computing power server, so as to solve the problems of the current high-frequency line energy consumption data collection method proposed in the above background technology: (1) when collecting energy consumption data, it is necessary to collect each data separately and then merge them, and then the data is manually sorted and analyzed by the staff, which leads to slow data analysis and cumbersome operation, thereby reducing the work efficiency of data collection; (2) when collecting data on high-frequency line energy consumption, because the data in the normal working environment is tested, the data in different working environments cannot be collected, which makes the data collection one-sided and reduces the accuracy of data collection; (3) when collecting and simulating data, when the data is abnormal, the staff cannot discover it in time, and when a fault occurs, the fault cannot be urgently handled, thereby reducing the safety of the operation process.

[0005] To achieve the above object, the present invention provides the following technical solution: a method for collecting high-frequency line energy consumption data for a computing power server, the collection method comprising the following steps:

[0006] S1. Determine the measurement target

[0007] 1. Select the server. 2. Determine the collection target, current, voltage, power, transmission efficiency and line loss.

[0008] S2. Select sensor

[0009] 1. Select high-precision sensors, ① temperature and humidity sensor, ② multimeter, ③ power monitoring sensor, ④ active RF sensor, ⑤ fiber optic current sensor, 2. Wiring connection, ① detection sensor, ② cleaning sensor, ③ connecting sensor to high-frequency line.

[0010] S3. Set up data collection equipment

[0011] 1. Select data acquisition equipment, ① inspect data acquisition equipment, ② determine whether the connection is stable, ③ calibrate data acquisition equipment, 2. Use edge server for detection, ① output storage, processing and analysis, and display using charts, ② remote data monitoring and viewing, ③ use report generator.

[0012] S4. Test verification

[0013] 1. Prepare for the test, ① Check normal operation, ② Simulate the environment, ③ Working temperature 10℃-30℃, humidity 20-60%RH, server current 200-800 watts, 2. Detect device connection, ① Connect the device, ② Turn on the power, 3. Start detection and verification, 4. Data collection.

[0014] S5. Alarm mechanism

[0015] 1. Connect the alarm device, ① connect the sound and light alarm to the detection equipment, ② set the stable range of current, voltage and temperature, 2. Emergency stop equipment, ① stop the equipment from working, ② cut off the power supply, 3. Urgent measures, install perfluorohexanone fire extinguishing agent.

[0016] S6, Updates and Improvements

[0017] 1. Update data collection methods and systems, ① conduct regular inspections, ② detect the accuracy and efficiency of data collection, 2. Introduce new sensors, optimize data processing algorithms and improve data visualization tools, ① adopt new sensors, thermal sensors and optical sensors, ② adopt long short-term memory networks and use principal component analysis (PCA).

[0018] Preferably, the low-temperature working environment of the server battery has an ambient temperature of 10° C., a humidity of 20% RH, and a server working current of 300±50 watts.

[0019] Preferably, the normal working environment of the server battery is an ambient temperature of 20° C., a humidity of 40% RH, and a server working current of 500±50 watts.

[0020] Preferably, the high temperature working environment of the server battery is 30° C., 60% RH, and the server working current is 700±50 watts.

[0021] Preferably, the CPU computing power of the computing power server is: single-core performance (such as GHz) × number of cores × number of chips.

[0022] Preferably, the GPU computing power of the computing power server is: video memory bandwidth×number of 32-bit floating-point operations÷7.

[0023] Preferably, the unit of the video memory bandwidth is GB / s, and the number of 32-bit floating-point operations is TOPS.

[0024] Preferably, the calculation formula of the computing power server is A=(Z value^2*σ^2) / (E^2).

[0025] Preferably, in the formula, the Z value is the Z statistic value under the corresponding confidence level, for example, the Z value corresponding to the 95% confidence level is 1.96, σ is the population standard deviation, and E is the allowable error.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] Before collecting data on the high-frequency line, the present invention performs a detection on the data collection device to avoid the problem that the collection device cannot be used and affects normal data collection, and then calibrates the data collection device. Then, the sensor data is remotely monitored through the edge server and uploaded to the cloud when necessary to ensure the security and privacy protection of data transmission and prevent unauthorized data access. Then, the data analysis results are presented to the user in the form of a chart through the report generator, allowing the user to generate a customized energy consumption analysis report according to needs, so that the collected data can be quickly analyzed and the results can be generated, thereby improving the work efficiency of the collection work.

[0028] When collecting data on high-frequency line energy consumption, the present invention connects the equipment and then powers on the computing server. Then, through the environmental simulation box, three modes are set according to the low-temperature working environment, the ambient temperature is 10°C, the humidity is 20%RH, and the server working current is 300±50 watts; the normal working environment, the ambient temperature is 20°C, the humidity is 40%RH, and the server working current is 500±50 watts; and the high-temperature working environment, the ambient temperature is 30°C, the humidity is 60%RH, and the server working current is 700±50 watts. Then, the high-frequency lines are tested and verified one by one, thereby improving the comprehensiveness of the high-frequency line test and verification, and improving the accuracy of the high-frequency line energy consumption data.

[0029] When the high-frequency line is tested and verified, the data acquisition device is connected to the alarm device, and then the sound and light alarm is electrically connected to the sensor. Then, the temperature of the alarm device is set not to exceed 80 degrees Celsius and the current is set not to exceed 900 watts. When the setting is exceeded, a fire is likely to cause damage to the equipment. Then, an alarm is sounded through the alarm to remind the staff to test the terminal and cut off the power supply. Then, when the internal temperature reaches 100 degrees Celsius, the temperature directly triggers the perfluorohexanone fire extinguishing agent, and emergency fire extinguishing measures are taken for the internal electronic equipment, thereby reducing the damage to the internal electronic equipment caused by the fire, thereby improving the safety of the high-frequency line test. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 This is a flow chart of the high-frequency line energy consumption data collection of the computing power server of the present invention;

[0031] Figure 2 A table of test environment parameters for the server in the present invention;

[0032] Figure 3 It is the test logic code of the computing power server in the present invention;

[0033] Figure 4 It is the operating logic code of the computing power server in the present invention. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0035] Embodiment 1:

[0036] Please refer to the figure, a method for collecting high-frequency line energy consumption data for a computing power server, the collection method includes the following steps:

[0037] S1. Determine the measurement target

[0038] 1. Select the server. 2. Determine the collection target, current, voltage, power, transmission efficiency and line loss. At the beginning of the work, select the server to be tested, then connect the server to the power supply, then determine the target to be collected, and then determine the current, voltage, power, transmission efficiency and line loss data required for the high-frequency line energy consumption of the computing server, and select the sensor according to the data to be collected;

[0039] S2. Select sensor

[0040] 1. Select high-precision sensors, ① temperature and humidity sensors, ② multimeters, ③ power monitoring sensors, ④ active RF sensors, ⑤ fiber optic current sensors, 2. Wiring connections, ① detection sensors, ② cleaning sensors, ③ connecting sensors to high-frequency lines. At the beginning of the test, select high-precision sensors according to the data to be collected. Use the temperature and humidity sensors to collect the temperature of the high-frequency line and the humidity data of the server working environment. Then use the multimeter to collect the voltage of the high-frequency line. Then use the power detection sensor to detect the power used by the high-frequency line. Then use the active RF sensor to actively send signals and receive reflected signals. Then use the fiber optic current sensor to detect the current of the high-frequency line. Then install the sensor and clean the sensor before installation to reduce the impact of dust on the sensor. Then install the sensor with the high-frequency line.

[0041] S3. Set up data collection equipment

[0042] 1. Select the data acquisition device, ① check the data acquisition device, ② determine whether the connection is stable, ③ calibrate the data acquisition device, 2. Use the edge server for detection, ① output storage, processing and analysis, and display with charts, ② remote data monitoring and viewing, ③ use the report generator, before collecting data on the high-frequency line, detect the data acquisition device to avoid the problem that the acquisition device cannot be used and affects normal data collection, and then calibrate the data acquisition device, and then connect to the data acquisition device through the edge server, and then remotely monitor the sensor data through the edge server, and upload it to the cloud when necessary to ensure the security and privacy of data transmission, and prevent unauthorized data access, and then present the data analysis results to the user in the form of charts through the report generator, so that the user can generate customized energy consumption analysis reports according to needs, so that the collected data can be quickly analyzed and the results can be generated, which improves the work efficiency of the collection work.

[0043] S4. Test verification

[0044] 1. Prepare for the test, ① detect normal operation, ② simulate the environment, ③ work temperature 10℃-30℃, humidity 20-60%RH, server current 200-800 watts, 2. Detection equipment connection, ① connect the equipment, ② turn on the power, 3. Start detection and verification, 4. Data collection, when collecting data on high-frequency line energy consumption, connect the equipment, and then turn on the power of the computing server, and then use the environmental simulation box to set the low-temperature working environment, ambient temperature 10℃, humidity 20%RH, server working current 300±50 watts, normal working environment, ambient temperature 20℃, humidity 40%RH, server working current 500±50 watts and high-temperature working environment, ambient temperature 30℃, humidity 60%RH, server working current 700±50 watts. Then test and verify the high-frequency lines one by one, thereby improving the comprehensiveness of the high-frequency line test and verification, and improving the accuracy of the high-frequency line energy consumption data.

[0045] S5. Alarm mechanism

[0046] 1. Connect the alarm device, ① connect the sound and light alarm to the detection equipment, ② set the stable range of current, voltage and temperature, 2. Emergency stop the equipment, ① stop the equipment from working, ② cut off the power supply, 3. Urgent measures, install perfluorohexanone fire extinguishing agent, when testing and verifying the high-frequency line, connect the data acquisition equipment to the alarm device, and then electrically connect the sound and light alarm to the sensor, and then set the temperature of the alarm device not to exceed 80 degrees Celsius and the current not to exceed 900 watts. When exceeding the setting, it is easy to cause fire damage to the equipment. Then the alarm is sounded through the alarm to remind the staff to test the terminal and cut off the power supply. Then when the internal temperature reaches 100 degrees Celsius, the temperature directly triggers the perfluorohexanone fire extinguishing agent, and emergency fire extinguishing measures are taken on the internal electronic equipment to reduce the damage to the internal electronic equipment caused by the fire, thereby improving the safety of high-frequency line testing.

[0047] S6, Updates and Improvements

[0048] 1. Update data collection methods and systems, ① conduct regular inspections, ② detect the accuracy and efficiency of data collection, 2. Introduce new sensors, optimize data processing algorithms and improve data visualization tools, ① adopt new sensors, thermal sensors and optical sensors, ② adopt long short-term memory networks and use principal component analysis (PCA). When implementing the detection method, regularly check the sensors and data acquisition equipment to ensure the accuracy of the detection. Then, through the analysis of each test result, update and upgrade the collection method and server system to ensure the comprehensiveness and stability of subsequent data collection. At the same time, new sensors and new data processing algorithms can be used to improve the security and efficiency of data analysis.

[0049] In this embodiment: the low temperature working environment of the server battery is 10° C., the humidity is 20% RH, and the server working current is 300±50 watts.

[0050] In this embodiment: the normal working environment of the server battery is as follows: the ambient temperature is 20°C, the humidity is 40%RH, and the server working current is 500±50W.

[0051] In this embodiment: the high temperature working environment of the server battery is 30° C., 60% RH, and the server working current is 700±50 watts.

[0052] In this embodiment: the CPU computing power of the computing server: single-core performance (such as GHz) × number of cores × number of chips.

[0053] In this embodiment: the GPU computing power of the computing power server: video memory bandwidth×number of 32-bit floating-point operations÷7.

[0054] In this embodiment: the unit of video memory bandwidth is GB / s, and the number of 32-bit floating-point operations is TOPS.

[0055] In this embodiment: the calculation formula of the computing power server is A=(Z value^2*σ^2) / (E^2).

[0056] In this embodiment: Z value in the formula: Z statistic value under the corresponding confidence level, for example, the Z value corresponding to the 95% confidence level is 1.96, σ: population standard deviation, E: allowable error.

[0057] Embodiment 2:

[0058] As shown in the figure, different from the first embodiment, the high-frequency line energy consumption data collection method for a computing power server in this embodiment includes the following steps:

[0059] S1. Determine the measurement target

[0060] 1. Select the server. 2. Determine the collection target, current, voltage, power, transmission efficiency and line loss. At the beginning of the work, select the server to be tested, and then connect the server to the power supply. Then determine the target to be collected, and then determine the current, voltage, power, transmission efficiency and line loss data required to collect the high-frequency line energy consumption of the computing server, and select the sensor according to the data to be collected.

[0061] S2. Select sensor

[0062] 1. Select high-precision sensors, ① temperature and humidity sensors, ② multimeters, ③ power monitoring sensors, 2. Wiring connections. At the beginning of the test, select high-precision sensors according to the data to be collected. Use the temperature and humidity sensors to collect the temperature of the high-frequency line and the humidity data of the server working environment. Then use the multimeter to collect the voltage of the high-frequency line, and then use the power detection sensor to detect the power used by the high-frequency line. Then use the active RF sensor to actively send signals and receive reflected signals. Then use the optical fiber current sensor to detect the current of the high-frequency line. Then install the sensor and clean the sensor before installation to reduce the impact of dust on the sensor. Then install the sensor and the high-frequency line.

[0063] S3. Set up data collection equipment

[0064] 1. Select the data acquisition device, ① inspect the data acquisition device, ② determine whether the connection is stable, ③ calibrate the data acquisition device. Before collecting data on the high-frequency line, test the data acquisition device to avoid the problem that the acquisition device cannot be used and affects normal data collection. Then calibrate the data acquisition device and then connect it to the data acquisition device through the edge server.

[0065] S4. Updates and improvements

[0066] 1. Update data collection methods and systems, ① conduct regular inspections, ② detect the accuracy and efficiency of data collection, 2. Introduce new sensors, optimize data processing algorithms and improve data visualization tools, ① adopt new sensors, thermal sensors and optical sensors, ② adopt long short-term memory networks and use principal component analysis (PCA). When implementing the detection method, regularly check the sensors and data acquisition equipment to ensure the accuracy of the detection. Then, through the analysis of each test result, update and upgrade the collection method and server system to ensure the comprehensiveness and stability of subsequent data collection. At the same time, new sensors and new data processing algorithms can be used to improve the security and efficiency of data analysis.

[0067] In summary: compared with embodiment 2, before collecting data on the high-frequency line, the data collection device is tested to avoid the problem that the collection device cannot be used and affects normal data collection, and then the data collection device is calibrated, and then connected to the data collection device through the edge server, and then the sensor data is remotely monitored through the edge server and uploaded to the cloud when necessary to ensure the security and privacy of data transmission and prevent unauthorized data access, and then the data analysis results are presented to the user in the form of charts through the report generator, allowing the user to generate customized energy consumption analysis reports according to needs, so that the collected data can be quickly analyzed and the results can be generated, thereby improving the work efficiency of the collection work.

[0068] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0069] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for collecting high-frequency line energy consumption data for a computing power server, characterized in that the collection method comprises the following steps: S1. Determine the measurement target 1. Select the server.

2. Determine the collection target, current, voltage, power, transmission efficiency and line loss. S2. Select sensor 1. Select high-precision sensors, ① temperature and humidity sensor, ② multimeter, ③ power monitoring sensor, ④ active RF sensor, ⑤ fiber optic current sensor, 2. Wiring connection, ① detection sensor, ② cleaning sensor, ③ connecting sensor to high-frequency line; S3. Set up data collection equipment 1. Select data acquisition equipment, ① check data acquisition equipment, ② determine whether the connection is stable, ③ calibrate data acquisition equipment, 2. Use edge server for detection, ① output storage, processing and analysis, and display using charts, ② remote data monitoring and viewing, ③ use report generator; S4. Test verification 1. Prepare for testing, ① Test normal operation, ② Simulate environment, ③ Working temperature 10℃-30℃, humidity 20-60%RH, server current 200-800W, 2. Test equipment connection, ① Connect the equipment, ② Turn on the power, 3. Start testing and verification, 4. Data collection; S5. Alarm mechanism 1. Connect the alarm device, ① connect the sound and light alarm to the detection equipment, ② set the stable range of current, voltage and temperature, 2. Emergency stop the equipment, ① stop the equipment from working, ② cut off the power supply, 3. Urgently seek measures, install perfluorohexanone fire extinguishing agent; S6, Updates and Improvements 1. Update data collection methods and systems, ① conduct regular inspections, ② detect the accuracy and efficiency of data collection, 2. Introduce new sensors, optimize data processing algorithms and improve data visualization tools, ① adopt new sensors, thermal sensors and optical sensors, ② adopt long short-term memory networks and use principal component analysis (PCA).

2. According to claim 1, a high-frequency line energy consumption data collection method for a computing power server is characterized in that: The low-temperature working environment of the server battery is 10° C., 20% RH, and the server working current is 300±50 watts.

3. The high-frequency line energy consumption data collection method for a computing power server according to claim 1 is characterized in that: The normal working environment of the server battery is as follows: the ambient temperature is 20° C., the humidity is 40% RH, and the server working current is 500±50 watts.

4. The high-frequency line energy consumption data collection method for a computing power server according to claim 1 is characterized in that: The high temperature working environment of the server battery is 30° C., 60% RH, and the server working current is 700±50 watts.

5. The high-frequency line energy consumption data collection method for a computing power server according to claim 1 is characterized in that: The CPU computing power of the computing server is: single-core performance (such as GHz) × number of cores × number of chips.

6. The high-frequency line energy consumption data collection method for a computing power server according to claim 1, characterized in that: The GPU computing power of the computing server is: video memory bandwidth × number of 32-bit floating-point operations ÷ 7.

7. The method for collecting high-frequency line energy consumption data for a computing power server according to claim 6, characterized in that: The unit of the video memory bandwidth is GB / s, and the number of 32-bit floating point operations is TOPS.

8. The method for collecting high-frequency line energy consumption data for a computing power server according to claim 1, characterized in that: The calculation formula of the computing power server is A=(Z value^2*σ^2) / (E^2).

9. The method for collecting high-frequency line energy consumption data for a computing power server according to claim 8, characterized in that: In the formula, Z value: Z statistic value under the corresponding confidence level, for example, the Z value corresponding to 95% confidence level is 1.96, σ: population standard deviation, E: allowable error.

Images