Server heat dissipation control method, device and computer storage medium

Abstract

The present invention relates to the technical field of server heat dissipation, and in particular to a method, device and computer storage medium for heat dissipation control of a server. The method comprises: obtaining temperature values, access flow data to be processed and processing speed data; determining the state value of the server; forward displacing the temperature sequence, and performing temperature delay correction according to the correlation between the displaced sequence and the state sequence obtained after the dislocation, to obtain a target sequence; determining the abnormal change moment according to the target sequence and the state sequence, and then determining the temperature change time period, and determining the change extension index of each abnormal change moment according to the numerical distribution of the temperature value in each temperature change time period; determining the heat dissipation power at the current moment according to the change extension index of the abnormal change moment before the current moment, and the state value and temperature value at the current moment. The present invention can reduce the delay effect of temperature control, optimize the operation strategy of heat dissipation, realize active heat dissipation control, and improve energy efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of server heat dissipation, and in particular to a server heat dissipation control method, device and computer storage medium. Background Art

[0002] With the rapid development of information technology and the growing demand for high-performance computing such as cloud computing and artificial intelligence, server performance and power consumption issues have attracted more and more attention. In high-density computing environments, server heat dissipation management has become an important factor in ensuring stable system operation and extending equipment life. Traditional heat dissipation control methods mainly rely on fixed cooling solutions, which usually cannot effectively cope with instantaneous load changes, resulting in energy waste and potential overheating risks.

[0003] In the related technology, by configuring a temperature sensor, real-time heat dissipation regulation is achieved according to the temperature data of the temperature sensor. However, due to the excessive power consumption of the server, the temperature rises and is transmitted to the temperature sensor to produce a reaction. The system needs time to respond and initiate cooling measures, which has a certain delay effect. This delay effect will cause large temperature fluctuations, resulting in unstable heat dissipation regulation effect and low cooling efficiency. Summary of the invention

[0004] In order to solve the technical problem that the cooling measures in the related art have a certain delay effect, which causes large temperature fluctuations, thereby causing unstable heat dissipation control effect and low cooling efficiency, the present invention provides a server heat dissipation control method, device and computer storage medium, and the technical solutions adopted are as follows:

[0005] The present invention proposes a server heat dissipation control method, the method comprising:

[0006] Obtaining the temperature value of the server at different times, the access flow data to be processed, and the processing speed data; determining the state value of the server at each time according to the access flow data and the processing speed data at each time; sorting the state values ​​according to time to obtain a state sequence, and sorting the temperature values ​​to obtain a temperature sequence;

[0007] The temperature sequence is shifted forward according to the time, and the temperature delay correction sequence with the strongest correlation is determined according to the correlation between the shifted sequence obtained after the shifting of different sequence bits and the state sequence; each moment of the temperature value in the temperature delay correction sequence is adjusted to be the same as the corresponding one in the state sequence to obtain the target sequence;

[0008] According to the value changes of the target sequence and the state sequence at adjacent moments, the abnormal change moment is determined, the temperature change time period is determined with the abnormal change moment as the reference point, and the change extension index of each abnormal change moment is determined according to the value distribution of the temperature value in each temperature change time period;

[0009] The heat dissipation power at the current moment is determined according to the change extension index at the abnormal change moment before the current moment, and the state value and temperature value at the current moment.

[0010] Further, determining the state value of the server at each moment according to the access traffic data and processing speed data at each moment includes:

[0011] The product of the access flow data and the processing speed data is used as the state value of the server at each moment.

[0012] Furthermore, the forward shifting is to shift the temperature value to the left in time sequence according to different sequence digits, and delete the temperature value of the corresponding sequence digit before the temperature sequence to obtain a shifted sequence corresponding to each sequence digit; wherein, according to the correlation between the shifted sequence obtained after the shifting of different sequence digits and the state sequence, the temperature delay correction sequence with the strongest correlation is determined, including:

[0013] Calculating the Pearson correlation coefficient between the dislocation sequence and the state sequence, and performing normalization processing to obtain a correlation index;

[0014] The misaligned sequence with the largest correlation index value is used as the temperature delay correction sequence.

[0015] Further, determining the abnormal change moment according to the value changes of the target sequence and the state sequence at adjacent moments includes:

[0016] Take any data moment as the moment to be analyzed, and take a preset number of other moments that are closest to the moment to be analyzed as comparison moments;

[0017] According to the temperature difference between the time to be analyzed and all the comparison times in the target sequence, the neighborhood temperature change index of the time to be analyzed is determined;

[0018] Determine the neighborhood state change index of the time to be analyzed according to the difference in state values ​​between the time to be analyzed and all the comparison times in the state sequence;

[0019] Calculate the product of the neighborhood temperature change index and the neighborhood state change index at the time to be analyzed, and normalize them as the state temperature impact index;

[0020] The moment when the value of the state temperature impact index is greater than a preset impact threshold is regarded as the abnormal change moment.

[0021] Furthermore, the method for obtaining the neighborhood temperature change index and the neighborhood state change index at the time to be analyzed includes:

[0022] In the target sequence, the absolute value of the difference between the temperature value corresponding to the time to be analyzed and each comparison time is calculated as the temperature change value between the time to be analyzed and the corresponding comparison time;

[0023] The average of the temperature change values ​​at the time to be analyzed and all the comparison times is used as the neighborhood temperature change index at the time to be analyzed;

[0024] In the state sequence, the absolute value of the difference between the state value corresponding to the time to be analyzed and each comparison time is calculated as the state change value between the time to be analyzed and the corresponding comparison time;

[0025] The average of the state change values ​​of the time to be analyzed and all the comparison times is used as the neighborhood state change index of the time to be analyzed.

[0026] Further, determining the temperature change time period with the abnormal change moment as a reference point includes:

[0027] The preset time range with the abnormal change moment as the center point is used as the temperature change time period corresponding to the abnormal change moment.

[0028] Furthermore, the step of determining the change extension index at each abnormal change moment according to the numerical distribution of the temperature values ​​in each temperature change time period includes:

[0029] Calculate the range of temperature values ​​within the same temperature change period;

[0030] The standard deviation of all temperature values ​​within the same temperature change period is taken as the change trend influence coefficient;

[0031] The product of the range and the change trend influence coefficient is calculated and normalized as a change extension index corresponding to the abnormal change moment.

[0032] Further, the determining of the heat dissipation power at the current moment according to the change extension index at the abnormal change moment before the current moment, and the state value and temperature value at the current moment, includes:

[0033] The change abnormal moment closest to the current moment in time sequence is taken as the current impact moment, and the time interval between the current impact moment and the current moment is determined as the impact interval; the ratio of the change extension index of the current impact moment to the impact interval is calculated as the change impact degree of the current moment;

[0034] Perform maximum and minimum normalization processing on the state value and temperature value at the current moment respectively to obtain the state parameter and temperature parameter;

[0035] The difference between the state parameter and the temperature parameter is used as the indicator to be changed;

[0036] Calculating the product of the indicator to be changed and the degree of influence of the change, and normalizing the product to be used as the power adjustment weight;

[0037] The product of the power adjustment weight and the preset power at the current moment is calculated to obtain the heat dissipation power at the current moment.

[0038] On the other hand, the present invention also provides a server heat dissipation control system, the system comprising:

[0039] An acquisition module is used to acquire the temperature value of the server at different times, the access flow data to be processed, and the processing speed data; determine the state value of the server at each time according to the access flow data and the processing speed data at each time; sort the state values ​​according to the time to obtain a state sequence, and sort the temperature values ​​to obtain a temperature sequence;

[0040] The delay analysis module is used to shift the temperature sequence forward according to the time, determine the temperature delay correction sequence with the strongest correlation according to the correlation between the shifted sequence and the state sequence after the shifting of different sequence bits; adjust each moment of the temperature value in the temperature delay correction sequence to be the same as the corresponding one in the state sequence to obtain the target sequence;

[0041] A temperature change determination module is used to determine the abnormal change moment according to the value changes of the target sequence and the state sequence at adjacent moments, determine the temperature change time period with the abnormal change moment as the reference point, and determine the change extension index of each abnormal change moment according to the value distribution of the temperature value in each temperature change time period;

[0042] The heat dissipation control module is used to determine the heat dissipation power at the current moment according to the change extension index at the abnormal change moment before the current moment, as well as the state value and temperature value at the current moment.

[0043] On the other hand, the present invention further provides a computer-readable storage medium having a computer program stored thereon, and when the computer program is executed, the steps of any of the aforementioned server heat dissipation control methods are implemented.

[0044] The present invention has the following beneficial effects:

[0045] The embodiment of the present invention analyzes the temperature and state at different times, and then analyzes the correlation between temperature change and state change through dislocation, adjusts the timing characteristics of the temperature sequence to correct the delay effect of the temperature, thereby obtaining a more accurate and effective target sequence of temperature change at different times, and determines the abnormal change moment through the numerical change in the target sequence and the numerical change in the state sequence. The abnormal change moment characterizes the moment when the temperature and state suddenly change, and analyzes the abnormal change moment in the timing neighborhood to analyze the instability of the temperature change at each abnormal change moment, and determines the change extension index. The heat dissipation power at the current moment is adaptively adjusted through the change extension index of the abnormal change moment before the current moment, as well as the state value and temperature value at the current moment. The present invention can eliminate the influence of the temperature delay effect, effectively reduce the delay effect of temperature control, and determine the heat dissipation power that is more suitable for the current moment in combination with the temperature change and the state change, so as to effectively reduce the temperature fluctuation and improve the cooling efficiency, thereby optimizing the heat dissipation operation strategy, realizing active heat dissipation control, and improving the energy efficiency ratio. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0047] Figure 1 A flow chart of a server heat dissipation control method provided by an embodiment of the present invention;

[0048] Figure 2 A schematic diagram of the principle of a server liquid cooling and heat dissipation system provided by an embodiment of the present invention;

[0049] Figure 3 A schematic diagram of a CPU occupancy rate curve provided by an embodiment of the present invention;

[0050] Figure 4 A schematic diagram of a CPU processing speed curve provided by an embodiment of the present invention;

[0051] Figure 5 A schematic diagram of server status values ​​at different times provided by an embodiment of the present invention;

[0052] Figure 6 A schematic diagram of the change of correlation under different sequence bit numbers provided by an embodiment of the present invention;

[0053] Figure 7 A structural diagram of a server heat dissipation control system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0054] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation, structure, features and effects of a server heat dissipation control method, device and computer storage medium proposed in accordance with the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0055] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0056] In order to enable those skilled in the art to understand the content of the present invention, some basic terms will be explained and elaborated in detail below.

[0057] The server is mainly used to store, read and process data. The server in this solution is a physical server. The hard disk in the physical server reads and writes data at high speed, and the power supply converts electrical energy, which will generate heat. The accumulation of heat from many components causes the server to heat up.

[0058] The heat dissipation of the server generally adopts the physical cooling method. When the temperature of the server is collected by the temperature sensor, when the temperature exceeds a certain threshold, the system will issue an alarm and trigger the corresponding cooling operation.

[0059] A specific solution of a server heat dissipation control method provided by the present invention is described in detail below with reference to the accompanying drawings.

[0060] See also Figure 1 , which shows a flow chart of a server heat dissipation control method provided by an embodiment of the present invention, the method comprising:

[0061] S101: Obtain the temperature value of the server at different times, the access traffic data to be processed, and the processing speed data; determine the state value of the server at each time according to the access traffic data and the processing speed data at each time; sort the state values ​​according to time to obtain a state sequence, and sort the temperature values ​​to obtain a temperature sequence.

[0062] It is understandable that servers are usually in a closed environment, and the heat dissipation of the server is generally not affected by changes in the environment. The heat of the server mainly comes from the heat generated by components such as the processor and power supply. When performing complex calculations, data analysis or graphics rendering, the heat generated will increase significantly.

[0063] As computing density and data processing requirements continue to increase, servers generate a lot of heat during operation. If heat dissipation is not properly managed, it is easy to cause system overheating, which will affect the stability and performance of the equipment and even shorten its service life. Overheating can not only cause hardware failures, but also cause system crashes, data loss, and even have a significant impact on business operations.

[0064] Traditional fixed cooling methods cannot meet the rapidly changing load requirements and environmental conditions. As the workload fluctuates, the server's cooling requirements will also change. Adaptive adjustment technology can monitor temperature and load conditions in real time, dynamically adjust cooling strategies, and ensure that the optimal temperature level is always maintained under different working conditions. This not only improves cooling efficiency and reduces energy consumption, but also prevents overheating, thereby protecting the safety and reliability of hardware equipment. Figure 2 , Figure 2 A schematic diagram of the principle of a server liquid cooling system provided by an embodiment of the present invention.

[0065] When controlling the heat dissipation temperature of the server, a temperature sensor may be installed inside the server to monitor the temperature changes of different components (such as CPU, GPU, etc.) in real time and analyze the current heat dissipation requirements. That is, by installing a temperature sensor inside the server to obtain temperature values ​​at different times, in the embodiment of the present invention, if only one temperature sensor is installed, the corresponding temperature sensor reading is directly obtained as the temperature value, and if multiple temperature sensors are installed, the average of all temperature sensor readings is used as the temperature value.

[0066] In the embodiment of the present invention, in order to analyze the load conditions at different times, it is also necessary to obtain the access flow data and processing speed data to be processed at different times, and the data is directly obtained from the server.

[0067] The present invention specifically obtains access flow data and processing speed data based on the CPU status. Figure 3 and Figure 4 , Figure 3 A schematic diagram of a CPU occupancy rate curve provided by an embodiment of the present invention. Figure 4 A schematic diagram of a CPU processing speed curve provided by an embodiment of the present invention. The CPU occupancy rate represents the size of the access traffic, and the CPU processing speed represents the size of the processing speed.

[0068] The embodiment of the present invention aims to dynamically adjust the heat dissipation measures according to the temperature value and the operating status of the server, such as adjusting the speed of the fan, starting or shutting down the liquid cooling system, etc., so as to avoid the occurrence of over-temperature.

[0069] It should be noted that in the embodiment of the present invention, the temperature value, the access flow data to be processed and the processing speed data can be periodically obtained at the same time. For example, the acquisition frequency is 20 Hz, that is, the temperature value, the access flow data and the processing speed data are obtained every 0.05 seconds. Of course, the time interval of data acquisition is adjusted according to the actual detection needs, and there is no restriction on this.

[0070] In order to indicate the running status of the server, it is necessary to determine the status value of the server at each moment according to the access flow data and the processing speed data at each moment. Further, in some embodiments of the present invention, determining the status value of the server at each moment according to the access flow data and the processing speed data at each moment includes: taking the product of the access flow data and the processing speed data as the status value of the server at each moment.

[0071] The access traffic data refers to the traffic data to be processed that is accumulated at a corresponding moment, which can specifically represent the amount of access data. The larger the amount of access data, the greater the load on the server.

[0072] Among them, the processing speed data represents the processing capacity of the server. If the processing speed data value is high, it means that the processing speed is fast. When high-speed processing is performed, the server will generate heat due to high-speed operation, which will make the server heat up more seriously.

[0073] Therefore, in the embodiment of the present invention, the product of the access flow data and the processing speed data at the same time is calculated to obtain the status value of the server. Figure 5 , Figure 5 This is a schematic diagram of the status values ​​of a server at different times provided by an embodiment of the present invention. The status value represents the current operating status of the server, and the larger the value, the more serious the server heating.

[0074] In the embodiment of the present invention, the state values ​​are sorted according to time to obtain a state sequence, and the temperature values ​​are sorted to obtain a temperature sequence.

[0075] S102: The temperature sequence is shifted forward according to the time, and the temperature delay correction sequence with the strongest correlation is determined according to the correlation between the shifted sequences obtained after the shifting of different sequence bits and the state sequence; each moment of the temperature value in the temperature delay correction sequence is adjusted to be the same as the corresponding temperature value in the state sequence, so as to obtain the target sequence.

[0076] It should be noted that since temperature feedback has a delay effect, the analysis of status and temperature in related technologies does not reflect this delay effect. That is to say, when the server is operating at high power consumption, it generates severe heat, but it takes a certain amount of time for the heat to be transferred to the temperature sensor and the corresponding cooling system to generate a cooling plan for the cooling operation. If cooling cannot be performed in advance, overheating may occur during the delay time, causing the server to enter the overheating protection stage.

[0077] Therefore, in order to ensure the timeliness of server heat dissipation, it is also necessary to analyze the delay effect. Since the delay effects of different servers in different scenarios are inconsistent, the analysis needs to be adaptively processed for different servers.

[0078] Among them, the forward shift is to shift the temperature value to the left in time series according to different sequence digits, and delete the temperature value of the corresponding sequence digit in front of the temperature sequence to obtain a shifted sequence corresponding to each sequence digit. It can be understood that since the effect of temperature is delayed, that is, the state change occurs first, which leads to the temperature change, and there is a certain lag in the middle. Therefore, it is necessary to analyze the temperature delay in the embodiment of the present invention.

[0079] The specific analysis process is to align the leftmost (first) data point of the temperature sequence with the leftmost (first) data point of the state sequence, calculate the correlation between the two data sequences, then translate the temperature sequence to the left by one data point position, calculate the correlation between the two data sequences after one translation, and then continue to translate and calculate the correlation between the two data sequences, so as to determine the translation effect with the strongest correlation based on the correlation, which corresponds to the temperature delay correction sequence.

[0080] For example, if the state sequence is {1, 2, 3, 4, 5} and the temperature sequence is {6, 7, 8, 9, 0}, the first forward shift is to get {7, 8, 9, 0, 0} (if the number of digits is insufficient, the last digit is used to complete the gap), and the correlation between {1, 2, 3, 4, 5} and {7, 8, 9, 0, 0} is calculated. The second forward shift is to get {8, 9, 0, 0, 0}, and the correlation between {1, 2, 3, 4, 5} and {8, 9, 0, 0, 0} is calculated, and so on.

[0081] Furthermore, in some other embodiments of the present invention, the temperature delay correction sequence with the strongest correlation is determined based on the correlation between the misaligned sequence and the state sequence obtained after misaligning different sequence bit numbers, including: calculating the Pearson correlation coefficient of the misaligned sequence and the state sequence, and normalizing them to obtain a correlation index; and taking the misaligned sequence with the largest correlation index value as the temperature delay correction sequence.

[0082] Among them, the Pearson correlation coefficient is a correlation calculation method well known to those skilled in the art. In an embodiment of the present invention, the Pearson correlation coefficient can be used to realize the calculation process of the correlation between the misaligned sequence and the state sequence. The larger the value of the Pearson correlation coefficient, the more positive the correlation. Therefore, the Pearson correlation coefficient can be directly normalized to obtain the correlation index.

[0083] See also Figure 6 , Figure 6 A schematic diagram of the change in correlation under different sequence bit numbers provided for an embodiment of the present invention. It can be understood that the larger the value of the correlation index is, the stronger the positive correlation between the misaligned sequence and the state sequence is, and the more consistent with the normal temperature-state change is. Therefore, the misaligned sequence with the largest correlation index value is used as the temperature delay correction sequence.

[0084] The temperature delay correction sequence may represent a temperature sequence obtained after correction to eliminate the delay effect, that is, when a state value changes at a certain moment, it also represents a change in the temperature value in the temperature delay correction sequence at the same moment.

[0085] S103: Determine the abnormal change moment according to the numerical changes of the target sequence and the state sequence at adjacent moments, determine the temperature change time period with the abnormal change moment as the reference point, and determine the change extension index of each abnormal change moment according to the numerical distribution of the temperature values ​​in each temperature change time period.

[0086] After analyzing the correlation to obtain the temperature delay correction sequence, it is necessary to conduct a specific analysis at each moment, that is, to analyze the moment when the temperature and state suddenly change, and determine the impact of different state changes on the temperature through the moment of the sudden change.

[0087] Furthermore, in some embodiments of the present invention, the abnormal change moment is determined according to the numerical change of the target sequence and the state sequence at adjacent moments, including: taking any data moment as the moment to be analyzed, and taking a preset number of other moments that are after the moment to be analyzed and closest to the moment to be analyzed as comparison moments; determining the neighborhood temperature change index of the moment to be analyzed according to the difference in temperature values ​​between the moment to be analyzed and all the comparison moments in the target sequence; determining the neighborhood state change index of the moment to be analyzed according to the difference in state values ​​between the moment to be analyzed and all the comparison moments in the state sequence; calculating the product of the neighborhood temperature change index of the moment to be analyzed and the neighborhood state change index, and normalizing it as the state temperature influence index; taking the moment when the value of the state temperature influence index is greater than the preset influence threshold as the abnormal change moment.

[0088] Among them, the preset number is the number of comparison moments corresponding to the moment to be analyzed. Specifically, the preset number is 10. In the embodiment of the present invention, the other 10 moments that are closest to and after the moment to be analyzed are used as comparison moments. That is to say, for example, when data is collected every 0.1 seconds, the moments within 1 second after the moment to be analyzed are used as comparison moments, so as to facilitate the analysis of temperature changes and state changes at the moment to be analyzed.

[0089] Among them, the neighborhood temperature change index can be determined by numerical analysis of temperature value changes, and the neighborhood state change index can be determined by numerical analysis of state value changes. Further, in some embodiments of the present invention, the method for obtaining the neighborhood temperature change index and the neighborhood state change index at the time to be analyzed includes: in the target sequence, calculating the absolute value of the difference between the temperature value corresponding to the time to be analyzed and each comparison time as the temperature change value between the time to be analyzed and the corresponding comparison time; taking the average of the temperature change values ​​between the time to be analyzed and all comparison times as the neighborhood temperature change index at the time to be analyzed; in the state sequence, calculating the absolute value of the difference between the state value corresponding to the time to be analyzed and each comparison time as the state change value between the time to be analyzed and the corresponding comparison time; taking the average of the state change values ​​between the time to be analyzed and all comparison times as the neighborhood state change index at the time to be analyzed.

[0090] Among them, the temperature change value and the state change value both represent the data difference between the time to be analyzed and the comparison time. The corresponding data differences are averaged to obtain the neighborhood temperature change index and the neighborhood state change index.

[0091] That is to say, the larger the value of the neighborhood temperature change index is, the greater the temperature difference between the time to be analyzed and the comparison time is, and the subsequent temperature will fluctuate greatly at the time to be analyzed. The larger the value of the neighborhood state change index is, the greater the state difference between the time to be analyzed and the comparison time is, and the subsequent server state will fluctuate greatly at the time to be analyzed.

[0092] Therefore, the product of the neighborhood temperature change index and the neighborhood state change index at the time to be analyzed is directly calculated and normalized as the state temperature impact index. The state temperature impact index is the index information obtained by combining the state change and the temperature change. In the embodiment of the present invention, the larger the value of the state temperature impact index is, the greater the state change and temperature change are at the time to be analyzed.

[0093] Among them, the preset impact threshold is the threshold value of the state temperature impact index. In the embodiment of the present invention, the preset impact threshold can be specifically 0.5, that is, the moment when the value of the state temperature impact index is greater than 0.5 is regarded as the moment of abnormal change.

[0094] Among them, the abnormal change moment refers to the moment when a large temperature fluctuation and state fluctuation occur within a local time range. When performing power regulation, such moments need to generate greater attention in order to perform more effective power regulation.

[0095] Further, in some embodiments of the present invention, determining the temperature change time period with the abnormal change moment as a reference point includes: taking a preset time range with the abnormal change moment as a center point as the temperature change time period corresponding to the abnormal change moment.

[0096] Among them, the preset time range is the neighborhood time range corresponding to the moment of abnormal change. It should be noted that not only analysis is needed after the moment of abnormal change, but also specific analysis is needed before the moment of abnormal change to realize the complete analysis process within the time neighborhood corresponding to the moment of abnormal change.

[0097] The preset time range can be specifically 1 second. Taking the abnormal change moment as the center point of the time range, the corresponding first 0.5 seconds and last 0.5 seconds form a time range, which is the temperature change time period at the abnormal change moment. The temperature change time period is used to specifically analyze the time range of the abnormal change moment in the front and back neighborhoods.

[0098] Furthermore, in some embodiments of the present invention, the change extension index of each abnormal change moment is determined according to the numerical distribution of temperature values ​​in each temperature change time period, including: calculating the range of temperature values ​​in the same temperature change time period; taking the standard deviation of all temperature values ​​in the same temperature change time period as the change trend influence coefficient; calculating the product of the range and the change trend influence coefficient, and normalizing it as the change extension index corresponding to the abnormal change moment.

[0099] Among them, the larger the range, the larger the overall fluctuation range of the temperature value in the corresponding temperature change time period, and the larger the standard deviation, the greater the fluctuation degree of the temperature value in the corresponding temperature change time period. Therefore, in the embodiment of the present invention, the product of the range and the change trend influence coefficient can be directly calculated and normalized as the change extension index corresponding to the abnormal change moment. That is, the larger the value of the change extension index, the more drastic the temperature change in the temperature change time period.

[0100] S104: Determine the heat dissipation power at the current moment according to the change extension index at the abnormal change moment before the current moment, and the state value and temperature value at the current moment.

[0101] By analyzing the change extension indicators of the abnormal change moments before the current moment, it is convenient to analyze the severity of the temperature change before the current moment, and determine the instability of the temperature change at the current moment. Through this analysis process, the heat dissipation power can be effectively adjusted to improve the planning and reliability of the heat dissipation power adjustment, and avoid unstable temperature changes that cause the server to be unable to effectively dissipate heat.

[0102] Furthermore, in some embodiments of the present invention, the heat dissipation power at the current moment is determined based on the change extension index of the abnormal change moment before the current moment, and the state value and temperature value at the current moment, including: taking the abnormal change moment that is closest in time sequence to the current moment as the current impact moment, and determining the time interval between the current impact moment and the current moment as the impact interval; calculating the ratio of the change extension index of the current impact moment to the impact interval as the degree of change impact at the current moment; performing maximum and minimum normalization processing on the state value and temperature value at the current moment, respectively, to obtain state parameters and temperature parameters; taking the difference between the state parameters and the temperature parameters as the indicator to be changed; calculating the product of the indicator to be changed and the degree of change impact, and normalizing it as the power adjustment weight; calculating the product of the power adjustment weight and the preset power at the current moment, to obtain the heat dissipation power at the current moment.

[0103] Among them, the current impact moment is the abnormal change moment closest to the current moment in time sequence. Therefore, the closer the current impact moment is to the current moment, the more unstable the temperature fluctuation at the current moment is. By determining the impact interval, the smaller the impact interval is, the more unstable the temperature change at the current moment is. Therefore, the ratio of the change extension index of the current impact moment to the impact interval is calculated as the degree of change impact at the current moment.

[0104] In the embodiment of the present invention, the state value at the current moment is normalized to obtain the state parameter. The normalization is the maximum and minimum normalization combined with the state value at all moments. The same is true for the temperature parameter. The larger the value of the state parameter and the temperature parameter, the larger the value of the corresponding state value and the temperature value as a whole. By calculating the difference between the state parameter and the temperature parameter, the indicator to be changed is obtained.

[0105] The indicator to be changed represents the intensity of the subsequent temperature change. The larger the value of the indicator to be changed, the greater the state parameter is, which means that the access traffic at the current moment is large and the processing speed is fast, which will emit more heat. If the value of the temperature parameter is small, it will still result in a very high temperature gap and a larger temperature value to be changed. Therefore, the subsequent temperature change will be more intense.

[0106] The product of the index to be changed and the degree of change impact is normalized and used as the power adjustment weight. In the embodiment of the present invention, the power adjustment weight is obtained by combining the temperature fluctuation and the current state, and the power adjustment weight can perform more accurate and timely adaptive analysis on the current moment.

[0107] Among them, the preset power at the current moment is the power value preset at the current moment, which can be specifically the maximum power of the entire condenser. That is to say, the product of the power adjustment weight and the maximum power is calculated as the adaptive heat dissipation power at the current moment. The heat dissipation power is an adaptive value obtained according to the server status and temperature at the current moment, as well as the fluctuation within the timing range at the current moment. The heat dissipation power can effectively avoid the delay effect of temperature, improve the timeliness of heat dissipation, optimize cooling efficiency, realize active heat dissipation regulation, and improve energy efficiency.

[0108] The embodiment of the present invention analyzes the temperature and state at different moments, and then analyzes the correlation between temperature change and state change through dislocation, adjusts the timing characteristics of the temperature sequence to correct the delay effect of the temperature, thereby obtaining a more accurate and effective target sequence of temperature change at different moments, and determines the abnormal change moment through the numerical change in the target sequence and the numerical change in the state sequence. The abnormal change moment characterizes the moment when the temperature and state undergo a sudden change, and analyzes the instability of the temperature change at each abnormal change moment through the analysis of the abnormal change moment in the time sequence neighborhood, and determines the change extension index, and adaptively adjusts the heat dissipation power at the current moment through the change extension index of the abnormal change moment before the current moment, as well as the state value and temperature value at the current moment. The present invention can eliminate the influence of the temperature delay effect, and determine the heat dissipation power that is more suitable for the current moment by combining the temperature change and the state change, so as to effectively reduce the temperature fluctuation and improve the cooling efficiency, thereby optimizing the heat dissipation operation strategy, realizing active heat dissipation regulation, and improving the energy efficiency ratio.

[0109] See also Figure 7 , Figure 7 A server heat dissipation control system structure diagram is provided, and the system includes:

[0110] The acquisition module 701 is used to acquire the temperature value of the server at different times, the access flow data to be processed, and the processing speed data; determine the state value of the server at each time according to the access flow data and the processing speed data at each time; sort the state values ​​according to the time to obtain a state sequence, and sort the temperature values ​​to obtain a temperature sequence;

[0111] The delay analysis module 702 is used to shift the temperature sequence forward according to the time, determine the temperature delay correction sequence with the strongest correlation according to the correlation between the shifted sequence obtained after the shifting of different sequence bits and the state sequence; adjust each moment of the temperature value in the temperature delay correction sequence to be the same as the corresponding one in the state sequence, and obtain the target sequence;

[0112] The temperature change determination module 703 is used to determine the abnormal change moment according to the value changes of the target sequence and the state sequence at adjacent moments, determine the temperature change time period based on the abnormal change moment, and determine the change extension index of each abnormal change moment according to the value distribution of the temperature value in each temperature change time period;

[0113] The heat dissipation control module 704 is used to determine the heat dissipation power at the current moment according to the change extension index at the abnormal change moment before the current moment, as well as the state value and temperature value at the current moment.

[0114] Furthermore, in the acquisition module 701, the state value of the server at each moment is determined according to the access flow data and the processing speed data at each moment, including:

[0115] The product of the access traffic data and the processing speed data is used as the status value of the server at each moment.

[0116] Further, in the delay analysis module 702, the forward shift is to shift the temperature value to the left in the time sequence according to different sequence digits, and delete the temperature value of the corresponding sequence digit before the temperature sequence to obtain a shifted sequence corresponding to each sequence digit; wherein, according to the correlation between the shifted sequence obtained after the shift of different sequence digits and the state sequence, the temperature delay correction sequence with the strongest correlation is determined, including:

[0117] Calculate the Pearson correlation coefficient between the dislocation sequence and the state sequence, and perform normalization to obtain the correlation index;

[0118] The misaligned sequence with the largest correlation index value is taken as the temperature delay correction sequence.

[0119] Furthermore, in the temperature change determination module 703, the abnormal change moment is determined according to the value changes of the target sequence and the state sequence at adjacent moments, including:

[0120] Take any data moment as the moment to be analyzed, and take a preset number of other moments that are closest to the moment to be analyzed as comparison moments;

[0121] According to the temperature difference between the time to be analyzed and all the comparison times in the target sequence, the neighborhood temperature change index of the time to be analyzed is determined;

[0122] Determine the neighborhood state change index of the time to be analyzed according to the difference in state values ​​between the time to be analyzed and all the comparison times in the state sequence;

[0123] Calculate the product of the neighborhood temperature change index and the neighborhood state change index at the time to be analyzed, and normalize them as the state temperature impact index;

[0124] The moment when the value of the state temperature impact index is greater than the preset impact threshold is regarded as the abnormal change moment.

[0125] Furthermore, the method for obtaining the neighborhood temperature change index and the neighborhood state change index at the time to be analyzed includes:

[0126] In the target sequence, the absolute value of the difference between the temperature value corresponding to the time to be analyzed and each comparison time is calculated as the temperature change value between the time to be analyzed and the corresponding comparison time;

[0127] The average of the temperature change values ​​between the time to be analyzed and all the comparison times is taken as the neighborhood temperature change index of the time to be analyzed;

[0128] In the state sequence, the absolute value of the difference between the state value corresponding to the time to be analyzed and each comparison time is calculated as the state change value between the time to be analyzed and the corresponding comparison time;

[0129] The mean of the state change values ​​between the time to be analyzed and all the comparison times is taken as the neighborhood state change index of the time to be analyzed.

[0130] Furthermore, in the temperature change determination module 703, the temperature change time period is determined based on the abnormal change moment as a reference point, including:

[0131] The preset time range with the abnormal change moment as the center point is used as the temperature change time period corresponding to the abnormal change moment.

[0132] Furthermore, in the temperature change determination module 703, the change extension index at each abnormal change moment is determined according to the numerical distribution of the temperature value in each temperature change time period, including:

[0133] Calculate the range of temperature values ​​within the same temperature change period;

[0134] The standard deviation of all temperature values ​​within the same temperature change period is taken as the change trend influence coefficient;

[0135] The product of the range and the change trend influence coefficient is calculated and normalized as the change extension indicator corresponding to the abnormal change moment.

[0136] Furthermore, in the heat dissipation control module 704, the heat dissipation power at the current moment is determined according to the change extension index at the abnormal change moment before the current moment, and the state value and temperature value at the current moment, including:

[0137] The change abnormal moment closest to the current moment in time sequence is taken as the current impact moment, and the time interval between the current impact moment and the current moment is determined as the impact interval; the ratio of the change extension index of the current impact moment to the impact interval is calculated as the change impact degree of the current moment;

[0138] Perform maximum and minimum normalization processing on the state value and temperature value at the current moment respectively to obtain the state parameter and temperature parameter;

[0139] The difference between the state parameter and the temperature parameter is used as the indicator to be changed;

[0140] Calculate the product of the indicator to be changed and the degree of change impact, and normalize it as the power adjustment weight;

[0141] The product of the power adjustment weight and the preset power at the current moment is calculated to obtain the heat dissipation power at the current moment.

[0142] The embodiment of the present invention analyzes the temperature and state at different moments, and then analyzes the correlation between temperature change and state change through dislocation, adjusts the timing characteristics of the temperature sequence to correct the delay effect of the temperature, thereby obtaining a more accurate and effective target sequence of temperature change at different moments, and determines the abnormal change moment through the numerical change in the target sequence and the numerical change in the state sequence. The abnormal change moment characterizes the moment when the temperature and state undergo a sudden change, and analyzes the instability of the temperature change at each abnormal change moment through the analysis of the abnormal change moment in the time sequence neighborhood, and determines the change extension index, and adaptively adjusts the heat dissipation power at the current moment through the change extension index of the abnormal change moment before the current moment, as well as the state value and temperature value at the current moment. The present invention can eliminate the influence of the temperature delay effect, and determine the heat dissipation power that is more suitable for the current moment by combining the temperature change and the state change, so as to effectively reduce the temperature fluctuation and improve the cooling efficiency, thereby optimizing the heat dissipation operation strategy, realizing active heat dissipation regulation, and improving the energy efficiency ratio.

[0143] It should be noted that the server heat dissipation control device provided in the above embodiment and the server heat dissipation control method provided in the above embodiment belong to the same concept, wherein the specific manner in which each module and unit performs the operation has been described in detail in the method embodiment and will not be repeated here. In actual application, the server heat dissipation control device provided in the above embodiment can allocate the above functions to different functional modules as needed, that is, divide the internal structure of the device into different functional modules to complete all or part of the functions described above, and this is not limited here.

[0144] On the other hand, the present invention also provides a computer-readable storage medium. It should be noted that the computer-readable storage medium shown in the embodiment of the present application may be a computer-readable signal medium or a physical computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable computer program. This propagated data signal may take a variety of forms, including but not limited to an electromagnetic signal, an optical signal, or any suitable combination of the above. A computer-readable signal medium may also be any computer-readable medium other than a physical computer-readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The computer program contained on the computer-readable medium may be transmitted using any suitable medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0145] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the system, method and computer-readable storage medium according to various embodiments of the present application. Among them, each box in the flowchart or block diagram can represent a module, a program segment, or a part of the code, and the above-mentioned module, program segment, or a part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0146] The units involved in the embodiments described in this application may be implemented by software or hardware, and the units described may also be set in a processor. The names of these units do not constitute limitations on the units themselves in some cases.

[0147] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0148] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A server heat dissipation control method, characterized in that: The method comprises: Obtaining the temperature value of the server at different times, the access flow data to be processed, and the processing speed data; determining the state value of the server at each time according to the access flow data and the processing speed data at each time; sorting the state values ​​according to time to obtain a state sequence, and sorting the temperature values ​​to obtain a temperature sequence; The temperature sequence is shifted forward according to the time, and the temperature delay correction sequence with the strongest correlation is determined according to the correlation between the shifted sequence obtained after the shifting of different sequence bits and the state sequence; each moment of the temperature value in the temperature delay correction sequence is adjusted to be the same as the corresponding one in the state sequence to obtain the target sequence; According to the value changes of the target sequence and the state sequence at adjacent moments, the abnormal change moment is determined, the temperature change time period is determined with the abnormal change moment as the reference point, and the change extension index of each abnormal change moment is determined according to the value distribution of the temperature value in each temperature change time period; Determine the heat dissipation power at the current moment according to the change extension index at the abnormal change moment before the current moment, as well as the state value and temperature value at the current moment; Determining the state value of the server at each moment according to the access traffic data and the processing speed data at each moment includes: The product of the access flow data and the processing speed data is used as the state value of the server at each moment; The step of determining the abnormal change moment according to the value changes of the target sequence and the state sequence at adjacent moments includes: Take any data moment as the moment to be analyzed, and take a preset number of other moments that are closest to the moment to be analyzed as comparison moments; According to the temperature difference between the time to be analyzed and all the comparison times in the target sequence, the neighborhood temperature change index of the time to be analyzed is determined; Determine the neighborhood state change index of the time to be analyzed according to the difference in state values ​​between the time to be analyzed and all the comparison times in the state sequence; Calculate the product of the neighborhood temperature change index and the neighborhood state change index at the time to be analyzed, and normalize them as the state temperature impact index; The moment when the value of the state temperature impact index is greater than a preset impact threshold is regarded as the abnormal change moment.

2. A server heat dissipation control method according to claim 1, characterized in that: The forward shifting is to shift the temperature value to the left in time sequence according to different sequence digits, and delete the temperature value of the corresponding sequence digit before the temperature sequence to obtain a shifted sequence corresponding to each sequence digit; wherein, according to the correlation between the shifted sequence obtained after the shifting of different sequence digits and the state sequence, the temperature delay correction sequence with the strongest correlation is determined, including: Calculating the Pearson correlation coefficient between the dislocation sequence and the state sequence, and performing normalization processing to obtain a correlation index; The misaligned sequence with the largest correlation index value is used as the temperature delay correction sequence.

3. A server heat dissipation control method according to claim 1, characterized in that: The method for obtaining the neighborhood temperature change index and the neighborhood state change index at the time to be analyzed includes: In the target sequence, the absolute value of the difference between the temperature value corresponding to the time to be analyzed and each comparison time is calculated as the temperature change value between the time to be analyzed and the corresponding comparison time; The average of the temperature change values ​​at the time to be analyzed and all the comparison times is used as the neighborhood temperature change index at the time to be analyzed; In the state sequence, the absolute value of the difference between the state value corresponding to the time to be analyzed and each comparison time is calculated as the state change value between the time to be analyzed and the corresponding comparison time; The average of the state change values ​​of the time to be analyzed and all the comparison times is used as the neighborhood state change index of the time to be analyzed.

4. A server heat dissipation control method according to claim 1, characterized in that: Determining the temperature change time period based on the abnormal change moment as a reference point includes: The preset time range with the abnormal change moment as the center point is used as the temperature change time period corresponding to the abnormal change moment.

5. A server heat dissipation control method according to claim 1, characterized in that: Determining the change extension index at each abnormal change moment according to the numerical distribution of the temperature value in each temperature change time period includes: Calculate the range of temperature values ​​within the same temperature change period; The standard deviation of all temperature values ​​within the same temperature change period is taken as the change trend influence coefficient; The product of the range and the change trend influence coefficient is calculated and normalized as a change extension index corresponding to the abnormal change moment.

6. A server heat dissipation control method according to claim 1, characterized in that: The step of determining the heat dissipation power at the current moment according to the change extension index at the abnormal change moment before the current moment and the state value and temperature value at the current moment includes: The change abnormal moment closest to the current moment in time sequence is taken as the current impact moment, and the time interval between the current impact moment and the current moment is determined as the impact interval; the ratio of the change extension index of the current impact moment to the impact interval is calculated as the change impact degree of the current moment; Perform maximum and minimum normalization processing on the state value and temperature value at the current moment respectively to obtain the state parameter and temperature parameter; The difference between the state parameter and the temperature parameter is used as the indicator to be changed; Calculating the product of the indicator to be changed and the degree of influence of the change, and normalizing the product to be used as the power adjustment weight; The product of the power adjustment weight and the preset power at the current moment is calculated to obtain the heat dissipation power at the current moment.

7. A server heat dissipation control system, characterized in that: The system comprises: An acquisition module is used to acquire the temperature value of the server at different times, the access flow data to be processed, and the processing speed data; determine the state value of the server at each time according to the access flow data and the processing speed data at each time; sort the state values ​​according to the time to obtain a state sequence, and sort the temperature values ​​to obtain a temperature sequence; The delay analysis module is used to shift the temperature sequence forward according to the time, determine the temperature delay correction sequence with the strongest correlation according to the correlation between the shifted sequence and the state sequence after the shifting of different sequence bits; adjust each moment of the temperature value in the temperature delay correction sequence to be the same as the corresponding one in the state sequence to obtain the target sequence; A temperature change determination module is used to determine the abnormal change moment according to the value changes of the target sequence and the state sequence at adjacent moments, determine the temperature change time period with the abnormal change moment as the reference point, and determine the change extension index of each abnormal change moment according to the value distribution of the temperature value in each temperature change time period; The heat dissipation control module is used to determine the heat dissipation power at the current moment according to the change extension index at the abnormal change moment before the current moment, as well as the state value and temperature value at the current moment; Determining the state value of the server at each moment according to the access traffic data and the processing speed data at each moment includes: The product of the access flow data and the processing speed data is used as the state value of the server at each moment; The step of determining the abnormal change moment according to the value changes of the target sequence and the state sequence at adjacent moments includes: Take any data moment as the moment to be analyzed, and take a preset number of other moments that are closest to the moment to be analyzed as comparison moments; According to the temperature difference between the time to be analyzed and all the comparison times in the target sequence, the neighborhood temperature change index of the time to be analyzed is determined; Determine the neighborhood state change index of the time to be analyzed according to the difference in state values ​​between the time to be analyzed and all the comparison times in the state sequence; Calculate the product of the neighborhood temperature change index and the neighborhood state change index at the time to be analyzed, and normalize them as the state temperature impact index; The moment when the value of the state temperature impact index is greater than a preset impact threshold is regarded as the abnormal change moment.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed, the steps of a server heat dissipation control method described in any one of claims 1 to 6 are implemented.

Images