Fan speed acquisition method, server heat dissipation method, device and equipment

Abstract

The present application discloses a fan speed acquisition method, a server heat dissipation method, an apparatus, and a device, and relates to the field of server technology. In the method of the present application, on the one hand, when the operating system is in normal operation, since the operating system has relatively abundant computing resources, the operating system can predict the temperature of each location of the server in the next time period by running a model, and control the fan speed of the server based on the predicted temperature. In this way, the fan speed can be ensured to match the temperature of each location, thereby improving the heat dissipation capacity of the server and solving the problem of poor server heat dissipation effect in some technologies. On the other hand, when the operating system is abnormal, the baseboard management control system determines the fan speed based on simple control logic, which can reduce the temperature rise rate of the server, thereby minimizing the impact of high temperature on the server.

Description

Technical Field

[0001] The present application relates to the field of server technology, and in particular to a method for obtaining fan speed, a server heat dissipation method, device, and equipment. Background Art

[0002] Good heat dissipation in a server ensures full performance. However, poor heat dissipation can, on the one hand, reduce the performance of various components within the server. For example, components like chips may experience frequency throttling due to excessive temperatures, resulting in a decrease in computing speed. Furthermore, electronic components within the server may malfunction due to excessive temperatures, leading to server freezes, lags, or even hardware damage, hindering the timely processing of various online services that rely on the server.

[0003] Currently, some cooling technologies typically monitor server temperatures regularly and then control the speed of internal fans based on the server temperature. For example, when the server temperature is high, the fan can be controlled to run at a higher speed to dissipate heat in a timely manner. When the server temperature is low, the fan can be controlled to run at a lower speed to reduce power consumption. However, these technologies suffer from hysteresis in fan speed control, resulting in poor cooling performance. Summary of the Invention

[0004] The present application provides a fan speed acquisition method, a server heat dissipation method, a fan speed acquisition device, a server heat dissipation device, an electronic device and a computer-readable storage medium, so as to at least solve the problem of poor server heat dissipation effect in the related art.

[0005] This application provides a method for obtaining fan speed, including:

[0006] receiving current operation information of the server sent by the heat dissipation controller, the current operation information including fan speed, server power, and temperature at at least one location of the server at multiple time points in a current period;

[0007] Determining, based on the current operating information, a first correlation between the temperature at each location of the server and the fan speed, and a second correlation between the temperature at each location of the server and the server power;

[0008] The current operation information, the first correlation and the second correlation are input into a temperature prediction model to obtain target temperatures of various positions of the server in the next period, and based on the target temperatures, a fan speed in the server is obtained.

[0009] The present application also provides a server heat dissipation method, comprising:

[0010] Acquire current operating information of the server, the current operating information including fan speed, server power, and temperature at at least one location of the server at multiple time points in a current period;

[0011] If the operating system in the server is in a normal operating state, the current operating information is sent to the operating system, a first fan speed is obtained from the operating system, and a fan in the server is controlled according to the first fan speed to dissipate heat for the server, where the first fan speed is obtained by the operating system based on the above-mentioned fan speed obtaining method.

[0012] If the operating system is not in a normal operating state and the baseboard management control system in the server is in a normal operating state, the temperature in the current operating information is sent to the baseboard management control system, the second fan speed returned by the baseboard management control system is obtained, and the fan in the server is controlled according to the second fan speed to dissipate heat for the server.

[0013] The present application also provides a fan speed acquisition device, comprising:

[0014] an information receiving module, configured to receive current operating information of the server sent by the heat dissipation controller, the current operating information including fan speed, server power, and temperature at at least one location of the server at multiple time points within a current period;

[0015] a correlation determination module, configured to determine, based on the current operating information, a first correlation between the temperature at each location of the server and the fan speed, and to determine a second correlation between the temperature at each location of the server and the server power;

[0016] The speed determination module is used to input the current operating information, the first correlation and the second correlation into a temperature prediction model to obtain the target temperature of each position of the server in the next time period, and obtain the fan speed in the server based on the target temperature.

[0017] The present application also provides a server heat dissipation device, comprising:

[0018] an information acquisition module, configured to acquire current operating information of the server, the current operating information including fan speed, server power, and temperature at at least one location of the server at multiple time points within a current period;

[0019] a first fan control module configured to, if the operating system in the server is in a normal operating state, send the current operating information to the operating system, obtain a first fan speed returned by the operating system, and control the fan in the server to dissipate heat from the server according to the first fan speed, where the first fan speed is obtained by the operating system based on the above-mentioned fan speed acquisition method;

[0020] The second fan control module is configured to send the temperature in the current operation information to the baseboard management control system if the operating system is not in a normal operating state and the baseboard management control system in the server is in a normal operating state, obtain a second fan speed returned by the baseboard management control system, and control the fan in the server according to the second fan speed to dissipate heat from the server.

[0021] The present application also provides an electronic device, comprising: a memory for storing a computer program; and a processor for implementing the steps of the above-mentioned fan speed acquisition method or server heat dissipation method when executing the computer program.

[0022] The present application also provides a computer-readable storage medium, in which a computer program is stored. When the computer program is executed by a processor, the steps of the above-mentioned fan speed acquisition method or server heat dissipation method are implemented.

[0023] In the technical solutions of some embodiments of the present application, on the one hand, when the operating system is in normal operation, due to its relatively abundant computing resources, the operating system can predict the temperature of each location of the server in the next time period by running a model and control the server's fan speed based on the predicted temperature. This ensures that the fan speed matches the temperature at each location, thereby improving the server's heat dissipation capacity. On the other hand, when the operating system is abnormal, the baseboard management control system determines the fan speed based on simple control logic, which can reduce the temperature rise rate of the server, thereby minimizing the impact of high temperature on the server. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0025] Figure 1 A schematic diagram of a server cooling system module in some technologies;

[0026] Figure 2A flowchart of a method for obtaining fan speed provided in some embodiments of the present application;

[0027] Figure 3 A schematic diagram of a module of a temperature prediction model provided in some embodiments of the present application;

[0028] Figure 4 A schematic flow chart of a server heat dissipation method provided in some embodiments of the present application;

[0029] Figure 5 for Figure 4 Schematic diagram of module interaction of the server cooling method in FIG.

[0030] Figure 6 A schematic diagram of a module of a fan speed acquisition device provided in some embodiments of the present application;

[0031] Figure 7 A schematic diagram of a module of a server heat dissipation device provided in some embodiments of the present application;

[0032] Figure 8 A schematic diagram of a module of an electronic device provided for some embodiments of the present application. DETAILED DESCRIPTION

[0033] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0034] It should be noted that, in the description of this application, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or device comprising 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. The terms "first," "second," etc., in this application are used to distinguish similar objects and are not used to describe a specific order or precedence.

[0035] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below with reference to the accompanying drawings and specific implementation methods.

[0036] See also Figure 1 , which is a module diagram of a server cooling system in some technologies. Figure 1The server cooling system in the system includes a baseboard management and control system, a cooling controller, multiple temperature sensors, and fans. Temperature sensors can be deployed in key heat-sensitive areas of the server to sense their temperatures. For example, temperature sensors can be deployed around components such as the central processing unit (CPU), memory modules, hard drives, and power supply units, as well as at the air inlets and outlets of the chassis. Fans are used to accelerate air circulation within the server to achieve server cooling. The fan deployment locations can correspond one-to-one with the temperature sensor deployment locations, or they do not need to correspond one-to-one.

[0037] The baseboard management and control system (BMS) runs in the baseboard management controller (BMC) and is used for out-of-band management of servers. Specifically, the BMC can be physically connected to temperature sensors. Based on this physical connection, the BMS can obtain temperatures at different locations on the server from various temperature sensors at preset intervals. Because the BMC's hardware resources (such as storage) are limited, the BMS can determine the fan speed based on simple control logic after obtaining the temperatures. For example, the fan speed can be determined based on a pre-established mapping between temperature and fan speed. In this mapping, temperature and fan speed are directly proportional. This allows for higher temperatures at or near a server location, allowing for timely heat dissipation. Conversely, lower temperatures at or near a server location can reduce fan power consumption.

[0038] The cooling controller is a complex programmable logic device (CPLD). The baseboard management controller and the cooling controller are connected via I2C (Inter-Integrated Circuit) and WDT (Watchdog Timer) lines. Each fan is connected to the cooling controller via PWM (Pulse Width Modulation Signal) and TACH (Tachometer Signal) lines, respectively.

[0039] Using the I2C circuit, the baseboard management and control system can transmit the determined fan speed to the thermal controller. The thermal controller converts the fan speed into a PWM signal and transmits it to the fan via the PWM circuit. This allows the fan speed to be controlled. While the fan is rotating, a TACH signal is returned to the thermal controller via the TACH circuit. The TACH signal indicates the current fan speed. Based on the TACH signal, the thermal controller can determine whether the fan is rotating at the planned speed. Furthermore, the thermal controller can collect fan speed statistics and return the statistical results to the baseboard management and control system via the I2C circuit.

[0040] Furthermore, during the normal operation of the baseboard management and control system, a WDT signal can be sent to the heat dissipation controller via the WDT line. The WDT signal can be a periodically changing signal. If the WDT signal does not change periodically for a long time, the heat dissipation controller can determine that the baseboard management and control system or the baseboard management controller is abnormal. At this time, the heat dissipation controller can take over the control of the fan speed. Specifically, the heat dissipation controller can store a pre-set target fan speed. The target fan speed can be the maximum speed of each fan. When the baseboard management and control system or the baseboard management controller is abnormal, since the heat dissipation controller cannot obtain the temperature at various locations inside the server, it can control the speed of each fan according to the preset target fan speed. At this time, although each fan consumes more electricity, it can at least ensure that the server can dissipate heat normally.

[0041] With the rapid development and widespread application of artificial intelligence, cloud computing, big data, etc., the operating power of servers is getting higher and higher. When the operating power of the server increases, the speed at which heat is generated during operation also increases. That is, the temperature at various locations in the server may rise to very high temperatures in a short period of time. Figure 1In the illustrated solution, fan speed is controlled based on the temperature at various locations within the server. For example, after the temperature has risen to 50 degrees Celsius, the fan is controlled at the speed corresponding to 50 degrees Celsius. This fan speed control method has a lag effect, and when the temperature inside the server rises rapidly, it cannot effectively dissipate heat from the server, resulting in poor cooling. For example, at time 1, the baseboard management control system detects that the temperature inside the server is 50 degrees Celsius. After simple logical calculation, it determines that the fan speed corresponding to 50 degrees Celsius is 15 rpm and sends this fan speed to the cooling controller. The cooling controller generates a PWM signal based on the fan speed of 15 rpm and adjusts the fan speed to 15 rpm at time 2, which is later than time 1. However, due to the relatively rapid temperature rise within the server, at time 2, the temperature inside the server may have risen to 60 degrees Celsius, but the fan is still operating at the speed corresponding to 50 degrees Celsius, resulting in poor cooling.

[0042] In view of this, the present application first provides a method for obtaining fan speed, which can predict the temperature at various locations of the server within a period of time in the future. In this way, the fan speed can be controlled in advance based on the predicted temperature, so that the fan speed matches the temperature inside the server, thereby improving the heat dissipation capacity of the server. For example, take the above-mentioned time point 1 and time point 2 as an example. Assume that at time point 1, it is predicted that at time point 2, the temperature inside the server will rise to 60 degrees Celsius. Then, the target fan speed corresponding to 60 degrees Celsius (for example, 20 revolutions per second) can be determined at time point 1, and a PWM signal can be generated according to the target fan speed. In this way, at time point 2, the fan rotation can be controlled according to the target fan speed, so that the fan speed matches 60 degrees Celsius, which can greatly improve the heat dissipation capacity of the server.

[0043] Specifically, considering that excessive computing resources may be consumed during temperature prediction and the operating system has sufficient computing resources, the program code of the operating system can be updated so that the fan speed acquisition method of the present application can be executed during operation. Since the operating system mainly runs in the central processing unit of the server, the method of the present application can be applied to the operating system, or to the central processing unit running the operating system. Figure 2 , which is a flow chart of a method for obtaining fan speed provided in some embodiments of the present application.

[0044] Step S201: receiving current operation information of a server sent by a heat dissipation controller, where the current operation information includes fan speed, server power, and temperature at at least one location of the server at multiple time points in a current period.

[0045] Specifically, the current time period refers to the specified duration before the current time. For example, if the specified duration is 10 minutes, and the current time is 10:45, the current time period is from 10:35 to 10:45. If the current time is 10:46, the current time period is from 10:36 to 10:46.

[0046] In this embodiment, the heat dissipation controller can be directly connected to the temperature sensor. The heat dissipation controller can collect fan speed, server power and temperature at various locations every preset time period (for example, every 10 seconds). In this way, the fan speed, server power and temperature at various locations at multiple time points can be obtained. Based on the collected data, the heat dissipation controller can send the fan speed, server power and temperature at various locations collected at multiple time points in the current time period as current operation information to the operating system. For example, assuming that the current time period is from 10:35 to 10:45, the heat dissipation controller can send the fan speed, server power and temperature at various locations collected at the following time points as current operation information to the operating system:

[0047] The fan speeds S11, S12, ..., S1m of each fan, the server power P1, and the temperatures T11, T12, ..., T1n at various locations collected at 10:35:10;

[0048] The fan speeds S21, S22, ..., S2m of each fan, the server power P2, and the temperatures T21, T22, ..., T2n at each location collected at 10:35:20;

[0049] …

[0050] The fan speeds St1, St2, ..., Stm of each fan, the server power Pt, and the temperatures Tt1, Tt2, ..., Ttn at each location collected at 10:45:00.

[0051] Step S202 : determining a first correlation between the temperature at each location of the server and the fan speed, and determining a second correlation between the temperature at each location of the server and the server power based on the current operation information.

[0052] Specifically, the first correlation is used to characterize the degree of correlation between the temperature at each position and the fan speed of each fan. The greater the first correlation, the greater the correlation. For the temperature at any position A and the fan speed of any fan F, if there is a large correlation between the temperature at position A and the fan speed of fan F, then when the fan speed of fan F changes, the temperature at position A will also change accordingly. On the contrary, if there is no large correlation between the temperature at position A and fan F, then when the fan speed of fan F changes, the temperature at position A will not change or will change slightly. For example, assuming that the first correlation between the temperature at position A1 and the fan speed of fan F1 is 0.3, it means that there is no large correlation between the temperature at position A1 and the fan speed of fan F1, that is, when the fan speed of fan F1 changes, the temperature at position A1 will not change or will change slightly. For another example, assuming that the first correlation between the temperature at position A1 and the fan speed of fan F2 is 0.8, it means that the temperature at position A1 and the fan speed of fan F2 have a large correlation, that is, when the fan speed of fan F2 changes, the temperature at position A1 will also change accordingly.

[0053] Similarly, the second correlation is used to characterize the degree of correlation between the temperature at each location and the server power. The larger the second correlation, the greater the correlation. For the temperature at any location A, if the temperature at location A has a large correlation with the server power, it means that when the server power changes, the temperature at location A will also change accordingly. Conversely, if the temperature at location A has no large correlation with the server power, it means that when the server power changes, the temperature at location A will not change or will change slightly. For example, assuming that the second correlation between the temperature at location A1 and the server power is 0.2, it means that the temperature at location A1 has no large correlation with the server power, that is, when the server power changes, the temperature at location A1 will not change or will change slightly. For another example, assuming that the second correlation between the temperature at location A2 and the server power is 0.9, it means that the temperature at location A2 has a large correlation with the server power, that is, when the server power changes, the temperature at location A will also change accordingly.

[0054] In this embodiment, the first correlation between the temperature at each position and the fan rotation speed of each fan can be calculated based on Expression (1).

[0055]

[0056] in, represents the first correlation between the temperature at the i-th position of the server and the fan speed of the j-th fan, p represents the number of time points in the current period, Indicates the fan speed of the jth fan at the kth time point in the current period, represents the average fan speed of the jth fan in the current period, represents the temperature of the i-th location of the server at the k-th time point in the current period, Represents the average temperature of the i-th server location during the current period.

[0057] And, based on Expression (2), a second correlation between the temperature at each location and the server power can be calculated.

[0058]

[0059] Among them, about 、 、 , please refer to the description of expression (1), which will not be repeated here. represents the second correlation between the temperature at the i-th position of the server and the server power, represents the server power at the kth time point in the current period, Indicates the average server power during the current period.

[0060] Step S203 , inputting the current operation information, the first correlation and the second correlation into a temperature prediction model, obtaining the target temperature of each position of the server in the next period, and obtaining the fan speed in the server based on the target temperature.

[0061] Specifically, similar to the current time period, the next time period is the time period of a specified duration after the current time point. For example, if the specified duration is 10 minutes, if the current time point is 10:45, the next time period is between 10:45 and 10:55. If the current time point is 10:46, the next time period is between 10:46 and 10:56.

[0062] Specifically, the next time period can also be divided into multiple time points. The temperature prediction model can output the predicted temperature of each location of the server at each time point in the next time period. Furthermore, the fan speed at each time point can be determined in sequence according to the time point sequence, and the fan can be controlled according to the fan speed corresponding to each time point before each time point. For example, assuming that the next time period is between 10:46 and 10:56, the temperature prediction model can output the following:

[0063] The predicted temperature T11 at 10:46:10 at position A1, the predicted temperature T12 at 10:46:20, ..., the predicted temperature T1p at 10:56:00;

[0064] The predicted temperature T21 at 10:46:10 at position A2, the predicted temperature T22 at 10:46:20, ..., the predicted temperature T2p at 10:56:00;

[0065] ...;

[0066] The predicted temperature Tn1 at the position An at 10:46:10, the predicted temperature Tn2 at 10:46:20, ..., the predicted temperature Tnp at 10:56:00.

[0067] Furthermore, based on the predicted temperature T11 of position A1 at 10:46:10, the predicted temperature T21 of position A2 at 10:46:10, ..., and the predicted temperature Tn1 of position An at 10:46:10, it can be determined that the fan speed of fan F1 at 10:46:10 is 10 rpm, the fan speed of fan F2 at 10:46:10 is 15 rpm, ..., and the fan speed of fan Fm at 10:46:10 is 20 rpm.

[0068] Based on the predicted temperature T12 at position A1 at 10:46:20, the predicted temperature T22 at position A2 at 10:46:20, ..., the predicted temperature Tn2 at position An at 10:46:20, it can be determined that the fan speed of fan F1 at 10:46:20 is 11 rpm, the fan speed of fan F2 at 10:46:20 is 13 rpm, ..., and the fan speed of fan Fm at 10:46:20 is 22 rpm.

[0069] Furthermore, at the kth time point in the next period (e.g., 10:46:10), the fan speed can be controlled based on the fan speed determined at the k+1th time point (e.g., 10:46:20). In this way, at the k+1th time point, the fan speed matches the temperature at the k+1th time point, significantly improving the server's heat dissipation capacity.

[0070] Of course, it's understandable that the temperature prediction model can also output the average temperature at each location on the server for the next period. Thus, at the end of the current period, the fan speed can be determined based on the average temperature for the next period. Before the next period begins, the fan speed is controlled based on the determined fan speed. Thus, when the next period begins, the fan speed matches the temperature at each location.

[0071] In summary, in the technical solutions of some embodiments of the present application, based on the current operating information of the server, a first correlation between the temperature at each location on the server and the fan speed, as well as a second correlation between the temperature at each location on the server and the server power, can be determined. Furthermore, based on the current operating information, the first correlation, and the second correlation, a target temperature for each location on the server in the next time period can be predicted. This allows the fan speed in the server to be controlled in advance according to the target temperature, so that the fan speed in the server matches the temperature. This significantly improves the server's heat dissipation capacity, thereby resolving the problem of poor server heat dissipation caused by the lag in fan control in some technologies.

[0072] See also Figure 3 , which is a module diagram of a temperature prediction model provided in some embodiments of the present application. Figure 3 The temperature prediction model consists of an input layer, a long short-term memory (LSTM) network layer, a self-attention layer, a fully connected layer, and an output layer. The input layer feeds the current operating information, the first correlation, and the second correlation into the temperature prediction model. The LSTM layer captures long-term dependencies in the input data, and the self-attention layer dynamically captures the importance of different parts of the input data. This significantly improves prediction accuracy. The fully connected layer integrates the outputs of the LSTM and self-attention layers to produce predicted temperatures for each server location. This predicted temperature can then be output from the output layer.

[0073] Furthermore, it is understandable that during the operation of the server, its operating environment, operating status, etc. are all changing dynamically. Accordingly, the correlation between the temperature at various locations on the server and the fan speed and server power should also change dynamically. For example, in the early stages of server use, the correlation between the temperature at location A1 and the server power may be small, but by the middle of the server's use, the correlation between the temperature at location A1 and the server power will gradually increase. Therefore, it is necessary to fine-tune the temperature prediction model at different stages of server use so that the temperature prediction model can capture the latest changes and characteristics of the server's operating status, thereby ensuring prediction accuracy.

[0074] Specifically, in some embodiments, each time a server is started up, the server's fan speed, server power, and temperature at at least one location can be obtained at multiple time points within a first preset duration. Based on the server's fan speed, server power, and temperature at at least one location at multiple time points within the first preset duration, a temperature prediction model can be fine-tuned and trained. The fine-tuned and trained temperature prediction model is then used to predict the target temperature for each location on the server for the next time period based on the server's current operating information. For example, if the server is restarted at 10:00, the cooling controller can obtain the server's fan speed, server power, and temperature at at least one location at preset intervals (e.g., every 10 seconds) between 10:00 and 10:30, and send the obtained data to the operating system. In this way, the operating system can fine-tune the temperature prediction model based on the obtained data, so that the temperature prediction model's reasoning logic can match the server's current operating environment and operating status, thereby significantly improving the model's prediction accuracy.

[0075] Furthermore, in some embodiments, the temperature prediction model may include a first temperature prediction model and a second temperature prediction model. In the current period, the current operating information, the first correlation and the second correlation may be input into the first temperature prediction model to obtain the target temperature of each location of the server in the next period. At the same time, the second temperature prediction model may be fine-tuned and trained based on the fan speed, server power and temperature at at least one location at multiple time points before the current period. After the current period, the second temperature prediction model may be used for temperature prediction, and the first temperature prediction model may be fine-tuned and trained based on the fan speed, server power and temperature at at least one location at multiple time points within the current period. In this way, during the operation of the server, the two temperature prediction models may be fine-tuned and trained alternately based on the current operating information of the server, thereby further ensuring the prediction accuracy of the temperature prediction model.

[0076] Specifically, each time the server is started, initial fine-tuning training can be performed simultaneously on the first and second temperature prediction models based on the server's fan speed, server power, and temperature at at least one location at multiple time points within a first preset duration. After the initial fine-tuning training is completed, the first and second temperature prediction models can be alternately fine-trained, and the first and second temperature prediction models can be used alternately for temperature prediction. For example, during the first period after the initial fine-tuning training, the first temperature prediction model is used for temperature prediction, and the second temperature prediction model is fine-tuned based on the server's operating information during the first period. During the second period, the second temperature prediction model is used for temperature prediction, and the first temperature prediction model is fine-tuned based on the server's operating information during the second period. During the third period, the first temperature prediction model is used for temperature prediction, and the second temperature prediction model is fine-tuned based on the server's operating information during the third period. And so on. In this way, the inference logic of the temperature prediction model can be highly consistent with the server's operating environment, operating status, and other aspects, thereby significantly improving prediction accuracy.

[0077] In some embodiments, determining a first correlation between the temperature at each location of the server and the fan speed, and determining a second correlation between the temperature at each location of the server and the server power based on the current operating information in step S202 may include:

[0078] Normalize the fan speed, server power, and temperature at each location in the current operating information.

[0079] A first correlation is determined based on the normalized fan speed and the temperature at each location, and a second correlation is determined based on the normalized server power and the temperature at each location.

[0080] Specifically, the fan speed, server power, and temperature at each location in the current operation information may be normalized based on Expression (3).

[0081]

[0082] Take the temperature of the i-th location on the server as an example. represents the original temperature of the i-th position at the k-th time point, Indicates the lowest temperature of the i-th position in the current period, represents the maximum temperature of the i-th position in the current period, and r represents the normalized temperature of the i-th position at the k-th time point.

[0083] After normalization, all data can be converted to the same scale, reducing numerical instability problems caused by large scale differences.

[0084] See also Figure 4 and Figure 5 . Figure 4 A flowchart of a server heat dissipation method provided in some embodiments of the present application. Figure 5 for Figure 4 Schematic diagram of the module interaction of the server cooling method in [1]. Figure 4 The server cooling method in this article can be applied to Figure 5 The thermal controller in the . Figure 5 In the system, the heat dissipation controller is connected to the temperature sensor, fan, baseboard management controller and central processing unit. The connection lines between the heat dissipation controller and the temperature sensor, fan and baseboard management controller are similar to Figure 1 , which will not be described here. The baseboard management controller is used to run the baseboard management control system, and the central processing unit is used to run the operating system. The central processing unit and the heat dissipation controller are connected through the SMBUS (System Management Bus) line and the WDT line. Among them, based on the SMBUS line, data can be transmitted between the operating system and the heat dissipation controller; based on the WDT line, the heat dissipation controller can detect whether the operating system is in normal operation. The relevant principles are similar to Figure 1 The WDT circuit in is similar and will not be described here.

[0085] based on Figure 5 The structural relationship of Figure 4 The server cooling method in the embodiment may include the following steps:

[0086] Step S401: obtaining current operation information of the server, where the current operation information includes the fan speed, server power, and temperature at at least one location of the server at multiple time points in a current period.

[0087] Specifically, because the heat dissipation controller is directly connected to the temperature sensor, the temperature at various locations on the server at multiple points in time during the current period can be obtained based on the temperature sensor. Furthermore, because the heat dissipation controller is directly connected to the fan, the fan speed at multiple points in time during the current period can be obtained based on the connection line between the heat dissipation controller and the fan.

[0088] Furthermore, the baseboard management and control system, as an out-of-band management system for the server, usually stores the server power at various time points, so the heat dissipation controller can obtain the server power at multiple time points in the current period from the baseboard management and control system.

[0089] In step S402, if the operating system in the server is in normal operation, the current operation information is sent to the operating system, the first fan speed returned by the operating system is obtained, and the fan in the server is controlled according to the first fan speed to dissipate heat for the server. The first fan speed is obtained by the operating system based on the above-mentioned fan speed acquisition method.

[0090] Specifically, since the operating system has abundant computing resources, it can run a temperature prediction model to predict the temperature of each location of the server in the next period, and generate the fan speed based on the temperature in the next period. In this way, the fan speed can be guaranteed to match the temperature of each location of the server, thereby improving the heat dissipation capacity of the server. Figure 2 The relevant description is not repeated here.

[0091] Step S403: If the operating system is not in normal operation and the baseboard management control system in the server is in normal operation, the temperature in the current operation information is sent to the baseboard management control system, the second fan speed returned by the baseboard management control system is obtained, and the fan in the server is controlled according to the second fan speed to dissipate heat for the server.

[0092] Specifically, when the operating system is abnormal, after sending the temperature in the current operating information to the baseboard management control system, the baseboard management control system can determine the fan speed based on simple control logic due to the limited computing power resources of the baseboard management controller. Figure 1 As can be seen from the relevant description, although the fan speed determined by the baseboard management control system has a hysteresis, it can at least ensure the heat dissipation of the server as much as possible, thereby reducing the temperature rise rate of the server.

[0093] Based on the above description, in the technical solutions of some embodiments of the present application, on the one hand, when the operating system is in normal operation, due to the operating system's relatively abundant computing resources, the operating system can predict the temperature of each location of the server in the next time period by running a model, and control the server's fan speed based on the predicted temperature. In this way, the fan speed can be ensured to match the temperature of each location, thereby improving the server's heat dissipation capacity. On the other hand, when the operating system is abnormal, the baseboard management control system determines the fan speed based on simple control logic, which can reduce the temperature rise rate of the server, thereby minimizing the impact of high temperature on the server.

[0094] In some embodiments, if the operating system and the baseboard management control system are not in normal operation, the fan in the server is controlled according to the preset third fan speed to dissipate heat for the server. The third fan speed is irrelevant to the current operation information. Figure 1 , I will not go into details here.

[0095] In some embodiments, when the operating system in the server is in a normal operating state, it is determined whether the operating time of the server after being started has reached a first preset time. If not, the sending of the current operating information to the operating system is paused, and the temperature in the current operating information is sent to the baseboard management control system. The fan in the server is controlled to dissipate heat for the server according to the second fan speed returned by the baseboard management control system. Specifically, Figure 2 To improve the prediction accuracy of the temperature prediction model, the model needs to be fine-tuned within a predetermined period of time after the server is started. During this period, the baseboard management and control system can control the fan speed to reduce the server's temperature rise. Once the temperature prediction model is fine-tuned, the server's fan speed can be precisely controlled based on the temperature prediction model, thereby improving the server's heat dissipation capacity.

[0096] Through the description of the above implementation methods, those skilled in the art can clearly understand that the method according to the above embodiment can be implemented by means of software plus the necessary general hardware platform, and of course it can also be implemented by hardware, but in many cases the former is a better implementation method.

[0097] See also Figure 6 , is a module schematic diagram of a fan speed acquisition device provided in some embodiments of the present application. Figure 6 In the embodiment, the fan speed obtaining device includes:

[0098] An information receiving module 601 is configured to receive current operating information of the server sent by the heat dissipation controller, the current operating information including the fan speed, server power, and temperature at at least one location of the server at multiple time points in a current period;

[0099] a correlation determination module 602 for determining, based on current operating information, a first correlation between the temperature at each location of the server and the fan speed, and a second correlation between the temperature at each location of the server and the server power;

[0100] The speed determination module 603 is used to input the current operation information, the first correlation and the second correlation into the temperature prediction model to obtain the target temperature of each position of the server in the next period, and obtain the fan speed in the server based on the target temperature.

[0101] In some embodiments, the rotation speed determination module 603 is further configured to:

[0102] After the server is started each time, obtaining the fan speed, server power and temperature at at least one location of the server at multiple time points within a first preset time period;

[0103] Based on the fan speed, server power and temperature at at least one location of the server at multiple time points within a first preset time period, the temperature prediction model is fine-tuned and trained. The fine-tuned temperature prediction model is used to predict the target temperature of each location of the server in the next time period based on the current operating information of the server.

[0104] In some embodiments, the temperature prediction model includes a first temperature prediction model and a second temperature prediction model; the speed determination module 603 is further configured to:

[0105] Inputting the current operation information, the first correlation, and the second correlation into a first temperature prediction model to obtain a target temperature for each location of the server in the next time period;

[0106] Fine-tuning and training a second temperature prediction model based on fan speeds, server power, and temperature at at least one location at multiple time points before a current period;

[0107] After the current period, the second temperature prediction model is used to perform temperature prediction, and the first temperature prediction model is fine-tuned and trained based on the fan speed, server power, and temperature at at least one location at multiple time points within the current period.

[0108] In some embodiments, the relevance determination module 602 is specifically configured to:

[0109] Normalize the fan speed, server power, and temperature at each location in the current operating information.

[0110] A first correlation is determined based on the normalized fan speed and the temperature at each location, and a second correlation is determined based on the normalized server power and the temperature at each location.

[0111] See also Figure 7 , which is a module schematic diagram of a server heat dissipation device provided in some embodiments of the present application. Figure 7 In the server cooling device, the server cooling device includes:

[0112] An information acquisition module 701 is configured to acquire current operating information of the server, including fan speed, server power, and temperature at at least one location of the server at multiple time points within a current period.

[0113] A first fan control module 702 is configured to, if the operating system in the server is in a normal operating state, send current operating information to the operating system, obtain a first fan speed returned by the operating system, and control the fans in the server to dissipate heat from the server according to the first fan speed, where the first fan speed is obtained by the operating system based on the above-mentioned fan speed acquisition method;

[0114] The second fan control module 703 is used to send the temperature in the current operation information to the baseboard management control system if the operating system is not in normal operation and the baseboard management control system in the server is in normal operation, obtain the second fan speed returned by the baseboard management control system, and control the fan in the server according to the second fan speed to dissipate heat for the server.

[0115] In some embodiments, the second fan control module 703 is further configured to:

[0116] If the operating system and the baseboard management control system are not in normal operation, the fan in the server is controlled according to a preset third fan speed to dissipate heat for the server. The third fan speed is irrelevant to the current operation information.

[0117] In some embodiments, the first fan control module 702 is further configured to:

[0118] When the operating system in the server is in normal operating state, determine whether the operating time of the server after being started reaches a first preset time. If not, pause sending the current operating information to the operating system, and send the temperature in the current operating information to the baseboard management control system, and control the fan in the server to dissipate heat for the server according to the second fan speed returned by the baseboard management control system.

[0119] For the description of the features in the embodiment corresponding to the fan speed acquisition device, please refer to the relevant description of the embodiment corresponding to the fan speed acquisition method. For the description of the features in the embodiment corresponding to the server heat dissipation device, please refer to the relevant description of the embodiment corresponding to the server heat dissipation method. They will not be repeated here one by one.

[0120] See also Figure 8 An embodiment of the present application also provides an electronic device, including a memory 10 and a processor 20, wherein the memory 10 stores a computer program, and the processor 20 is configured to run the computer program to execute the steps in any of the above-mentioned fan speed acquisition methods and server heat dissipation method embodiments.

[0121] An embodiment of the present application further provides a computer-readable storage medium storing a computer program, wherein the computer program is configured to execute the steps of any of the above-mentioned fan speed acquisition method and server heat dissipation method embodiments when running.

[0122] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk, or an optical disk.

[0123] An embodiment of the present application further provides a computer program product, which includes a computer program. When the computer program is executed by a processor, the steps in any of the above-mentioned fan speed acquisition method and server heat dissipation method embodiments are implemented.

[0124] An embodiment of the present application also provides another computer program product, including a non-volatile computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the steps in any of the above-mentioned fan speed acquisition methods and server heat dissipation method embodiments.

[0125] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0126] The above is a detailed introduction to a fan speed acquisition method, server heat dissipation method, device, equipment and storage medium provided by this application. This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method of this application and its core idea. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of this application, several improvements and modifications can be made to this application, and these improvements and modifications also fall within the scope of protection of the claims of this application.

Claims

1. A method for obtaining fan speed, characterized in that: The method comprises: receiving current operation information of the server sent by the heat dissipation controller, the current operation information including fan speed, server power, and temperature at at least one location of the server at multiple time points in a current period; Determining, based on the current operating information, a first correlation between the temperature at each location of the server and the fan speed, and a second correlation between the temperature at each location of the server and the server power; Inputting the current operating information, the first correlation, and the second correlation into a temperature prediction model to obtain a target temperature for each position of the server in the next time period, and obtaining a fan speed in the server based on the target temperature; and, after the server is started each time, obtaining the fan speed, server power, and temperature at at least one location of the server at multiple time points within a first preset time period; Fine-tuning the temperature prediction model based on the fan speed, server power, and temperature at at least one location of the server at multiple time points within the first preset time period, wherein the fine-tuned temperature prediction model is used to predict a target temperature at each location of the server in a next time period based on current operating information of the server; Wherein, the temperature prediction model includes a first temperature prediction model and a second temperature prediction model; Inputting the current operation information, the first correlation, and the second correlation into the first temperature prediction model to obtain a target temperature for each location of the server in the next time period; Fine-tuning the second temperature prediction model based on the fan speed, server power, and temperature at at least one location at multiple time points before the current period; After the current period, the second temperature prediction model is used to perform temperature prediction, and the first temperature prediction model is fine-tuned and trained based on the fan speed, server power, and temperature at at least one location at multiple time points within the current period.

2. The method according to claim 1, characterized in that Determining a first correlation between the temperature at each position of the server and the fan speed, and determining a second correlation between the temperature at each position of the server and the server power based on the current operation information, includes: Normalizing the fan speed, server power, and temperature at each location in the current operation information; The first correlation is determined based on the normalized fan speed and the temperature at each location, and the second correlation is determined based on the normalized server power and the temperature at each location.

3. A server heat dissipation method, characterized in that: Applied to a heat dissipation controller, the method includes: Acquire current operating information of the server, the current operating information including fan speed, server power, and temperature at at least one location of the server at multiple time points in a current period; If the operating system in the server is in a normal operating state, sending the current operating information to the operating system, obtaining a first fan speed returned by the operating system, and controlling the fan in the server to dissipate heat for the server according to the first fan speed, where the first fan speed is obtained by the operating system based on the fan speed acquisition method according to any one of claims 1 to 2; If the operating system is not in a normal operating state and the baseboard management control system in the server is in a normal operating state, the temperature in the current operating information is sent to the baseboard management control system, the second fan speed returned by the baseboard management control system is obtained, and the fan in the server is controlled according to the second fan speed to dissipate heat for the server.

4. The method according to claim 3, characterized in that The method further comprises: If the operating system and the baseboard management control system are not in normal operation, the fan in the server is controlled according to a preset third fan speed to dissipate heat for the server, and the third fan speed is irrelevant to the current operation information.

5. The method according to claim 3, characterized in that The method further comprises: When the operating system in the server is in a normal operating state, determine whether the operating time of the server after being started reaches a first preset time; if not, pause sending the current operating information to the operating system, send the temperature in the current operating information to the baseboard management and control system, and control the fan in the server to dissipate heat for the server according to the second fan speed returned by the baseboard management and control system.

6. A fan speed acquisition device, characterized in that: The device comprises: an information receiving module, configured to receive current operation information of the server sent by the heat dissipation controller, the current operation information including fan speed, server power, and temperature at at least one location of the server at multiple time points in a current period; a correlation determination module, configured to determine, based on the current operation information, a first correlation between the temperature at each location of the server and the fan speed, and to determine a second correlation between the temperature at each location of the server and the server power; A speed determination module is used to input the current operating information, the first correlation and the second correlation into a temperature prediction model to obtain the target temperature of each position of the server in the next time period, and based on the target temperature, obtain the fan speed of the server, and obtain the fan speed, server power and temperature of at least one position of the server at multiple time points within a first preset time period each time after the server is started and run. Based on the fan speed, server power and temperature of at least one position of the server at multiple time points within the first preset time period, fine-tune the temperature prediction model. The fine-tuned temperature prediction model is used to predict the temperature based on the current operating information of the server. the target temperature of each location of the server in the next time period; wherein the temperature prediction model includes a first temperature prediction model and a second temperature prediction model; the current operating information, the first correlation and the second correlation are input into the first temperature prediction model to obtain the target temperature of each location of the server in the next time period; based on the fan speed, server power and temperature at at least one location at multiple time points before the current time period, the second temperature prediction model is fine-tuned and trained; after the current time period, the second temperature prediction model is used to perform temperature prediction, and based on the fan speed, server power and temperature at at least one location at multiple time points within the current time period, the first temperature prediction model is fine-tuned and trained.

7. A server heat dissipation device, characterized in that: The device comprises: an information acquisition module, configured to acquire current operating information of the server, the current operating information including fan speed, server power, and temperature at at least one location of the server at multiple time points within a current period; a first fan control module, configured to, if the operating system in the server is in a normal operating state, send the current operating information to the operating system, obtain a first fan speed returned by the operating system, and control the fan in the server to dissipate heat for the server according to the first fan speed, wherein the first fan speed is obtained by the operating system based on the fan speed acquisition method according to any one of claims 1 to 2; The second fan control module is configured to send the temperature in the current operation information to the baseboard management control system if the operating system is not in a normal operating state and the baseboard management control system in the server is in a normal operating state, obtain a second fan speed returned by the baseboard management control system, and control the fan in the server according to the second fan speed to dissipate heat from the server.

8. An electronic device, characterized in that: include: Memory for storing computer programs; A processor, configured to implement the steps of the fan speed acquisition method according to any one of claims 1 to 2, or the steps of the server heat dissipation method according to any one of claims 3 to 5, when executing the computer program.

Images