Heat dissipation control method and device of server, electronic equipment and storage medium
By identifying the thermal control mode and load information of high-density server nodes, and independently adjusting the heat dissipation components of each node, the power consumption waste problem caused by overall temperature control in high-density servers is solved, and more efficient thermal control and energy consumption reduction is achieved.
Patent Information
- Application Number
- CN202510872421.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-26
- Publication Date
- 2025-07-29
- Estimated Expiration
- 2045-06-26
AI Technical Summary
In high-density servers, in the prior art, the heat dissipation components are uniformly controlled by only considering the maximum value of the overall temperature of the server, the heat dissipation components with lower actual heat dissipation requirements of some server nodes are still running at higher speeds or power, resulting in unnecessary power consumption and waste.
By identifying the heat dissipation control mode of the server node, combining temperature information and load information, determining and writing the heat dissipation control parameters in the single point control mode, and independently adjusting the heat dissipation components of each server node to avoid all heat dissipation components operating at the maximum temperature.
It significantly reduces the overall energy consumption of the server, solves unnecessary power consumption, improves the accuracy and efficiency of heat dissipation control, and ensures the stability and security of the system.
Smart Images

Figure CN120386441A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of heat dissipation, and in particular to a heat dissipation control method, device, electronic device and storage medium for a server. Background Art
[0002] In the related art, the design of high-density server models is different from that of general server models. High-density servers are configured with multiple server nodes. In high-density servers, generally only the overall temperature of the server is considered, and generally all heat dissipation components are uniformly controlled according to the maximum value of the temperatures of all server nodes. This results in that even if the actual heat dissipation requirements of some server nodes are low, the heat dissipation components will operate at a higher rotation speed or power, leading to an increase in unnecessary power consumption.
[0003] Therefore, there is a technical problem of power consumption waste in the related art. Summary of the Invention
[0004] The present application provides a heat dissipation control method, device, electronic device and storage medium for a server, so as to solve at least the problem of power consumption waste in the related art.
[0005] The present application provides a heat dissipation control method for a server, including:
[0006] Identifying a heat dissipation control mode of a group of server nodes;
[0007] Determining a set of heat dissipation control parameters of the group of server nodes according to the temperature information of the group of server nodes, the heat dissipation control mode and the load information of the group of server nodes;
[0008] When the heat dissipation control mode indicated by the mode register in the control component is the single-point control mode, writing the set of heat dissipation control parameters into a set of parameter setting registers in the control component, so as to control a specified heat dissipation component to perform heat dissipation control according to the heat dissipation control parameters of the parameter setting registers in the set of parameter setting registers, where the specified heat dissipation component is the heat dissipation component in the area where the server node to be cooled in the group of server nodes is located.
[0009] The present application also provides a heat dissipation control device for a server, including:
[0010] An identification module, configured to identify a heat dissipation control mode of a group of server nodes;
[0011] A determination module, configured to determine a set of heat dissipation control parameters of the group of server nodes according to the temperature information of the group of server nodes, the heat dissipation control mode and the load information of the group of server nodes;
[0012] A first control module, configured to write the set of heat dissipation control parameters into a set of parameter setting registers in the control component when the heat dissipation control mode indicated by the mode register in the control component is the single-point control mode, so as to control a specified heat dissipation component to perform heat dissipation control according to the heat dissipation control parameters of the parameter setting registers in the set of parameter setting registers, where the specified heat dissipation component is the heat dissipation component in the area where the server node to be cooled in the set of server nodes is located.
[0013] The present application also provides an electronic device, including: a memory, configured to store a computer program; a processor, configured to implement the steps of any one of the above-mentioned server heat dissipation control methods when executing the computer program.
[0014] The present application also provides a computer-readable storage medium, in which a computer program is stored, where the computer program implements the steps of any one of the above-mentioned server heat dissipation control methods when being executed by a processor.
[0015] The present application also provides a computer program product, including a computer program, where the computer program implements the steps of any one of the above-mentioned server heat dissipation control methods when being executed by a processor.
[0016] Through the present application, the heat dissipation control mode of a set of server nodes is identified; according to the temperature information of the set of server nodes, the heat dissipation control mode, and the load information of the set of server nodes, a set of heat dissipation control parameters of the set of server nodes is determined; when the heat dissipation control mode indicated by the mode register in the control component is the single-point control mode, the set of heat dissipation control parameters is written into a set of parameter setting registers in the control component, so as to control a specified heat dissipation component to perform heat dissipation control according to the heat dissipation control parameters of the parameter setting registers in the set of parameter setting registers. By independently adjusting the heat dissipation components of each server node, the situation where all heat dissipation components operate according to the maximum value of the temperature is avoided, the overall energy consumption of the server is significantly reduced, the technical problem of unnecessary power consumption waste existing in the related art is solved, and the technical effect of reducing energy consumption is achieved. Description of the Drawings
[0017] To more clearly illustrate the embodiments of the present application, the following will briefly introduce the drawings required for the embodiments. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.
[0018] Figure 1 It is a schematic diagram of an application scenario of a server heat dissipation control method provided by an embodiment of the present application.
[0019] Figure 2Schematic flowchart of an optional server heat dissipation control method provided by an embodiment of the present application.
[0020] Figure 3 Schematic diagram of an optional server heat dissipation control method provided by an embodiment of the present application.
[0021] Figure 4 Schematic diagram of an optional control component control logic register provided by an embodiment of the present application.
[0022] Figure 5 Schematic diagram of the structure of a server provided by an embodiment of the present application.
[0023] Figure 6 Block diagram of the structure of an optional server heat dissipation control device provided by an embodiment of the present application. Detailed implementation manners
[0024] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.
[0025] It should be noted that in the description of the present application, the terms "including", "comprising" or any other variants thereof are intended to cover non-exclusive inclusions, so that a process, method, article or device including a series of elements not only includes those elements but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. The terms "first", "second", etc. in the present application are used to distinguish similar objects and not to describe a specific order or sequence.
[0026] In order to enable those skilled in the art of the present technology to better understand the solution of the present application, the present application will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.
[0027] With the advent of the cloud computing and big data era, the performance requirements of servers in data centers are constantly increasing, and at the same time, they are also facing the dual challenges of heat dissipation efficiency and cost control. In the design of high-density servers, each chassis can accommodate multiple server nodes, and these server nodes share heat dissipation resources.
[0028] In related technologies, the design of high-density server models is different from that of general server models. High-density servers are configured with multiple server nodes. In high-density servers, the Baseboard Management Controller (BMC) generally only considers the overall temperature of the server. Generally, all heat dissipation components are uniformly controlled according to the maximum value of the temperatures of all server nodes. This results in the situation that even if the actual heat dissipation requirements of some server nodes are relatively low, the heat dissipation components will still operate at a relatively high rotation speed or power, which undoubtedly causes a waste of power consumption.
[0029] To solve the above problems, an embodiment of the present application provides a method for controlling the heat dissipation of a server. According to the temperature information of a group of server nodes, the heat dissipation control mode, and the load information of a group of server nodes, a group of heat dissipation control parameters for a group of server nodes is determined; when the heat dissipation control mode indicated by the mode register in the control component is the single-point control mode, the group of heat dissipation control parameters is written into a group of parameter setting registers in the control component, so as to control the specified heat dissipation component to perform heat dissipation control according to the heat dissipation control parameters of the parameter setting registers in the group of parameter setting registers. By independently adjusting the heat dissipation components of each server node, the situation where all heat dissipation components operate according to the maximum temperature is avoided, the overall energy consumption of the server is significantly reduced, the technical problem of unnecessary power consumption waste existing in related technologies is solved, and the technical effect of reducing energy consumption is achieved.
[0030] According to one aspect of the embodiment of the present application, a method for controlling the heat dissipation of a server is provided. Optionally, in this embodiment, the above method for controlling the heat dissipation of a server may but is not limited to be applied to a hardware environment including a terminal device 102 and a server 104 as shown in Figure 1 Figure. The server 104 can be connected to the terminal device 102 through a network and can be used to provide services for the terminal device 102 or a client installed on the terminal device 102 (for example, application services, etc.). A database can be set on the server 104 or independently of the server 104 for providing data storage services for the server 104.
[0031] The above network may include but is not limited to at least one of the following: a wired network, a wireless network. The above wired network may include but is not limited to at least one of the following: a wide area network, a metropolitan area network, a local area network. The above wireless network may include but is not limited to at least one of the following: Wireless Fidelity (WIFI), Bluetooth. The terminal device 102 may but is not limited to be a personal computer (PC), a mobile phone, a tablet computer, etc. The server 104 may but is not limited to be a cloud server, a server cluster, or other server types.
[0032] The heat dissipation control method of the server according to the embodiments of the present application can be executed by the server 104, or can be executed by the terminal device 102, or can be jointly executed by the server 104 and the terminal device 102. Among them, the execution of the heat dissipation control method of the server according to the embodiments of the present application by the terminal device 102 can also be executed by the client installed thereon.
[0033] Taking the execution of the heat dissipation control method of the server in this embodiment by the server 104 as an example, Figure 2 FIG. is a schematic flowchart of an optional heat dissipation control method of a server according to an embodiment of the present application. As Figure 2 shown, the process of this method can include the following steps:
[0034] Step S202, identify the heat dissipation control mode of a group of server nodes.
[0035] It should be noted that the server can be a high-density server and can include a group of server nodes. A server node can be a basic component unit in a high-density server. Each server node can be equivalent to a small independent server and has complete computing and storage functions. Each server node can run application programs independently. A server node can include a central processing unit, a graphics processing unit, a storage module, a network interface, and so on. A heat dissipation system can be deployed on each server node, and the heat dissipation system can include components such as fans, heat sinks, and liquid cooling implementation.
[0036] Identifying the heat dissipation control mode of a group of server nodes means identifying the heat dissipation control mode in which the high-density server is located. The heat dissipation control mode can include a global control mode or a node control mode. Among them, the node control mode can be a mode in which each server node can independently adjust the heat dissipation components. The global control mode can be a mode in which a group of server nodes jointly adjust the heat dissipation components.
[0037] Optionally, the server can include a BMC and a control component. The BMC can detect the mode register in the control component through serial bus (Inter-Integrated Circuit, abbreviated as I2C) communication to identify the heat dissipation control mode. Of course, in the presence of the latest configuration instructions, the heat dissipation control mode of a group of server nodes can also be identified according to the latest configuration command, or according to the temperature difference of a group of server nodes in the current server, the heat dissipation control mode of a group of server nodes can be identified.
[0038] Step S204, determine a set of heat dissipation control parameters for a group of server nodes according to the temperature information of a group of server nodes, the heat dissipation control mode, and the load information of a group of server nodes.
[0039] It should be noted that the temperature information of a group of server nodes may include the temperature information of each server node. The temperature information of each server node may be the real-time temperature data of various components on the server node (such as the central processing unit, memory, and hard disk), which can be obtained through temperature sensors deployed on the server node. It should be noted that in a server node, temperature sensors can be respectively deployed at different positions of the server node. When the number of temperature sensors deployed on the server node is greater than 1, the temperature information of the server node may be the average value on the temperature sensors, or the weighted average value of the temperature obtained according to a preset weight ratio. The size of the preset weight ratio can be determined according to the actual situation and can generally be obtained according to the empirical values measured in experiments.
[0040] The load information of a group of server nodes may be the actual workload of the server nodes at the current time point, and generally may include the processor utilization rate, memory usage rate, input / output (I / O) activities, etc.
[0041] Optionally, the heat dissipation strategies corresponding to different heat dissipation control modes are different. Specifically, after determining the heat dissipation control mode, the heat dissipation strategy corresponding to the heat dissipation control mode can be loaded.
[0042] According to the heat dissipation strategy corresponding to the heat dissipation control mode, analyze the temperature information and load information of each server node in a group of server nodes to determine a group of heat dissipation control parameters, where one heat dissipation control parameter in the group of heat dissipation control parameters corresponds to one server node one by one.
[0043] It can be seen that in this embodiment, the method of using temperature information, heat dissipation control mode, and load information to determine heat dissipation control parameters can, to a certain extent, adapt to the fluctuations of the server workload compared with the method of only using the maximum value of the temperature information to determine the heat dissipation control parameters. Whether it is a short-term high-load peak or a long-term low-load state, it can provide an optimized heat dissipation strategy, improve the heat dissipation control accuracy to a certain extent, and also improve the overall adaptability and efficiency of the system.
[0044] Step S206, when the heat dissipation control mode indicated by the mode register in the control component is the single-point control mode, write a group of heat dissipation control parameters into a group of parameter setting registers in the control component to control the specified heat dissipation component to perform heat dissipation control according to the heat dissipation control parameters of the parameter setting registers in the group of parameter setting registers, where the specified heat dissipation component is the heat dissipation component in the area where the server node to be cooled in a group of server nodes is located.
[0045] It should be noted that the control component can generally be a Complex Programmable Logic Device (CPLD for short). In a high-density server, the control component can receive instructions sent by the BMC, parse the commands sent by the BMC, and execute control strategies for heat dissipation components (such as fans).
[0046] The control component can be provided with a mode register and a set of parameter setting registers. Among them, the mode register can be a register for storing the identifier of the current heat dissipation control mode. The parameter setting registers in a set of parameter setting registers can be registers for storing heat dissipation control parameters. Each server node can have one or more such registers for controlling the heat dissipation components associated with it.
[0047] In the case where the heat dissipation components corresponding to a set of server nodes are fans, the corresponding heat dissipation control parameters are the rotational speed control parameters of the fans, which are usually adjusted in the form of Pulse Width Modulation (PWM) signals. The duty cycle of the PWM signal determines the voltage and rotational speed of the fan motor, thereby controlling the air volume and the intensity of the air flow. Among them, the PWM signal is used to set the rotational speed of the fan motor, and the range is usually from 0% to 100%, where 0% means the fan stops rotating and 100% means the fan runs at full speed.
[0048] In the case where the heat dissipation components corresponding to a set of server nodes are the heat dissipation components in a liquid cooling system, the corresponding heat dissipation control parameters can include the rotational speed of the liquid cooling pump, the heat exchanger or cooler adjusting the temperature of the coolant, etc. Among them, the rotational speed of the liquid cooling pump is used to control the flow rate of the coolant through the server node. The heat exchanger or cooler adjusting the temperature of the coolant is used to adapt to different heat dissipation requirements.
[0049] In one example, in the case where the heat dissipation component is a fan, as Figure 3 shown, in a high-density server configured with 2 server nodes, 2 BMCs, 6 fans and a midplane CPLD, in the case where the heat dissipation control mode is the single-point heat dissipation control mode, according to the temperature information and load information of the 2 server nodes, determine the heat dissipation control parameters of the 2 server nodes, and write them into a set of parameter setting registers in the control component respectively, so as to control the fans of the 2 server nodes to dissipate heat according to the set of parameter setting registers respectively.
[0050] Through the embodiments of the present application, identify the heat dissipation control mode of a group of server nodes; determine a group of heat dissipation control parameters of a group of server nodes according to the temperature information, heat dissipation control mode, and load information of the group of server nodes; when the heat dissipation control mode indicated by the mode register in the control component is the single-point control mode, write the group of heat dissipation control parameters into a group of parameter setting registers in the control component to control the specified heat dissipation component to perform heat dissipation control according to the heat dissipation control parameters of the parameter setting registers in the group of parameter setting registers. By independently adjusting the heat dissipation components of each server node, the situation where all heat dissipation components operate according to the maximum value of the temperature is avoided, the overall energy consumption of the server is significantly reduced, the technical problem of unnecessary power consumption waste existing in the related art is solved, and the technical effect of reducing energy consumption is achieved.
[0051] In an exemplary embodiment, step S202 includes: identifying the heat dissipation control mode of a group of server nodes according to the configuration identification code; or, when it is determined according to the temperature information of a group of server nodes that the difference between the temperatures of any two server nodes in the group of server nodes is less than or equal to a preset difference, determining that the heat dissipation control mode is the global control mode; when there is at least one difference greater than the preset difference among the differences between the temperatures of any two server nodes in the group of server nodes determined according to the temperature information of the group of server nodes, determining that the heat dissipation control mode is the single-point control mode.
[0052] It should be noted that the configuration identification code can be the configuration code obtained by sending a command to the BMC to control the heat dissipation control mode in which a group of server nodes are located. The configuration code can be stored in the BMC of the server. According to the current configuration identification code in the BMC of the server, determine the current heat dissipation control mode in which a group of server nodes are located. Through the configuration identification code, the heat dissipation control mode can be quickly identified to load the heat dissipation strategy corresponding to the heat dissipation control mode.
[0053] Of course, the heat dissipation control mode can also be identified by the difference between the temperatures of any two server nodes in a group of server nodes. Among them, the heat dissipation control mode can include the single-point control mode and the global control mode. The single-point control mode can be a mode in which a server node independently controls the heat dissipation strategy according to its own temperature information and load information. The adjustment of the heat dissipation parameters of each server node does not affect other server nodes, and it is suitable for scenarios with large temperature and load differences. The global control mode can be a mode in which the heat dissipation strategies (such as fan speed control) of all server nodes are uniformly controlled.
[0054] To implement the automatic mode switching method, a preset difference can be set, namely the preset temperature difference. The preset difference can be used to determine the temperature difference between server nodes to determine whether to adopt the global control mode or the single-point control mode.
[0055] Optionally, the size of the preset difference can be determined based on multiple factors to more accurately match the operating environment and specific requirements of the server. For example, the heat capacity and heat load characteristics of server nodes. Among them, the heat capacity of each server node is different. The heat capacity of each server node depends on the power consumption level of the central processing unit of the server node, the memory size, and the layout of internal components. According to the architecture and configuration of each node, there will also be differences in the heat load generated when performing the same type of task. For example, a node containing a Graphics Processing Unit (GPU) has a significantly higher heat load when running graphics-intensive applications than a node without a GPU. By determining the preset difference based on the heat capacity and heat load characteristics, the heat dissipation strategy can be controlled more precisely, avoiding all fans from running at full speed under slight heat differences, and effectively reducing energy consumption.
[0056] Specifically, the size of the preset difference can be determined according to experimental data and the heat load of server nodes in historical operation records.
[0057] Through this embodiment, by using the single-point control mode, energy waste on nodes with small temperature differences or light loads can be avoided, significantly reducing the overall energy consumption. Moreover, each server node can adjust the heat dissipation control parameters of the heat dissipation components according to its actual needs, improving the overall efficiency of the heat dissipation system.
[0058] In an exemplary embodiment, the above method further includes: when the heat dissipation control mode indicated by the mode register is the global control mode, controlling to write the maximum heat dissipation control parameter in a set of heat dissipation control parameters into the global parameter setting register in the control component, so as to control the heat dissipation component to perform heat dissipation control according to the heat dissipation control parameters in the global parameter setting register.
[0059] It should be noted that by reading the status information of the mode register in the control component, the currently set heat dissipation control mode of the control component is determined. When it is determined that the heat dissipation control mode is the global control mode, the heat dissipation strategy corresponding to the global control mode is loaded to obtain a set of heat dissipation control parameters according to the heat dissipation strategy in the global control mode. Specifically, when the heat dissipation component is a fan, the heat dissipation control parameter in a set of heat dissipation control parameters can be the PWM value, that is, the heat dissipation control parameter with the largest selected PWM value is written into the global parameter setting register in the control component.
[0060] The global parameter setting register can be used to store the unified heat dissipation control parameters applicable to all server nodes. When in the global control mode, the maximum heat dissipation control parameter is written to guide the operation of all heat dissipation components.
[0061] Optionally, when the heat dissipation component is a fan, the heat dissipation control parameter can be a PWM value. Specifically, in actual heat dissipation, the temperature information of a group of server nodes and the load information of a group of server nodes are monitored in real time to gradually determine a group of heat dissipation control parameters. Generally, it can be set that the heat dissipation control mode remains unchanged during the heat dissipation control process. Of course, a temperature difference threshold during the heat dissipation control process can also be set to determine whether to switch the temperature control mode according to the temperature difference threshold during the heat dissipation control process. Specifically, the temperature difference threshold during the heat dissipation control process is generally greater than the preset difference to simplify the complexity of heat dissipation control.
[0062] Through this embodiment, when the heat dissipation control mode indicated by the mode register is the global control mode, control writes the maximum heat dissipation control parameter in a group of heat dissipation control parameters to the global parameter setting register in the control component, so as to control the heat dissipation component to perform heat dissipation control according to the heat dissipation control parameter in the global parameter setting register, ensuring that all heat dissipation components (such as fans) can respond to the server node with the highest heat demand, thereby effectively preventing the systematic risk caused by local overheating.
[0063] In an exemplary embodiment, step S204 includes: calculating a group of heat dissipation control parameters according to the temperature information of a group of server nodes, the load information of a group of server nodes, and the heat dissipation strategy corresponding to the heat dissipation control mode, where, when the temperature information and the load information are the same, the parameter value of the heat dissipation control parameter obtained based on the heat dissipation strategy corresponding to the single-point control mode is greater than the parameter value of the heat dissipation control parameter obtained based on the heat dissipation strategy corresponding to the global control mode.
[0064] It should be noted that different heat dissipation strategies (such as heat dissipation control logic and rules) can be preset according to different heat dissipation control modes (global control mode or single-point control mode). The heat dissipation strategy under each heat dissipation control mode will calculate different heat dissipation control parameters based on the temperature information and the load information.
[0065] When the heat dissipation component is a fan or a fan group, the heat dissipation control parameter can be a PWM value to control the rotation speed of the fan or the fan group. In this embodiment, there are differences in the corresponding heat dissipation control parameters under different heat dissipation control modes.
[0066] Through this embodiment, when the temperature and load appear to be the same, the heat dissipation control parameters corresponding to the single-point control mode are higher than those of the global control mode, which can increase the heat dissipation redundancy of the system and improve the safety and stability of thermal management. Moreover, in the single-point control mode, the parameter values of the heat dissipation control parameters are set higher, which can prevent overheating caused by local heat sources early and ensure the long-term stable operation of the server under high-density deployment.
[0067] In an exemplary embodiment, the policy information of the heat dissipation policy includes a temperature adjustment table and a load adjustment table. The temperature adjustment table is used to record a set of temperature ranges and the initial parameter values corresponding to the temperature ranges in the set of temperature ranges. The load adjustment table is used to record a set of load ranges and the adjustment values corresponding to the load ranges in the set of load ranges.
[0068] According to the temperature information of a set of server nodes, the load information of a set of server nodes, and the heat dissipation policy corresponding to the heat dissipation control mode, calculate a set of heat dissipation control parameters, including: respectively searching in the temperature adjustment table for the temperature ranges that match the temperature information of different server nodes in the set of server nodes to obtain a set of matching temperature ranges and a set of initial parameter values corresponding to the set of matching temperature ranges; respectively searching in the load adjustment table for the load ranges that match the load information of different server nodes in the set of server nodes to obtain a set of matching load ranges and a set of adjustment values corresponding to the set of matching load ranges; and determining the sum of the initial parameter value and the adjustment value corresponding to the same server node in the set of initial parameter values and the set of adjustment values as the heat dissipation control parameter in the set of heat dissipation control parameters.
[0069] It should be noted that the heat dissipation policy can be a set of predefined rule sets for guiding the heat dissipation operation of the server heat dissipation system. The heat dissipation policy can be determined based on the temperature information and load information of a set of server nodes.
[0070] Optionally, the policy information of the heat dissipation policy includes a temperature adjustment table and a load adjustment table. The temperature adjustment table is used to record a set of temperature ranges and the initial parameter values corresponding to the temperature ranges in the set of temperature ranges. The load adjustment table is used to record a set of load ranges and the adjustment values corresponding to the load ranges in the set of load ranges. In the case where the heat dissipation component is a fan or a fan group, the temperature adjustment table can be used to record a series of temperature ranges and the initial fan rotation speed parameters (PWM values) corresponding to each temperature range. These parameters are designed to adjust the operating efficiency of the heat dissipation system according to the real-time temperature of the server node. The load adjustment table can be used to record a series of load ranges and a table of fan speed adjustment values corresponding to each load range. These adjustment values are used to further fine-tune the fan speed according to the load (computing task volume) of the server node to optimize the heat dissipation effect and energy utilization.
[0071] The temperature information of a group of server nodes includes the temperature information of each server node in the group of server nodes. Optionally, the temperature information of each server node can be determined by the temperature information collected by one or more temperature sensors. In the case where there are multiple temperature sensors in a server node, the temperature information of a server node can be determined according to the average value or weighted average value of the temperatures collected by the multiple temperature sensors. In the case where the temperature information of a server node is determined according to the weighted average value, the weight for calculating the weighted average value can be determined according to the area detected by the temperature sensor.
[0072] The load information of a group of server nodes can include the load information of each server node in the group of server nodes. Among them, the load information of each server node can reflect the real-time computing or resource occupancy situation of the server node, and can be evaluated according to a pre-deployed model, or indicators such as CPU usage rate and memory occupancy.
[0073] In specific practice, the temperature adjustment table and the load adjustment table can be determined according to the heat dissipation curves for temperature and load in the heat dissipation test. Specifically, in the heat dissipation curve for temperature information, there is a temperature and the corresponding preliminary heat dissipation control parameter at that temperature. In the heat dissipation curve for load information, there is a load and the adjustment value of the heat dissipation parameter at that load. It should be noted that in the case where the load information indicates low load, the adjustment value of the corresponding heat dissipation parameter is relatively small and can be a value less than 0.
[0074] Through the policy information of the temperature adjustment table and the load adjustment table, the most suitable heat dissipation control parameters, such as fan speed, coolant flow rate, etc., can be calculated according to the actual temperature and load conditions of the server nodes, ensuring the balance between heat dissipation efficiency and energy consumption. For example, when the temperature of server node C is in the range of 30°C - 40°C and the load is 50% - 70%, by looking up the temperature adjustment table and the load adjustment table, the parameters of the heat dissipation components can be automatically adjusted to meet the heat dissipation requirements of node C.
[0075] Through this embodiment, through the combined adjustment of temperature and load, personalized heat dissipation control can be provided for each server node, ensuring the best match between heat dissipation and performance requirements. The operating state of the heat dissipation components is dynamically adjusted according to the actual temperature and load, avoiding overheat dissipation or insufficient heat dissipation, and achieving the maximization of energy efficiency. Moreover, the use of the load adjustment value ensures that in the high-load state, the heat dissipation components can provide sufficient cooling to prevent performance degradation or failures caused by overheating, enhancing the overall stability of the system.
[0076] In an exemplary embodiment, a data processing model is deployed on a server node in a group of server nodes; the method further includes: obtaining the processing duration of the server node in the group of server nodes, where the processing duration of the server node in the group of server nodes is the response duration for the data processing model deployed on the server node in the group of server nodes to process the data to be processed; determining the processing duration of the server node in the group of server nodes as the load information of the server node in the group of server nodes, and obtaining the load information of the group of server nodes.
[0077] It should be noted that the data processing model can be a software program or an algorithm model running on the server node and can be used to process data processing tasks. For example, data analysis, machine learning model training, data retrieval, etc. The running efficiency of these models can reflect the load status of the server node.
[0078] In a high-density server, there can be multiple server nodes, and each server node can independently run data processing tasks. The models and tasks of the tasks on each server node can be the same or different.
[0079] The processing duration can be the response duration for the data processing model deployed on the server node in the group of server nodes to process the data to be processed. The processing duration can be used to reflect the current load situation of the server node, such as the current computing load and resource occupancy. Among them, the longer the processing duration, the heavier the load of the server node.
[0080] Optionally, for the processing duration of each server node, it can be determined as the load information of the same server node, and the load information of the group of server nodes can be obtained. Of course, the load information of the group of server nodes can also be comprehensively evaluated through indicators such as processing duration, CPU usage rate, and memory occupancy.
[0081] In an example, assume a high-density server with 8 server nodes internally configured, and a data processing model dedicated to image processing is deployed on each node. The server's cooling system needs to dynamically adjust the cooling strategy according to the load information of the nodes to maintain system stability. Use BMC to continuously monitor the processing duration of the data processing model on each server node. For example, the processing duration of server node A is 50ms, the processing duration of server node B is 100ms, the processing duration of server node C is 15ms, etc. The monitored processing duration can be converted into the load information of the node. Specifically, the processing duration can be converted into the percentage of CPU occupancy or resource usage through a preset conversion rule. Of course, the preset conversion rule needs to consider the hardware configuration of different server nodes.
[0082] Through this embodiment, by monitoring the processing duration of the monitoring data processing model, the load status of each server node can be accurately judged, so as to realize the reasonable allocation of heat dissipation resources and avoid the inefficiency caused by resource waste or uneven allocation.
[0083] In an exemplary embodiment, the above method further includes: monitoring the configuration information at a specified position in the mode register of the control component, where the configuration information at the specified position in the mode register is used to indicate the heat dissipation control mode of the control component; when the configuration information at the specified position in the mode register is inconsistent with the configuration information corresponding to the heat dissipation control mode, triggering the generation of a configuration control instruction, where the configuration control instruction is used to indicate to switch the configuration information at the specified position in the mode register to the configuration information corresponding to the heat dissipation control mode.
[0084] It should be noted that the control component can be a component that executes the heat dissipation control strategy, and can be a CPLD on the backplane of the server. The mode register can be a special register inside the CPLD and can be used to store the configuration information of the current heat dissipation control mode, specifically stored at the specified position in the mode register. The mode register can be a single register or a part of a series of registers.
[0085] The configuration information at the specified position in the mode register can be the bit value of one or more bits in the mode register and can represent the status of the heat dissipation control mode. For example, if the bit at the specified position is 1, it represents the single-point control mode; if the bit at the specified position is 0, it represents the global control mode.
[0086] Optionally, the configuration control instruction can be issued by a controller (such as BMC) or management software, and is an instruction used to change the configuration information at the specified position in the mode register. It is usually implemented through I2C or other communication protocols to switch the heat dissipation control mode from one to another.
[0087] In specific practice, the controller can monitor the specified bit (such as bit 0) in the mode register of the control component. When the configuration information at the specified position in the mode register is inconsistent with the configuration information corresponding to the heat dissipation control mode, trigger the generation of a configuration control instruction and send the configuration control instruction to the control component to switch the configuration information at the specified position in the mode register to the configuration information corresponding to the heat dissipation control mode.
[0088] In one example, when the control component is a CPLD and the current heat dissipation control mode is the single-point control mode, the configuration information at the specified position of the mode register can be read by reading the status of the I2C register to check whether the single-point control mode has been enabled. If not, the configuration information corresponding to the single-point control mode is written to the specified position of the mode register. The code for reading the status of the I2C register can be as follows:
[0089] / / Read the status of the I2C register
[0090] i2c_read(FAN_STATUS_REG, _val);
[0091] if (reg_val & 0x80) { / / Check whether the single-point control mode has been enabled
[0092] return IPMI_CC_PARAM_OUT_OF_RANGE;
[0093] }
[0094] / / Write to the CPLD control register
[0095] i2c_write(CPLD_CTRL_REG, (mode << 7) | (pwm & 0x7F));
[0096] / / Return the current status
[0097] *((uint8_t*)response) = reg_val;
[0098] return IPMI_CC_OK;
[0099] Through this embodiment, seamless switching of the heat dissipation control mode can be easily achieved during the operation of the server without restarting the server or interrupting the service, improving the flexibility and availability of the system. Moreover, by monitoring the mode register in real time, it is ensured that the heat dissipation control mode always meets the current management requirements, avoiding problems such as insufficient heat dissipation or excessive heat dissipation caused by mode configuration conflicts or errors, and enhancing the overall reliability and security of the server system.
[0100] In an exemplary embodiment, the number of specified heat dissipation components is at least two; the method further includes: monitoring the component status of at least two specified heat dissipation components; and when there is a component status indicating that the corresponding specified heat dissipation component is in a fault state among the component statuses of the at least two specified heat dissipation components, increasing the heat dissipation control parameters of the other specified heat dissipation components except the specified heat dissipation component in the fault state.
[0101] It should be noted that the designated heat dissipation component is the heat dissipation component selected to perform the heat dissipation operation. The component status can be used to indicate the current operating condition of the heat dissipation component, including states such as normal operation, failure, maintenance, or performance degradation.
[0102] When the heat dissipation component cannot operate normally or its performance drops significantly, it is marked as a failure state. This can be caused by various reasons such as hardware failures, power problems, abnormal control circuits, etc. In the case where the heat dissipation component is a fan, the failures that can occur include fan redundancy loss and abnormal fan rotors.
[0103] The heat dissipation control parameters can be specific instructions for guiding how the designated heat dissipation component operates, such as the PWM control signal of the fan, the adjustment of the number of heat sink fins of the radiator, the coolant flow rate of the liquid cooling system, etc. These parameters directly affect the heat dissipation effect and system energy consumption.
[0104] Optionally, in the case where the heat dissipation component is a fan, the controller can continuously monitor the operating state of the fan, such as rotation speed, vibration, noise level, and power status. Based on the operating state of the fan, the component status of the fan is determined.
[0105] Optionally, the number of designated heat dissipation components is at least two. In the case where there is a component status indicating that the corresponding designated heat dissipation component is in a failure state among the component statuses of at least two designated heat dissipation components, the heat dissipation control parameters of the other designated heat dissipation components except the designated heat dissipation component in the failure state can be increased. For example, currently there are fan A and fan B. When it is detected that fan A is in a failure state, the heat dissipation control parameter of fan B can be adjusted (for example, raising the rotation speed of fan B to full speed) to compensate for the reduced heat dissipation capacity due to the failure of fan A. This ensures that the overall heat dissipation capacity of the server will not decrease significantly due to the failure of a single fan, maintaining the thermal balance and performance stability of the system.
[0106] Through this embodiment, by monitoring and responding to the failure state of the designated heat dissipation component, it is possible to still maintain efficient heat dissipation in the case of partial heat dissipation component failures, enhancing the robustness and stability of the server.
[0107] The heat dissipation control method of the server in the embodiments of the present application will be explained below in combination with optional examples. In this optional example, the server includes a temperature sensor, BMC, control components, and heat dissipation components. There is Intelligent Platform Management Interface (IPMI) firmware in the BMC. IPMI can be used to monitor and manage the hardware status of the server and related devices, including temperature, power status, fan rotation speed, etc.
[0108] When the heat dissipation control parameters and heat dissipation control mode are known, a heat dissipation control command is sent to the IPMI firmware in the BMC. Among them, the heat dissipation control command may include at least one of the following parameters: heat dissipation control parameter, heat dissipation control mode. When the heat dissipation control mode is included in the heat dissipation control command, the heat dissipation control mode indicated in the control component can be obtained for consistency detection. When only the heat dissipation control parameter is included in the heat dissipation control command, the heat dissipation control parameter can be sent to the control component to control the heat dissipation component to perform a heat dissipation operation according to the heat dissipation control parameter.
[0109] When the heat dissipation control command includes a heat dissipation control parameter and a heat dissipation control mode, the command parameter format corresponding to the heat dissipation control command may include two parts: the first part is used to indicate the heat dissipation control mode, and the second part is used to indicate the heat dissipation control parameter. For example, when the heat dissipation component is a fan, the heat dissipation control parameter is the PWM value. Among them, the heat dissipation control mode can be one byte. For example, 0x00 represents the global control mode, and 0x01 represents the single-point control mode. The PWM value occupies two bytes, and the settable range is 0x0000 to 0xFFFF, corresponding to a fan speed of 0% to 100%.
[0110] In order to enable the BMC to parse the heat dissipation control command, a new command definition can be added to the IPMI firmware, which can include a command code, a command parameter format, and a processing function. The processing function is used to judge and perform different operations according to the different parameters and command codes passed in when receiving the IPMI command. Among them, the command code can be a unique Original Equipment Manufacturer (OEM) command code, for example, 0x3C0x2C. The command parameter format can include the parameters of the first part and the parameters of the second part. The parameters of the first part can be the parameter values corresponding to the heat dissipation control mode, and the parameters of the second part can be the numerical values corresponding to the heat dissipation control parameters. For example, the PWM value (range 0x0000 to 0xFFFF, corresponding to a speed of 0% to 100%).
[0111] In specific practice, when the server includes a BMC, a CPLD, and a fan assembly, the BMC can use I2C communication to write the heat dissipation control parameter into the corresponding parameter setting register in the CPLD, and the CPLD reads these parameters and applies them to the heat dissipation components in the corresponding area. Specifically, as Figure 4 shown Figure 4A schematic diagram of an optional control component control logic register provided by an embodiment of the present application. The control component control logic register may include address information, function information, and bit definitions. For example, address information: 0x20; function information: mode register; bit definition: Bit7 (mode selection bit). Another example, address information: 0x21; function information: parameter setting register; bit definition: 0x00 - 0x64 (PWM value range, corresponding to the fan speed from 0% to 100%).
[0112] Among them, when Bit7 is 0, it represents that a group of server nodes are in the global control mode, and the rotation speed control of all fans is based on the temperature and load information of multiple nodes, and the maximum PWM value is selected for unified control. When Bit7 is 1, a group of server nodes are in the single-point control mode, allowing the BMC to independently control the rotation speed of the fans of each server node without considering the PWM values of other server nodes.
[0113] Optionally, the BMC communicates with the CPLD via the I2C bus. Then an example of the communication command between the BMC and the CPLD can be: i2cset -y7 0x50 0x20 0x01, and this communication command is used to set the single-point control mode. First, communicate with the CPLD device via the I2C bus (such as bus 7), and the address of the CPLD can be 0x50. Then, set the value of the specified position (such as Bit7) in the mode register to 1, indicating switching to the single-point control mode. Of course, via the I2C bus communication, another communication command can be set, and an example of this communication command can be: i2cset -y 7 0x50 0x21 0x32, and this communication command can be used to set the fan PWM value. Similarly, communicate with the CPLD device (address 0x50) via the I2C bus 7, and set the PWM value to 0x32, which corresponds to 50% of the fan speed, to achieve precise control of the fan rotation speed.
[0114] In specific practice, the CPLD will compare the PWM signals of all server nodes and select the maximum value to control all fans to ensure that the heat dissipation requirements of any server node can be met. In the single-point control mode, the CPLD will directly use the PWM signal provided by the BMC and only adjust the fans in the area where the selected node is located in this mode, improving the flexibility and energy efficiency of heat dissipation control.
[0115] The heat dissipation control method of the server in the embodiment of the present application will be explained below in combination with another optional example. In the optional example, as Figure 5 shown Figure 5 A schematic diagram of the structure of a server provided by an embodiment of the present application. Among them, the server may include a controller 502, a control component 504, a fan component 506, and a temperature sensor 508.
[0116] A controller 502, which can be a BMC, is used to load a heat dissipation strategy and monitor the status of a group of server nodes, including temperature information and load information, when the server is powered on; and determine heat dissipation control parameters according to the temperature information, load information, and the heat dissipation strategy corresponding to the heat dissipation control mode, and send the heat dissipation control parameters (such as PWM values) to the control component.
[0117] Specifically, temperature information is obtained through a temperature sensor 508. The temperature sensors can be distributed near key heat sources inside the server nodes and can be used to monitor the temperature and feedback it to the BMC.
[0118] A control component 504, which can be a CPLD, serves as a control unit on the midplane and is responsible for receiving control instructions from the BMC and processing these instructions to control the fans according to the current control mode (global control mode or single-point control mode).
[0119] A fan assembly 506, which can be a physical fan directly driven by the CPLD, is used for heat dissipation inside the server.
[0120] Specifically, after the server is powered on, the BMC will send corresponding control signals to the CPLD according to the selected heat dissipation control mode (global or single node). In the single-point control mode, the BMC will directly send the PWM value instruction of the fan to the CPLD, and the CPLD will only adjust the fan speed on one side of the selected node; while in the global mode, the CPLD will compare the PWM values of all nodes and select the maximum value to control all fans. This process reflects the collaborative work between the BMC and the CPLD to achieve efficient heat dissipation and energy consumption optimization inside the server.
[0121] Through the description of the above embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be implemented by means of software plus a necessary general hardware platform. Of course, it can also be implemented by hardware, but in many cases, the former is a better implementation method.
[0122] An embodiment of the present application also provides a heat dissipation control device for a server, as Figure 6 shown. The device includes:
[0123] An identification module 602, which is used to identify the heat dissipation control mode of a group of server nodes;
[0124] A determination module 604, which is used to determine a set of heat dissipation control parameters for a group of server nodes according to the temperature information of a group of server nodes, the heat dissipation control mode, and the load information of a group of server nodes;
[0125] The first control module 606 is configured to write a set of heat dissipation control parameters into a set of parameter setting registers in the control component when the heat dissipation control mode indicated by the mode register in the control component is the single-point control mode, so as to control the specified heat dissipation component to perform heat dissipation control according to the heat dissipation control parameters of the parameter setting registers in the set of parameter setting registers, where the specified heat dissipation component is the heat dissipation component of the area where the server node to be cooled in a set of server nodes is located. In an exemplary embodiment,
[0126] In an exemplary embodiment, the identification module 602 includes: an identification unit, configured to identify the heat dissipation control mode of a set of server nodes according to the configuration identification code; or, when it is determined according to the temperature information of a set of server nodes that the difference between the temperatures of any two server nodes in the set of server nodes is less than or equal to a preset difference, determine that the heat dissipation control mode is the global control mode; when there is at least one difference greater than the preset difference among the differences between the temperatures of any two server nodes in the set of server nodes determined according to the temperature information of the set of server nodes, determine that the heat dissipation control mode is the single-point control mode.
[0127] In an exemplary embodiment, the heat dissipation control device of the server further includes: a second control module, configured to control writing the maximum heat dissipation control parameter in a set of heat dissipation control parameters into the global parameter setting register in the control component when the heat dissipation control mode indicated by the mode register is the global control mode, so as to control the heat dissipation component to perform heat dissipation control according to the heat dissipation control parameter in the global parameter setting register.
[0128] In an exemplary embodiment, the first control module 606 includes: a calculation unit, configured to calculate a set of heat dissipation control parameters according to the temperature information of a set of server nodes, the load information of the set of server nodes, and the heat dissipation strategy corresponding to the heat dissipation control mode, where when the temperature information and the load information are the same, the parameter value of the heat dissipation control parameter obtained based on the heat dissipation strategy corresponding to the single-point control mode is greater than the parameter value of the heat dissipation control parameter obtained based on the heat dissipation strategy corresponding to the global control mode.
[0129] In an exemplary embodiment, the strategy information of the heat dissipation strategy includes a temperature adjustment table and a load adjustment table. The temperature adjustment table is used to record a set of temperature ranges and the initial parameter values corresponding to the temperature ranges in the set of temperature ranges, and the load adjustment table is used to record a set of load ranges and the adjustment values corresponding to the load ranges in the set of load ranges;
[0130] The computing unit includes: a temperature subunit, which respectively looks up temperature ranges in a temperature adjustment table that match the temperature information of different server nodes in a group of server nodes, to obtain a group of matching temperature ranges and a group of initial parameter values corresponding to the group of matching temperature ranges; a load subunit, which is used to respectively look up load ranges in a load adjustment table that match the load information of different server nodes in a group of server nodes, to obtain a group of matching load ranges and a group of adjustment values corresponding to the group of matching load ranges; and a determination subunit, which is used to determine the sum of the initial parameter value and the adjustment value corresponding to the same server node in the group of initial parameter values and the group of adjustment values as the heat dissipation control parameter in a group of heat dissipation control parameters.
[0131] In an exemplary embodiment, a data processing model is deployed on a server node in a group of server nodes. The heat dissipation control device of the server further includes: a duration acquisition module, which is used to acquire the processing duration of a server node in a group of server nodes, where the processing duration of a server node in a group of server nodes is the response duration for the data processing model deployed on the server node in the group of server nodes to process the data to be processed; and a load determination module, which is used to determine the processing duration of a server node in a group of server nodes as the load information of the server node in the group of server nodes, to obtain the load information of the group of server nodes.
[0132] In an exemplary embodiment, the heat dissipation control device of the server further includes: a monitoring module, which is used to: monitor the configuration information at a specified position in the mode register in the control component, where the configuration information at the specified position in the mode register is used to indicate the heat dissipation control mode of the control component;
[0133] A switching module, which is used to trigger the generation of a configuration control instruction when the configuration information at the specified position in the mode register is inconsistent with the configuration information corresponding to the heat dissipation control mode, where the configuration control instruction is used to indicate switching the configuration information at the specified position in the mode register to the configuration information corresponding to the heat dissipation control mode.
[0134] For the description of the features in the embodiments corresponding to the heat dissipation control device of the server, reference can be made to the relevant descriptions in the embodiments corresponding to the heat dissipation control method of the server, which will not be elaborated here one by one.
[0135] An embodiment of the present application further provides an electronic device, including a memory and a processor. A computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any of the above embodiments of the heat dissipation control method of the server.
[0136] An embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored, and the computer program is configured to execute the steps in any of the above-described embodiments of the heat dissipation control method for a server when running.
[0137] In an exemplary embodiment, the above computer-readable storage medium may include, but is not limited to: various media such as USB flash drives, read-only memories (ROMs for short), random access memories (RAMs for short), mobile hard disks, magnetic disks, or optical discs that can store computer programs.
[0138] An embodiment of the present application further provides a computer program product. The above computer program product includes a computer program, and when the computer program is executed by a processor, it implements the steps in any of the above-described embodiments of the heat dissipation control method for a server.
[0139] An embodiment of the present application further provides another computer program product, including a non-volatile computer-readable storage medium. The non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it implements the steps in any of the above-described embodiments of the heat dissipation control method for a server.
[0140] Those skilled in the art can further realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been generally described according to functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Skilled professionals can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.
[0141] The above has introduced in detail a heat dissipation control method, device, electronic device, and storage medium for a server provided by the present application. Specific examples are used in this article to elaborate on the principle and implementation manner of the present application. The description of the above embodiments is only used to help understand the method and its core idea of the present application. It should be noted that for those of ordinary skill in the art in the technical field, without departing from the principle of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the protection scope of the claims of the present application.
Claims
1. A heat dissipation control method for a server, characterized in that, Including: Identifying a heat dissipation control mode of a group of server nodes; Determining a group of heat dissipation control parameters of the group of server nodes according to the temperature information of the group of server nodes, the heat dissipation control mode, and the load information of the group of server nodes; When the heat dissipation control mode indicated by the mode register in the control component is the single-point control mode, writing the group of heat dissipation control parameters into a group of parameter setting registers in the control component to control a specified heat dissipation component to perform heat dissipation control according to the heat dissipation control parameters of the parameter setting registers in the group of parameter setting registers, where the specified heat dissipation component is the heat dissipation component in the area where the server node to be cooled in the group of server nodes is located.
2. The method according to claim 1, characterized in that, The identifying a heat dissipation control mode of a group of server nodes includes: Identifying the heat dissipation control mode of the group of server nodes according to a configuration identification code; or When it is determined according to the temperature information of the group of server nodes that the difference between the temperatures of any two server nodes in the group of server nodes is less than or equal to a preset difference, determining that the heat dissipation control mode is the global control mode; when there is at least one difference greater than the preset difference among the differences between the temperatures of any two server nodes in the group of server nodes determined according to the temperature information of the group of server nodes, determining that the heat dissipation control mode is the single-point control mode.
3. The method according to claim 2, wherein The method further includes: When the heat dissipation control mode indicated by the mode register is the global control mode, controlling to write the maximum heat dissipation control parameter in the group of heat dissipation control parameters into the global parameter setting register in the control component to control the heat dissipation component to perform heat dissipation control according to the heat dissipation control parameter in the global parameter setting register.
4. The method according to claim 1, wherein The determining a group of heat dissipation control parameters of the group of server nodes according to the temperature information of the group of server nodes, the heat dissipation control mode, and the load information of the group of server nodes includes: Calculating the group of heat dissipation control parameters according to the temperature information of the group of server nodes, the load information of the group of server nodes, and the heat dissipation strategy corresponding to the heat dissipation control mode, where when the temperature information and the load information are the same, the parameter value of the heat dissipation control parameter obtained based on the heat dissipation strategy corresponding to the single-point control mode is greater than the parameter value of the heat dissipation control parameter obtained based on the heat dissipation strategy corresponding to the global control mode.
5. The method according to claim 4, wherein The strategy information of the heat dissipation strategy includes a temperature adjustment table and a load adjustment table. The temperature adjustment table is used to record a group of temperature ranges and the initial parameter values corresponding to the temperature ranges in the group of temperature ranges, and the load adjustment table is used to record a group of load ranges and the adjustment values corresponding to the load ranges in the group of load ranges; The calculating the group of heat dissipation control parameters according to the temperature information of the group of server nodes, the load information of the group of server nodes, and the heat dissipation strategy corresponding to the heat dissipation control mode includes: Search for the temperature ranges in the temperature adjustment table that match the temperature information of different server nodes in the set of server nodes, to obtain a set of matching temperature ranges and a set of initial parameter values corresponding to the set of matching temperature ranges; Search for the load ranges in the load adjustment table that match the load information of different server nodes in the set of server nodes, to obtain a set of matching load ranges and a set of adjustment values corresponding to the set of matching load ranges; Determine the sum of the initial parameter value and the adjustment value corresponding to the same server node in the set of initial parameter values and the set of adjustment values as the heat dissipation control parameter in the set of heat dissipation control parameters.
6. The method according to claim 1, wherein A data processing model is deployed on the server nodes in the set of server nodes; the method further includes: Obtain the processing duration of the server nodes in the set of server nodes, where the processing duration of the server nodes in the set of server nodes is the response duration for the data processing model deployed on the server nodes in the set of server nodes to process the data to be processed; Determine the processing duration of the server nodes in the set of server nodes as the load information of the server nodes in the set of server nodes, to obtain the load information of the set of server nodes.
7. The method according to any one of claims 1 to 6, characterized in that, The method further includes: Monitor the configuration information at a specified position in the mode register in the control component, where the configuration information at the specified position in the mode register is used to indicate the heat dissipation control mode of the control component; When the configuration information at the specified position in the mode register is inconsistent with the configuration information corresponding to the heat dissipation control mode, trigger the generation of a configuration control instruction, where the configuration control instruction is used to indicate switching the configuration information at the specified position in the mode register to the configuration information corresponding to the heat dissipation control mode.
8. A heat dissipation control device for a server, characterized in that, Includes: An identification module, configured to identify the heat dissipation control mode of a set of server nodes; A determination module, configured to determine a set of heat dissipation control parameters for the set of server nodes according to the temperature information of the set of server nodes, the heat dissipation control mode, and the load information of the set of server nodes; A first control module, configured to write the set of heat dissipation control parameters into a set of parameter setting registers in the control component when the heat dissipation control mode indicated by the mode register in the control component is the single-point control mode, so as to control a specified heat dissipation component to perform heat dissipation control according to the heat dissipation control parameters of the parameter setting registers in the set of parameter setting registers, where the specified heat dissipation component is the heat dissipation component in the area where the server node to be cooled in the set of server nodes is located.
9. An electronic device, characterized in that, Includes: A memory, configured to store a computer program; A processor, configured to implement the steps of the heat dissipation control method of the server according to any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, characterized in that, A computer program is stored in the computer-readable storage medium, where the computer program implements the steps of the heat dissipation control method of the server according to any one of claims 1 to 7 when executed by a processor.
Citation Information
Patent Citations
Thermal dissipation management system for server cluster
CN102902328A
Multi-node server PSU access and energy consumption heat dissipation management structure and method
CN114461041A
Heat dissipation control method and device of server, storage medium and electronic device
CN116027868A
Server heat dissipation control method, electronic equipment and storage medium
CN120066922A
System, method, and architecture for dynamic server power management and dynamic workload management for multi-server environment
US20030196126A1