Server fan control method and device, storage medium and electronic equipment

Abstract

The embodiment of the invention provides a server fan control method and device, a storage medium and electronic equipment, and relates to the field of computers.The server fan control method comprises the steps that state information of a control assembly of a target server in a shutdown state is obtained, the state information of the control assembly comprises an in-place state and a non-in-place state, and the control assembly comprises N assemblies; under the condition that at least one component in the N components is in the in-place state, the duty ratio of the PWM signals corresponding to the state information of the N components is found out from the relation mapping table, and the higher the duty ratio of the PWM signals is, the higher the rotating speed of the server fan is; based on the duty ratio of the PWM signal, a server fan is controlled to operate at a preset rotating speed, the target server supports a plurality of partitions, each component in the N components is arranged in one of the partitions according to a preset arrangement rule, and heat dissipation is performed on one group of components in the target partition by operating the server fan.

Description

Technical Field

[0001] The embodiments of the present application relate to the field of computers, and in particular, to a method and device for controlling a server fan, a storage medium, and an electronic device. Background Art

[0002] With the development of technologies such as artificial intelligence, big data, and the Internet of Things, cloud computing has become a development trend in data centers. In order to meet the high performance requirements of cloud computing services, servers usually use a high-density layout of multiple CPUs (Central Processing Units) and multiple GPUs (Graphics Processing Units or accelerator cards) to provide powerful computing capabilities. However, the high-density layout of server components and a large number of computing tasks will generate high heat, which puts higher requirements on the heat dissipation capacity of the server.

[0003] In order to improve the heat dissipation performance of the server, multiple duct fans are usually installed inside the server for heat dissipation. In the current AI server, if there are multiple parts or components that have heat dissipation requirements in the S5 state, all fans are usually controlled to run at the same speed, and heat is dissipated from the heat-generating components to the external environment during the fan operation.

[0004] However, in order to support higher computing efficiency, large-scale model training, and more storage space, a large number of high-performance CPUs, GPUs, smart network cards and other supporting devices are usually configured inside the server, which will also increase the number of server fans. However, when the server is in S5 state, the various components inside the server have different heat dissipation requirements. If all fans are still controlled to run at full speed in the above-mentioned unified way, a large amount of electricity will be consumed, resulting in the technical problem of excessive energy consumption in the control process of the server fans.

[0005] In the related art, there is no effective solution to the problem of excessive energy consumption in the control process of server fans. Summary of the invention

[0006] The embodiments of the present application provide a method and device for controlling a server fan, a storage medium, and an electronic device, so as to at least solve the problem of excessive energy consumption occurring during the control process of the server fan.

[0007] According to an embodiment of the present application, a method for controlling a server fan is provided, comprising: obtaining status information of a control component of a target server in a shutdown state, wherein the status information of the control component includes an in-place state and an out-of-place state, and the control component includes N components, where N is a positive integer greater than or equal to 1; when at least one of the N components is in-place, finding a duty cycle of a PWM signal corresponding to the status information of the N components from a relationship mapping table, wherein the higher the duty cycle of the PWM signal, the higher the rotation speed of the server fan; based on the duty cycle of the PWM signal, controlling a server fan corresponding to a target partition to run at a preset rotation speed, wherein the target server supports multiple partitions, and each of the N components is respectively arranged in one of the multiple partitions according to a preset arrangement rule, and cooling a group of components in the target partition by running the server fan.

[0008] In an exemplary embodiment, the above-mentioned acquisition of status information of the control component of the target server in the shutdown state includes: sending a first PWM signal to the fan board CPLD of the target server through the switch board CPLD of the target server; parsing the first PWM signal through the fan board CPLD to obtain the status information of each of the N components.

[0009] In an exemplary embodiment, before sending the first PWM signal to the fan board CPLD of the target server through the switch board CPLD of the target server, the method further includes: when the N components include an accelerator board, a smart network card and an OCP network card, sending a smart network card in-place signal and an OCP network card in-place signal to the switch board CPLD through the mainboard of the target server, wherein both the smart network card in-place signal and the OCP network card in-place signal are valid at a low level; sending an accelerator card in-place signal to the switch board CPLD through the accelerator board, wherein the accelerator card in-place signal is valid at a low level.

[0010] In an exemplary embodiment, when at least one of the N components is in the in-place state, finding out from the relationship mapping table the duty cycle of the PWM signal corresponding to the status information of the N components includes at least one of the following: when the N components include an accelerator board, a smart network card, and an OCP network card, obtaining the first state information of the accelerator board, the second state information of the smart network card, and the third state information of the OCP network card; when the first state information indicates that the accelerator board is in the non-place state, the second state information indicates that the smart network card is in the non-place state, and the third state information indicates that the OCP network card is in the non-place state, finding out from the relationship mapping table that the duty cycle of the PWM signal is 0; when the first state information indicates that the accelerator board is in the non-place state, the second state information indicates that the smart network card is in the non-place state, and the third state information indicates that the OCP network card is in the in-place state, finding out from the relationship mapping table that the duty cycle of the PWM signal is 0. The duty cycle of the PWM signal is a first value; when the first state information indicates that the accelerator board is in a non-in-place state, the second state information indicates that the smart network card is in a position state, and the third state information indicates that the OCP network card is in a non-in-place state, the duty cycle of the PWM signal is found to be a second value from the relationship mapping table, wherein the second value is greater than the first value; when the first state information indicates that the accelerator board is in a position state, the second state information indicates that the smart network card is in a non-in-place state, and the third state information indicates that the OCP network card is in a non-in-place state, the duty cycle of the PWM signal is found to be a third value from the relationship mapping table, wherein the third value is greater than the second value; when the first state information indicates that the accelerator board is in a position state and the second state information indicates that the smart network card is in a position state, the duty cycle of the PWM signal is found to be a fourth value from the relationship mapping table, wherein the fourth value is greater than the third value.

[0011] In an exemplary embodiment, the above-mentioned control of the server fans corresponding to the target partition to run at a preset speed based on the duty cycle of the PWM signal includes: determining the position information of each group of server fans in M ​​groups of server fans through the first input signal and the second input signal on the fan board CPLD of the target server, wherein M is a positive integer greater than or equal to 2; determining the target position of at least one group of server fans for cooling the components in the target partition based on the status information of the N components, wherein the components in a partition require at least one group of server fans to cool them, and the M groups of server fans include the at least one group of server fans; based on the duty cycle of the PWM signal, controlling the at least one group of server fans at the target position to run at the preset speed.

[0012] In an exemplary embodiment, the above-mentioned determining the position information of each group of server fans in the M groups of server fans through the first input signal and the second input signal on the fan board CPLD of the target server includes: obtaining a group of server fans from the M groups of server fans in sequence as the current group of server fans; when the first input signal is a low level and the second input signal is a low level, determining that the current group of server fans is located at a first position in the fan partition; when the first input signal is a low level and the second input signal is a high level, determining that the current group of server fans is located at a second position in the fan partition; when the first input signal is a high level and the second input signal is a low level, determining that the current group of server fans is located at a third position in the fan partition; when the first input signal is a low level and the second input signal is a low level, determining that the current group of server fans is located at a fourth position in the fan partition.

[0013] In an exemplary embodiment, the method further includes: adding an in-place status signal of the N+1th component to the relationship mapping table; based on the in-place status signal of the N+1th component, updating the mapping relationship between the status information of the N+1 components and the duty cycle of the PWM signal to obtain an updated relationship mapping table; when the current status information of the N+1 components is obtained, finding the current duty cycle of the PWM signal corresponding to the current status information from the updated relationship mapping table; based on the current duty cycle of the PWM signal, controlling the current server fan corresponding to the current partition to run at the current speed, wherein the current partition includes the server partition where the components in the in-place state among the N+1 components at the current moment are located.

[0014] According to another embodiment of the embodiment of the present application, a control device for a server fan is also provided, including: a first acquisition unit, used to acquire status information of a control component of a target server in a shutdown state, wherein the status information of the control component includes an in-place state and an out-of-place state, and the control component includes N components, N being a positive integer greater than or equal to 1; a search unit, used to search for a duty cycle of a PWM signal corresponding to the status information of the N components from a relationship mapping table when at least one component among the N components is in-place, wherein the higher the duty cycle of the PWM signal, the higher the speed of the server fan; a first control unit, used to control the server fan corresponding to the target partition to run at a preset speed based on the duty cycle of the PWM signal, wherein the target server supports multiple partitions, and each of the N components is respectively arranged in one of the multiple partitions according to a preset arrangement rule, and a group of components in the target partition are cooled by running the server fan.

[0015] According to another embodiment of the present application, a computer-readable storage medium is provided, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when run.

[0016] According to another embodiment of the present application, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0017] According to another embodiment of the present application, a computer program product is provided. The computer program product includes a computer program. When the computer program is executed by a processor, the steps in any one of the above method embodiments are implemented.

[0018] Through the above-mentioned embodiment provided by the present application, after obtaining the status information of N components in the target server, at least one component that currently has a heat dissipation demand and the server fan in the target partition that has a position association relationship with the at least one component are intelligently identified according to the in-place status of the N components. Then, the duty cycle of the PWM signal corresponding to the combination mode of at least one component is found by searching the relationship mapping table, and the server fan in the target partition is controlled to operate at the speed indicated by the duty cycle of the found PWM signal, thereby realizing refined control of the server fan, achieving the technical effect that the server fan can run on demand, and reducing energy consumption. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0020] Figure 1 It is a hardware structure block diagram of a server device according to a method for controlling a server fan in an embodiment of the present application;

[0021] Figure 2 is a flow chart of an optional method for controlling a server fan according to an embodiment of the present application;

[0022] Figure 3 is an overall topology diagram of an optional method for controlling a server fan according to an embodiment of the present application;

[0023] Figure 4 is a flowchart of another optional method for controlling a server fan according to an embodiment of the present application;

[0024] Figure 5 is a specific example of an optional control strategy for a server fan according to an embodiment of the present application;

[0025] Figure 6 is a schematic diagram of an optional server heat dissipation and fan partitioning according to an embodiment of the present application;

[0026] Figure 7 is a structural block diagram of a control device for a server fan according to an embodiment of the present application;

[0027] Figure 8 It is a schematic diagram of the structure of an optional electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] The embodiments of the present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0029] It should be noted that the terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence.

[0030] The server fan control method embodiment provided in the embodiment of the present application can be executed in a server device or a similar computing device. Taking running on a server device as an example, Figure 1 FIG. 1 is a hardware structure block diagram of a server device of a method for controlling a server fan according to an embodiment of the present application. Figure 1 As shown, the server device may include one or more ( Figure 1Only one is shown in the figure) a processor 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the above-mentioned server device may also include a transmission device 106 and an input / output device 108 for communication functions. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the above server device. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0031] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the control method of the server fan in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, to implement the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the server device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0032] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the server device. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, referred to as NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0033] In order to better understand the above-mentioned server fan control method, the basic concepts of the professional terms and terminology used in the embodiments of the present application are briefly introduced below.

[0034] TDP: Thermal Design Power;

[0035] PSU: Power Supply Unit;

[0036] PWM:Pulse width modulation;

[0037] I2C: Inter-Integrated Circuit two-wire serial bus;

[0038] MUX: Multiplexer multiplexer;

[0039] PCA9546: an I2C switch chip from NXP;

[0040] NSCI: Network Controller Sideband Interface, a protocol for communication between host systems and network controllers, implemented through a sideband interface. It defines message formats, communication methods, and functions, and provides a standardized solution for the management and monitoring of network controllers;

[0041] CPLD: Complex Programmable logic device;

[0042] CLK buffer: clock buffer / expander;

[0043] CPU0_PE3: PCIe output from Port3 of CPU0;

[0044] CPU1_PE3: PCIe output from Port3 of CPU1;

[0045] MCIO: Multi-Channel Input / Output, is a multi-channel input / output connector;

[0046] VR:voltage regulator,voltage regulator or voltage stabilizer;

[0047] PWM:Pulse width modulation;

[0048] FAN board: fan board;

[0049] Efuse:Electronic fuse, electronic fuse;

[0050] OCP network card: a standard network card defined by the Open Compute Project (OCP) organization;

[0051] Server S5 state: a power state of the server. In this state, only some components of the server, such as CPLD and BMC, are powered.

[0052] In order to solve the above-mentioned problems existing in the related art, a method for controlling a server fan is provided in the present embodiment, and the execution subjects thereof include but are not limited to AI servers, high-end servers and other servers with complex architectures, many accessories, high machine heights and a large number of fans. For ease of understanding, the fan control of an AI server in S5 state is used as an example for explanation in the present application embodiment. Among them, the AI ​​server includes but is not limited to an MB control unit, a Switch board unit, a fan board and a fan adapter board unit, a GPU board unit, and an intelligent network card and an OCP network card unit, etc. Through the interaction between these five units, the in-place status of N components in the target server is obtained, and according to the in-place status information, the relationship mapping table is searched to determine the duty cycle of the PWM signal corresponding to the current in-place status information. Figure 2 As shown, the process includes the following steps S202-S206:

[0053] Step S202: Acquire status information of a control component of a target server in a shutdown state, wherein the status information of the control component includes an in-place state and an out-of-place state, and the control component includes N components, where N is a positive integer greater than or equal to 1;

[0054] Step S204: when at least one of the N components is in the on-state, searching the relationship mapping table for a duty cycle of a PWM signal corresponding to the state information of the N components, wherein the higher the duty cycle of the PWM signal, the higher the rotation speed of the server fan;

[0055] Step S206: Based on the duty cycle of the PWM signal, control the server fan corresponding to the target partition to run at a preset speed, wherein the target server supports multiple partitions, and each of the N components is arranged in one of the multiple partitions according to a preset arrangement rule, and the server fan is run to dissipate heat for a group of components in the target partition.

[0056] Before explaining the control method of the above-mentioned server fan, first Figure 3 The overall schematic diagram of the server fan control system is briefly introduced below.

[0057] like Figure 3 As shown in FIG. 1 , the control system of the server fan includes a server mainboard (MB), a switch board (Switch Board), an accelerator board (GPU Board), a fan board (FAN Board), a fan adapter board, a smart network card, and an OCP network card, etc. The following briefly introduces each of the above parts.

[0058] (1) Server motherboard (MB control unit)

[0059] The server motherboard includes but is not limited to CPLD, NCSI CONN (connector), CLK buffer, CPU, and MCIO. The NCSI connector is a key component for connecting the smart network card. Its main function is to identify whether the smart network card is in place (which can also be understood as whether the smart network card is inserted). Specifically, it determines whether the smart network card is in place by the level of the smart network card in-place signal (NIC_PRSNT_N).

[0060] CLK buffer is a clock buffer, there are two in total, the clock sources come from CPU0 and CPU respectively. Figure 3 CLK buffer (CPU0) and CLK buffer (CPU1) are shown. The clock source corresponds to the PCIE signal from the CPU. The PCIE of the Riser of the smart NIC comes from CPU1_PE3, that is, the PCIe output from Port3 of CPU1, and its clock depends on CLKbuffer (CPU1). The PCIE of the OCP NIC comes from CPU0_PE3, that is, the PCIe output from Port3 of CPU0, and its clock depends on CLKbuffer (CPU0).

[0061] The main function of the CPLD of the server motherboard (hereinafter referred to as MB CPLD) is to receive the NIC_PRSNT_N and OCP_PRSNT_N signals and transmit them to the CPLD of the switch board after processing. At the same time, the Enable output of VR0 and VR1 is controlled to power on the smart NIC and OCP NIC. That is, only when the smart NIC is determined to be in place can VR0 be controlled to power on the smart NIC; similarly, only when the OCP NIC is determined to be in place can VR1 be controlled to power on the OCP NIC.

[0062] MCIO is a high-speed connector commonly used in the industry, and its full name is Multi-Channel Input / Output. Figure 3 As shown, MCIO can be, but is not limited to, a bridge connecting the PCIe signals of the Smart NIC and the OCP NIC.

[0063] (2) Switch board unit

[0064] The switch board includes but is not limited to CPLD, P12V / P3V3 VR, and PWR CONN, wherein the switch board CPLD can, but is not limited to, receive the in-place status signals of the MB CPLD and GPU board, and perform logic processing and judgment internally, thereby outputting PWM signals with different duty cycles to the CPLD of the fan board (FAN Board).

[0065] The duty cycle of a PWM signal (Pulse Width Modulation) refers to the ratio of the time the signal is in a high level state to the entire cycle time in one cycle. The duty cycle is an important parameter of the PWM signal, which directly affects the average value and power output of the signal.

[0066] In power management, the average power of the output load can be controlled by adjusting the duty cycle. In motor drive, the duty cycle affects the speed and torque of the motor. In this embodiment, the speed of the server fan can be changed by adjusting the duty cycle.

[0067] (III) Fan plate and fan adapter plate unit

[0068] This unit contains CPLD, P54V_FAN Efuse circuit, fan (FAN), and FAN_1D0 / ID 1 signal. The CPLD of the fan board receives the PWM signal of the CPLD of the switch board. The PWM signal here is not used for fan speed control, but carries the information of the GPU board, smart network card, and OCP network card. The fan board CPLD determines the current status of the above three boards by analyzing PWM.

[0069] (IV) GPU board unit

[0070] like Figure 6 As shown, the GPU board (which can also be understood as an accelerator board) is located in the 2nd, 3rd, and 4th rows of the server. The accelerator card presence signal (GPU_PRSNT_N signal) of the GPU board is connected to the CPLD of the switch board to identify its presence status.

[0071] (V) Smart NIC and OCP NIC

[0072] The Smart NIC Riser provides the slot, power supply, and clock of the Smart NIC. The GND pin on the NCSI connector of the Smart NIC is short-circuited with the NIC_PRSNT_N signal through the NCSI cable. This is provided to the MB mainboard as in-place detection; the OCP_PRSNT_N signal of the OCP NIC is connected to the MB mainboard CPLD as in-place detection.

[0073] Among them, Smart NICs have powerful computing capabilities and can perform tasks such as encryption / decryption, firewalls, TCP / IP and HTTP processing. They can offload network functions, security functions and storage functions from the host server to release processing power; OCP NICs are mainly designed as bridges connecting servers with the outside world, focusing on providing high-speed, low-latency network communications. OCP NICs support multiple rate versions to meet different data center requirements. It can be seen that the functions of the two are different.

[0074] In the embodiment of the present application, the in-place status information of each component in the server can be received through the switch board CPLD, but is not limited to, for example, the GPU board is in-place, the smart network card is in-place, the OCP network card is not in-place, etc., and then the in-place status information of each component is sent to the fan board CPLD to query the duty cycle of the PWM signal corresponding to the N components in the current state in the relationship mapping table, and determine the position of at least one group of server fans that dissipate heat for at least part of the components currently in the in-place state, and finally control at least one group of server fans at this position to run at the speed indicated by the duty cycle of the PWM signal found. The relationship mapping table will be described in detail below in conjunction with specific embodiments.

[0075] Obviously, Figure 6 The partitioning of the three components in the server shown is only an example, and the situation of the server fan corresponding to each partition is also only an example and is not limited thereto.

[0076] Through the above method, after obtaining the status information of N components in the target server, at least one component that currently has a heat dissipation demand and the server fan in the target partition that has a position association relationship with the at least one component are intelligently identified according to the in-place status of the N components. Then, the duty cycle of the PWM signal corresponding to the combination mode of at least one component is found by searching the relationship mapping table, and the server fan in the target partition is controlled to operate at the speed indicated by the duty cycle of the found PWM signal, thereby realizing refined control of the server fan, achieving the technical effect that the server fan can run on demand, and reducing energy consumption.

[0077] In addition, the above method can also be used to make autonomous adjustments based on actual cooling needs, thereby improving the flexibility and response speed of the cooling system; at the same time, since the number of running fans is reduced and the speed of server fans is finely controlled, noise can be effectively reduced, the server environment is improved, and the server's operating efficiency and energy utilization are improved.

[0078] In an exemplary embodiment, the above-mentioned obtaining the status information of the control component of the target server in the shutdown state includes:

[0079] Sending a first PWM signal to a fan board CPLD of the target server through the switch board CPLD of the target server;

[0080] The first PWM signal is parsed by the fan board CPLD to obtain status information of each component in the N components.

[0081] It should be noted that the first PWM signal here is not used to control the rotation speed of the server fan, but carries the presence information of the GPU board, the smart network card and the OCP network card.

[0082] In an embodiment of the present application, components or parts with heat dissipation requirements are arranged in partitions to distinguish different areas with heat dissipation requirements; and the fan board is designed to be modular, separated and intelligent, so that the fan speed of the fan board can be automatically adjusted according to the in-place information of the smart network card, OCP network card and GPU board in the server architecture.

[0083] When the server is in S 5 state, some components such as CPLD and BMC may be in power supply state but are not limited to it. The partition heat dissipation control method uses CPLD to intelligently control the fan speed of the fan board according to the in-place information of each component, thereby reducing the energy consumption of the whole machine and optimizing the noise level.

[0084] At the same time, this method is particularly suitable for high-density data center environments and can effectively improve the operating efficiency of the server and the environmental comfort.

[0085] In an exemplary embodiment, before sending the first PWM signal to the fan board CPLD of the target server through the switch board CPLD of the target server, the method further includes:

[0086] In the case where the N components include an accelerator board, a smart network card, and an OCP network card, a smart network card in place signal and an OCP network card in place signal are sent to the switch board CPLD through the mainboard of the target server, wherein the smart network card in place signal and the OCP network card in place signal are both valid at a low level;

[0087] An acceleration card in-place signal is sent to the switch board CPLD through the acceleration board, wherein the acceleration card in-place signal is valid at a low level.

[0088] like Figure 4 As shown, the in-place status information of each component in the target server can be obtained in the following ways but not limited to.

[0089] S402, obtaining a smart network card presence signal (NIC_PRSNT_N) and an OCP network card presence signal (OCP_PRSNT_N) through a CPLD (MB CPLD) of a server mainboard;

[0090] For details, please refer to Figure 3 As shown, the main function of the MB CPLD is to receive the smart network card in-place signal and the OCP network card in-place signal, process the received smart network card in-place signal and the OCP network card in-place signal, and send the processed smart network card in-place signal and the OCP network card in-place signal to the CPLD of the switch board (hereinafter referred to as the switch board CPLD), that is, the smart network card in-place signal and the OCP network card in-place signal are sent to the switch board CPLD through the MB CPLD.

[0091] It should be noted that in the MB control unit, NIC_PRSNT_N is pulled up externally by P3V3_STBY and is high by default. When the NCSI cable is inserted, the NCSI cable will short the GND pin on the NCSI connector and the NIC_PRSNT_N signal together. A low level is generated, and the NIC_PRSNT_N low level signal is passed to the CPLD. The CPLD de-bounces the NIC_PRSNT_N and outputs CPLD_NIC_PRSNT_N to the Switch board CPLD.

[0092] S404, sending the processed smart network card in place signal and OCP network card in place signal to the switch board CPLD;

[0093] S406, receiving an acceleration card in place signal sent by the GPU board;

[0094] like Figure 3 As shown, the GPU_PRSNT_N signal of the GPU board is connected to the CPLD of the Switch board, and the user identifies its presence.

[0095] S408, outputting the duty cycle of the PWM signal;

[0096] like Figure 3 As shown in the figure, the CPLD of the Switch board receives NIC_PRSNT_N and OCP_PRSNT_N from the MB CPLD, and receives the Enable signals of VR0 and VR1 and the GPU_PRSNT_N signal from the GPU board. It also performs logical processing and judgment internally. It outputs PWM signals with different duty cycles to the CPLDs of FAN board0, 1, 2, and 3. The corresponding relationship between the combination of in-position signals and different PWM duty cycles is shown in the figure. Figure 5 The table shown.

[0097] S4 1 0, using the fan board CPLD, analyze the duty cycle of the PWM signal to obtain the in-position status information of each component.

[0098] It should be noted that the PWM signal here is not used to control the speed of the server fan, but carries the presence information of the GPU board, smart network card and OCP network card.

[0099] The fan board CPLD determines the current status of the three components or three boards by analyzing the PWM, for example, whether they are in the in-position state or not in the in-position state.

[0100] In this embodiment, the NIC_PRSNT_N signal can be processed by a pull-up resistor to ensure the stability of the signal. This optimization measure is particularly critical in the scenario where the server is frequently started and shut down, and can avoid misjudgment of heat dissipation control caused by signal fluctuations, ensuring the stable operation of the server in various working states.

[0101] Secondly, in the MB control unit, the MB CPLD performs de-jitter processing on the NIC_PRSNT_N signal to improve the accuracy of signal recognition. During signal transmission, de-jitter processing can eliminate signal noise, which can significantly improve the response speed and accuracy of the heat dissipation control system for server environments that require high-precision control, such as high-performance computing centers.

[0102] Furthermore, by introducing a clock buffer in the MB control unit, the transmission performance of the PCIe signal can be effectively enhanced. The addition of the clock buffer enables the server to maintain signal integrity and synchronization when processing large amounts of data, which is suitable for cloud computing and big data analysis scenarios that require high-speed data transmission.

[0103] In an exemplary embodiment, when at least one of the N components is in the on-state, finding the duty cycle of the PWM signal corresponding to the state information of the N components from the relationship mapping table includes at least one of the following:

[0104] In a case where the N components include an accelerator board, a smart network card, and an OCP network card, obtaining first state information of the accelerator board, second state information of the smart network card, and third state information of the OCP network card;

[0105] When the first state information indicates that the accelerator board is in a non-in-place state, the second state information indicates that the smart network card is in a non-in-place state, and the third state information indicates that the OCP network card is in a non-in-place state, finding out from the relationship mapping table that the duty cycle of the PWM signal is 0;

[0106] When the first state information indicates that the accelerator board is in a non-in-place state, the second state information indicates that the smart network card is in a non-in-place state, and the third state information indicates that the OCP network card is in-place, finding from the relationship mapping table that the duty cycle of the PWM signal is a first value;

[0107] When the first state information indicates that the accelerator board is in a non-in-place state, the second state information indicates that the smart network card is in a position state, and the third state information indicates that the OCP network card is in a non-in-place state, finding out from the relationship mapping table that the duty cycle of the PWM signal is a second value, wherein the second value is greater than the first value;

[0108] When the first state information indicates that the accelerator board is in the on-site state, the second state information indicates that the smart network card is in the off-site state, and the third state information indicates that the OCP network card is in the off-site state, finding out from the relationship mapping table that the duty cycle of the PWM signal is a third value, wherein the third value is greater than the second value;

[0109] When the first state information indicates that the accelerator board is in place and the second state information indicates that the smart network card is in place, the duty cycle of the PWM signal is found to be a fourth value from the relationship mapping table, wherein the fourth value is greater than the third value.

[0110] In a specific embodiment, it is assumed that the partitioning of N components is as follows: Figure 6 As shown, the Smart NIC and the OCP NIC are in the same area (which can also be understood as the same partition), and the GPU board is in another area.

[0111] The first row of fans is used to meet the heat dissipation requirements of the smart network card and the OCP network card. The second, third and fourth rows of fans are used to meet the heat dissipation requirements of the GPU board. After the CPLD of the server fan (referred to as the fan board CPLD) is powered on, the fan speed is controlled first. The fan board CPLD will monitor the status of the three components in real time.

[0112] Specifically, when the target server is in the S5 state, the specific method of controlling the server fans at various positions is as follows:

[0113] (1) If the CPLD detects that there are no components that require heat dissipation, such as OCP network cards, smart network cards, and GPUs, it will not power on the fan and set the fan to 0% duty output.

[0114] The basis for determining that there is no OCP network card, smart network card, or GPU is that the OCP network card presence signal is at a high level, the smart network card presence signal is at a high level, and the acceleration card presence signal is at a high level.

[0115] (2) When the OCP network card is detected to be in place, since the OCP is in the first row of the server, the CPLD on the fan board 0 powers on the fan and controls the fan speed to start from 10% duty, increase by 5% duty every 10 seconds, and reach a maximum speed of 20% duty. For details, refer to Figure 5 Strategy A in the table shown. The CPLD on the remaining fan boards controls the fans not to rotate, thereby optimizing the power consumption of the entire machine and reducing noise;

[0116] (3) When the smart network card is detected to be in place, the CPLD on fan board 0 powers on the fan because the smart network card is in the first row of the server. The fan speed is controlled to start from 10% duty and increase by 5% duty every 10 seconds, with a maximum speed of 40% duty. For details, refer to Figure 5 Strategy B in the table shown. The CPLD on the remaining fan boards controls the fans not to rotate, thereby optimizing the power consumption of the entire machine and reducing noise;

[0117] Combined with the description in the above embodiments, it can be seen that the smart network card is a flexible and programmable network card. It has computing power and can offload data processing functions in network, security, and storage that are not suitable for CPU processing to programmable hardware chips for execution, reducing CPU consumption and enabling servers to run key applications and operating systems more efficiently. The OCP network card is mainly designed to solve the scalability problem of data center networks and has the characteristics of high performance, strong scalability, and easy maintenance. The OCP network card improves data center efficiency through standardized and open hardware design, supports hot-swap functions, and reduces operation and maintenance costs.

[0118] Therefore, usually Smart NICs require more advanced cooling solutions, such as liquid cooling systems, to cope with the heat load generated by high-density computing. In contrast, OCP NICs may use air cooling design, transferring heat to the liquid cooling loop through a heat exchanger at the rear of the rack, which helps reduce cooling costs and complexity.

[0119] It is precisely because of the different functions and processing processes of Smart NICs and OCP NICs that the heat dissipation requirements of the two are different. For example, when only the Smart NIC is in place, the maximum speed of the server fan is 40% duty, and when only the OCP NIC is in place, the maximum speed is 20% duty.

[0120] (4) When it is detected that both the OCP network card and the smart network card are in place, since both the OCP network card and the smart network card are in the first row of the server, the CPLD on the fan board 0 recognizes this configuration and powers on the fan, controlling the fan speed to start from 10% duty and increase by 5% duty every 10 seconds, with a maximum speed of 40% duty. Figure 5Strategy B shown in the figure. The CPLD on the other fan boards controls the fans not to rotate, thereby optimizing the power consumption of the whole machine and reducing the noise.

[0121] (5) When it is detected that the GPU board is in place and the OCP network card or smart network card is not in place, since the GPU board is in the second, third, and fourth rows of the server, the CPLD on the fan board 1, fan board 2, and fan board 3 powers on the fans and controls the fan speed to start from 10% duty and increase by 5% duty every 10 seconds, with a maximum speed of 60% duty. Figure 5 Strategy C shown. The CPLD on the remaining fan boards controls the fans not to rotate, thereby optimizing the power consumption of the entire machine and reducing noise.

[0122] (6) When it is detected that the GPU board is in place and the OCP network card or smart network card is in place, the components that need heat dissipation are in the first, second, third, and fourth rows of the server. The CPLD on fan boards 0, 1, 2, and 3 powers on the fans and controls the fan speed to start at 10% duty and increase by 5% duty every 10 seconds, with a maximum speed of 100% duty. Figure 5 Strategy D shown;

[0123] (7) When it is detected that the GPU board is in place, the OCP network card is in place, and the smart network card is in place, the components that need heat dissipation are in the first, second, third, and fourth rows of the server. The CPLD on fan boards 0, 1, 2, and 3 powers on the fans and controls the fan speed to start at 10% duty, increase by 5% duty every 10 seconds, and reach a maximum speed of 100% duty. Figure 5 Strategy D shown.

[0124] Through the above embodiment, the corresponding PWM signal duty cycle can be found from the relationship mapping table according to the acquired status of each component, and the server fans at the positions corresponding to the areas where each component is located can be controlled to operate according to the found PWM signal duty cycle, thereby achieving on-demand heat dissipation and reducing server energy consumption.

[0125] In an exemplary embodiment, the above-mentioned controlling the server fan corresponding to the target partition to run at a preset speed based on the duty cycle of the PWM signal includes:

[0126] Determine the position information of each group of server fans in M ​​groups of server fans through a first input signal and a second input signal on a fan board CPLD of the target server, where M is a positive integer greater than or equal to 2;

[0127] Based on the status information of the N components, determine a target position of at least one group of server fans for cooling the components in the target partition, wherein the components in a partition require at least one group of server fans for cooling, and the M groups of server fans include the at least one group of server fans;

[0128] Based on the duty cycle of the PWM signal, the at least one group of server fans at the target location is controlled to operate at the preset speed.

[0129] like Figure 3 As shown, the fan board CPLD has two input signals, FAN_ID0 and FAN_ID1, which control the high and low level states on the fan adapter board. The fan board CPLD determines the position of the fan board by identifying the high and low levels.

[0130] For example, obtain the component that is currently in place, assuming it is a smart network card. Since a mapping relationship between the target partition where the smart network card is located and the input signals FAN_ID0 (low level) and FAN_ID1 (low level) of the fan board is created in advance, according to the component that is currently in place, the high and low levels of the input signals FAN_ID0 and FAN_ID1 of the fan board corresponding to these components can be directly obtained, and the high and low levels of the input signals FAN_ID0 and FAN_ID1 are fed back to the fan board CPLD.

[0131] The fan board CPLD identifies the position number of the fan board that meets the high and low level conditions of the input signals FAN_ID0 and FAN_ID 1, for example, fan board 0, fan board 1, or fan board 2, etc. Or identifies the first row of fans, the second row of fans, or the third row of fans, etc.

[0132] By adopting the above method, heat generated by at least some components in the in-place state can be dissipated only, while the server fans corresponding to the partitions where other components not in the in-place state are located are not started, thereby fundamentally achieving on-demand heat dissipation and reducing energy consumption.

[0133] In an exemplary embodiment, determining the position information of each group of server fans in the M groups of server fans by using the first input signal and the second input signal on the fan board CP LD of the target server includes:

[0134] Sequentially obtain a group of server fans from the M groups of server fans as a current group of server fans;

[0135] When the first input signal is at a low level and the second input signal is at a low level, determining that the current group of server fans is located at a first position in the fan partition;

[0136] When the first input signal is at a low level and the second input signal is at a high level, determining that the current group of server fans is located at a second position in the fan partition;

[0137] When the first input signal is at a high level and the second input signal is at a low level, determining that the current group of server fans is located at a third position in the fan partition;

[0138] When the first input signal is at a low level and the second input signal is at a low level, it is determined that the current group of server fans is located at a fourth position in the fan zone.

[0139] As described in the above embodiment, by creating a relationship between the partition where each component is located and the position of the server fan, the target position corresponding to the component currently in place can be quickly found, and then at least one group of server fans at the target position can be determined through the fan board CPLD.

[0140] Specifically, the fan board CPLD confirms the position of the fan board by identifying high and low levels. When it is identified that the input signal FAN_ID0 is a low level and FAN_ID 1 is a low level, the target position is determined to be fan board 0 corresponding to the first fan partition (which can also be understood as the first position), that is, the first row of fans that can meet the heat dissipation requirements of the smart network card and the OCP network card; when it is identified that the input signal FAN_1D0 is a low level and FAN_ID 1 is a high level, the target position is determined to be fan board 1 corresponding to the second fan partition (which can also be understood as the second position); when it is identified that the input signal FAN_ID0 is a high level and FAN_ID 1 is a low level, the target position is determined to be fan board 2 corresponding to the third fan partition (which can also be understood as the third position); when it is identified that the input signal FAN_1D0 is a high level and FAN_ID 1 is a high level, the target position is determined to be fan board 3 corresponding to the fourth fan partition (which can also be understood as the fourth position).

[0141] In other words, the combination information of the high and low levels of the two input signals FAN_ID0 and FAN_ID1 is 00 for fan board 0, 01 for fan board 1, 10 for fan board 2, and 11 for fan board 3.

[0142] The fan board CPLD searches for the following information based on the identified location information and the status information of the GPU board, smart network card, and OCP network card received by the switch board CPLD: Figure 5 The relationship mapping table shown in the relationship mapping table is used, and different PWM duty cycles are output to fans at different positions according to the control logic shown in the relationship mapping table, so as to control the speed of the server fans in each partition.

[0143] Obviously, Figure 6 The first row of fans shown is only an example. In actual application scenarios, the first row of fans may include one fan, or may include two or more fans, which is limited in the embodiments of the present application.

[0144] Similarly, the second row of fans, the third row of fans, and the fourth row of fans that provide heat dissipation for the GPU board are only examples, and the first row of fans that provide heat dissipation for the smart network card and the OCP network card are only examples. In actual application scenarios, they can be adjusted as needed.

[0145] By adopting the above method, it is possible to accurately locate the components that need heat dissipation, quickly determine the position of the server fan that provides heat dissipation for the component, and then control the server fan at that position to operate at the speed indicated by the duty cycle of the PWM signal output by the switch board. This method can automatically adjust the input voltage according to the load change, ensure that under different workload conditions, the server fan that needs to be operated is adaptively selected, and realize the dynamic adjustment of the overall power consumption of the server.

[0146] In an exemplary embodiment, the method further includes:

[0147] Adding an in-place status signal of the N+1th component in the relationship mapping table;

[0148] Based on the in-position state signal of the N+1th component, updating the mapping relationship between the state information of the N+1th component and the duty cycle of the PWM signal to obtain an updated relationship mapping table;

[0149] When the current state information of the N+1 components is obtained, searching the updated relationship mapping table for the current duty cycle of the PWM signal corresponding to the current state information;

[0150] Based on the current duty cycle of the PWM signal, the current server fan corresponding to the current partition is controlled to run at the current speed, wherein the current partition includes the server partition where the components in the N+1 components that are in place at the current moment are located.

[0151] The N components in the above embodiments are all taken as examples of GPU boards, smart network cards and OCP network cards. In addition, in actual application scenarios, other heat-generating components may also be included, such as power modules. Similar to the heat dissipation requirements of the above GPU boards, smart network cards and OCP network cards, the power modules also need to be cooled when the server is in the shutdown state.

[0152] If you want to control the server fans corresponding to the partition where the power module is located according to the above server fan control strategy to perform heat dissipation, you can, but are not limited to, use the following processing to achieve this:

[0153] S11, obtaining the partition where the power module is located;

[0154] S12, creating a correspondence between the partition where the power module is located and the position of the server fan;

[0155] For example, suppose Figure 6 The GPU area is shown as the first partition, and the smart NIC and OCP NIC areas are the second partition. Then the area where the power module is located is determined as the third partition, a fifth row of fans is added, and an association relationship is created between the location of the fifth row of fans and the third partition.

[0156] S13, adding the power module in-place signal, and updating the power module in-place signal according to the combination relationship between the power module in-place status signal, the acceleration card in-place signal, the smart network card in-place signal and the OCP network card in-place signal. Figure 5 The fan control strategy shown;

[0157] For example, when the power module on-site signal is low, the acceleration card on-site signal is high, the intelligent network card on-site signal is high, and the OCP network card on-site signal is high, the PWM signal duty cycle is 10%, and the position of the working fan is the fifth row where the fan board 4 is located; when the power module on-site signal is high, the acceleration card on-site signal is low, the intelligent network card on-site signal is high, and the OCP network card on-site signal is high, the PWM signal duty cycle is 60%, and the position of the working fan is the second to fourth rows where the fan boards 1, 2, and 3 are located, etc.

[0158] S14: According to the high and low levels of the in-position signals of the power module, GPU board, smart network card, and OCP network card at the current moment, find the PWM duty ratio corresponding to the current combination state of the in-position signals, and control the corresponding server fan to run at the corresponding speed.

[0159] For example, when only the power module is in place, the fan is powered on through the CPLD of fan board 4, and the speed of the fifth row of fans on fan board 4 is controlled to start from 10% duty, increase by 5% duty per second, and the maximum speed reaches 20% duty; when the power module is in place and the smart network card is in place, the fan is powered on through the CPLD of fan board 4 and the fan board 0, and the speed of the fifth row of fans on fan board 4 and the first row of fans on fan board 0 is controlled to start from 10% duty, increase by 5% duty per second, and the maximum speed reaches 45% duty.

[0160] It can be seen that in the embodiment of the present application, components with heat dissipation requirements in the server in the shutdown state can be adjusted in real time as needed, and the relationship mapping table is updated according to the adjusted components. Then, the corresponding PWM signal duty cycle is queried according to the in-place signals of each component at the current moment, and the fan at the corresponding position is controlled to operate according to the PWM signal duty cycle, which improves the flexibility of controlling the server fan and enhances the compatibility of the solution.

[0161] 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 a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods of each embodiment of the present application.

[0162] In this embodiment, a control device for a server fan is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware for a predetermined function. Although the modules described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0163] Figure 7 : is a structural block diagram of a control device for a server fan according to an embodiment of the present application, the device comprising:

[0164] A first acquisition unit 702 is used to acquire state information of a control component of a target server in a shutdown state, wherein the state information of the control component includes an in-place state and an out-of-place state, and the control component includes N components, where N is a positive integer greater than or equal to 1;

[0165] A search unit 704 is used to search the duty cycle of the PWM signal corresponding to the status information of the N components from the relationship mapping table when at least one component among the N components is in the in-position state, wherein the higher the duty cycle of the PWM signal, the higher the rotation speed of the server fan;

[0166] The first control unit 706 is used to control the server fan corresponding to the target partition to run at a preset speed based on the duty cycle of the PWM signal, wherein the target server supports multiple partitions, and each of the N components is arranged in one of the multiple partitions according to a preset arrangement rule, and a group of components in the target partition are cooled by running the server fan.

[0167] In an exemplary embodiment, the first obtaining unit 702 includes:

[0168] A first sending module, configured to send a first PWM signal to a fan board CPLD of the target server through a switch board CPLD of the target server;

[0169] The parsing module is used to parse the first PWM signal through the fan board CPLD to obtain the status information of each component in the N components.

[0170] In an exemplary embodiment, the above device further includes:

[0171] A first sending unit is used for sending a smart network card in place signal and an OCP network card in place signal to the switch board CPLD through the mainboard of the target server when the N components include an accelerator board, a smart network card and an OCP network card, wherein the smart network card in place signal and the OCP network card in place signal are both valid at a low level;

[0172] The second sending unit is used to send an acceleration card presence signal to the switch board CPLD through the acceleration board, wherein the acceleration card presence signal is valid at a low level.

[0173] In an exemplary embodiment, the search unit 704 includes:

[0174] The first processing module is configured to perform at least one of the following:

[0175] In a case where the N components include an accelerator board, a smart network card, and an OCP network card, obtaining first state information of the accelerator board, second state information of the smart network card, and third state information of the OCP network card;

[0176] When the first state information indicates that the accelerator board is in a non-in-place state, the second state information indicates that the smart network card is in a non-in-place state, and the third state information indicates that the OCP network card is in a non-in-place state, finding out from the relationship mapping table that the duty cycle of the PWM signal is 0;

[0177] When the first state information indicates that the accelerator board is in a non-in-place state, the second state information indicates that the smart network card is in a non-in-place state, and the third state information indicates that the OCP network card is in-place, finding from the relationship mapping table that the duty cycle of the PWM signal is a first value;

[0178] When the first state information indicates that the accelerator board is in a non-in-place state, the second state information indicates that the smart network card is in a position state, and the third state information indicates that the OCP network card is in a non-in-place state, finding out from the relationship mapping table that the duty cycle of the PWM signal is a second value, wherein the second value is greater than the first value;

[0179] When the first state information indicates that the accelerator board is in the on-site state, the second state information indicates that the smart network card is in the off-site state, and the third state information indicates that the OCP network card is in the off-site state, finding out from the relationship mapping table that the duty cycle of the PWM signal is a third value, wherein the third value is greater than the second value;

[0180] When the first state information indicates that the accelerator board is in place and the second state information indicates that the smart network card is in place, the duty cycle of the PWM signal is found to be a fourth value from the relationship mapping table, wherein the fourth value is greater than the third value.

[0181] In an exemplary embodiment, the first control unit 706 includes:

[0182] A second processing module, configured to determine position information of each group of server fans in M ​​groups of server fans through a first input signal and a second input signal on a fan board CPLD of the target server, wherein M is a positive integer greater than or equal to 2;

[0183] a third processing module, configured to determine, based on the status information of the N components, a target position of at least one group of server fans for cooling the components in the target partition, wherein the components in one partition require at least one group of server fans for cooling the components, and the M groups of server fans include the at least one group of server fans;

[0184] The fourth processing module is used to control the at least one group of server fans at the target position to operate at the preset speed based on the duty cycle of the PWM signal.

[0185] In an exemplary embodiment, the second processing module includes:

[0186] An acquisition submodule, used for sequentially acquiring a group of server fans from the M groups of server fans as a current group of server fans;

[0187] A first processing submodule, configured to determine that the current group of server fans is located at a first position in the fan partition when the first input signal is at a low level and the second input signal is at a low level;

[0188] A second processing submodule, configured to determine that the current group of server fans is located at a second position in the fan partition when the first input signal is at a low level and the second input signal is at a high level;

[0189] A third processing submodule, configured to determine that the current group of server fans is located at a third position in the fan partition when the first input signal is at a high level and the second input signal is at a low level;

[0190] The fourth processing submodule is used to determine that the current group of server fans is located at a fourth position in the fan partition when the first input signal is at a low level and the second input signal is at a low level.

[0191] In an exemplary embodiment, the above device further includes:

[0192] An adding unit, used for adding an in-place status signal of the N+1th component in the relationship mapping table;

[0193] an updating unit, configured to update the mapping relationship between the state information of the N+1th component and the duty cycle of the PWM signal based on the in-position state signal of the N+1th component, and obtain an updated relationship mapping table;

[0194] a first processing unit, configured to search the current duty cycle of the PWM signal corresponding to the current state information from the updated relationship mapping table when the current state information of the N+1 components is acquired;

[0195] The second processing unit is used to control the current server fan corresponding to the current partition to run at the current speed based on the current duty cycle of the PWM signal, wherein the current partition includes the server partition where the components in the N+1 components that are in place at the current moment are located.

[0196] By applying the above device to obtain the status information of N components in the target server, at least one component that currently has a heat dissipation demand and the server fan in the target partition that has a position association relationship with the at least one component are intelligently identified according to the in-place status of the N components. Then, the duty cycle of the PWM signal corresponding to the combination mode of at least one component is found by searching the relationship mapping table, and the server fan in the target partition is controlled to operate at the speed indicated by the duty cycle of the found PWM signal, thereby realizing refined control of the server fan, achieving the technical effect that the server fan can run on demand, and reducing energy consumption.

[0197] It should be noted that the above modules can be implemented by software or hardware. For the latter, it can be implemented in the following ways, but not limited to: the above modules are all located in the same processor; or the above modules are located in different processors in any combination.

[0198] An embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored, wherein the computer program is configured to execute the steps of any of the above method embodiments when running.

[0199] Optionally, in this embodiment, the computer program may be configured to perform the following steps by means of the computer program:

[0200] S1, obtaining status information of a control component of a target server in a shutdown state, wherein the status information of the control component includes an in-place state and an out-of-place state, and the control component includes N components, where N is a positive integer greater than or equal to 1;

[0201] S2, when at least one of the N components is in the on-state, searching the relationship mapping table for a duty cycle of a PWM signal corresponding to the state information of the N components, wherein a higher duty cycle of the PWM signal is, a higher rotation speed of the server fan is;

[0202] S3, based on the duty cycle of the PWM signal, controlling the server fan corresponding to the target partition to run at a preset speed, wherein the target server supports multiple partitions, and each of the N components is arranged in one of the multiple partitions according to a preset arrangement rule, and cooling a group of components in the target partition is performed by running the server fan.

[0203] 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.

[0204] The embodiment of the present application also provides an electronic device, such as Figure 8 As shown, the electronic device includes a memory 802 and a processor 804. The memory 802 stores a computer program, and the processor 804 is configured to execute the steps in any of the above method embodiments through the computer program.

[0205] Optionally, in this embodiment, the processor 804 may be configured to perform the following steps through a computer program:

[0206] S1, obtaining status information of a control component of a target server in a shutdown state, wherein the status information of the control component includes an in-place state and an out-of-place state, and the control component includes N components, where N is a positive integer greater than or equal to 1;

[0207] S2, when at least one of the N components is in the on-state, searching the relationship mapping table for a duty cycle of a PWM signal corresponding to the state information of the N components, wherein a higher duty cycle of the PWM signal is, a higher rotation speed of the server fan is;

[0208] S3, based on the duty cycle of the PWM signal, controlling the server fan corresponding to the target partition to run at a preset speed, wherein the target server supports multiple partitions, and each of the N components is arranged in one of the multiple partitions according to a preset arrangement rule, and cooling a group of components in the target partition is performed by running the server fan.

[0209] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail herein.

[0210] Alternatively, a person skilled in the art may understand that: Figure 8 The structure shown is for illustration only. Figure 8 The structure of the electronic device is not limited. For example, the electronic device may also include Figure 8 More or fewer components (such as network interfaces, etc.) as shown in, or with Figure 8 Different configurations shown.

[0211] Among them, the memory 802 can be used to store software programs and modules, such as the program instructions / modules corresponding to the control method of the server fan and the control device of the server fan in the embodiment of the present application. The processor 804 executes various functional applications and data processing by running the software programs and modules stored in the memory 802, that is, realizing the above-mentioned control method of the server fan. The memory 802 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 802 may further include a memory remotely arranged relative to the processor 804, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned networks include but are not limited to the Internet, corporate intranets, local area networks, mobile communication networks and combinations thereof. Among them, the memory 802 can be specifically but not limited to being used to store the duty cycle of the PWM signal, the status information of the control component, etc. As an example, such as Figure 8 As shown, the memory 802 may include but is not limited to the first acquisition unit 702, search unit 704 and first control unit 706 in the control device for the server fan. In addition, it may also include but is not limited to other module units in the control device for the server fan, which will not be repeated in this example.

[0212] Optionally, the transmission device 806 is used to receive or send data via a network. Specific examples of the network may include wired networks and wireless networks. In one example, the transmission device 806 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices and routers via a network cable so as to communicate with the Internet or a local area network. In one example, the transmission device 806 is a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0213] In addition, the electronic device further includes: a display 808; and a connection bus 810, which is used to connect various module components in the electronic device.

[0214] In other embodiments, the electronic device may be a node in a distributed system, wherein the distributed system may be a blockchain system, and the blockchain system may be a distributed system formed by connecting the multiple nodes in the form of network communication. Among them, the nodes may form a peer-to-peer (P2P) network, and any form of computing device, such as a server, terminal, or other electronic device, may become a node in the blockchain system by joining the peer-to-peer network.

[0215] 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 method embodiments are implemented.

[0216] An embodiment of the present application further provides another computer program product, including a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in any of the above method embodiments are implemented.

[0217] The embodiments of the present application also provide a computer program, which includes computer instructions, which are stored in a computer-readable storage medium; a processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device performs the steps of any one of the above method embodiments.

[0218] Obviously, those skilled in the art should understand that the above modules or steps of the present application can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order from that herein, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present application is not limited to any specific combination of hardware and software.

[0219] The above are only preferred embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for controlling a server fan, characterized in that: include: Acquire status information of a control component of a target server in a shutdown state, wherein the status information of the control component includes an in-place state and an out-of-place state, and the control component includes N components, where N is a positive integer greater than or equal to 1; When at least one of the N components is in the on-state, a duty cycle of a PWM signal corresponding to the state information of the N components is found from a relationship mapping table, wherein a higher duty cycle of the PWM signal indicates a higher rotation speed of the server fan; Based on the duty cycle of the PWM signal, the server fan corresponding to the target partition is controlled to run at a preset speed, wherein the target server supports multiple partitions, and each of the N components is arranged in one of the multiple partitions according to a preset arrangement rule, and the server fan is run to dissipate heat for a group of components in the target partition.

2. The method according to claim 1, characterized in that The step of obtaining the status information of the control component of the target server in the shutdown state includes: Sending a first PWM signal to a fan board CPLD of the target server through the switch board CPLD of the target server; The first PWM signal is parsed by the fan board CPLD to obtain status information of each component in the N components.

3. The method according to claim 2, characterized in that Before sending the first PWM signal to the fan board CPLD of the target server through the switch board CPLD of the target server, the method further includes: In the case where the N components include an accelerator board, a smart network card, and an OCP network card, a smart network card in place signal and an OCP network card in place signal are sent to the switch board CPLD through the mainboard of the target server, wherein the smart network card in place signal and the OCP network card in place signal are both valid at a low level; An acceleration card in-place signal is sent to the switch board CPLD through the acceleration board, wherein the acceleration card in-place signal is valid at a low level.

4. The method according to claim 1, characterized in that When at least one of the N components is in the on-state, searching the relationship mapping table for a duty cycle of a PWM signal corresponding to the state information of the N components includes at least one of the following: In a case where the N components include an accelerator board, a smart network card, and an OCP network card, obtaining first state information of the accelerator board, second state information of the smart network card, and third state information of the OCP network card; When the first state information indicates that the accelerator board is in a non-in-place state, the second state information indicates that the smart network card is in a non-in-place state, and the third state information indicates that the OCP network card is in a non-in-place state, finding out from the relationship mapping table that the duty cycle of the PWM signal is 0; When the first state information indicates that the accelerator board is in a non-in-place state, the second state information indicates that the smart network card is in a non-in-place state, and the third state information indicates that the OCP network card is in-place, finding from the relationship mapping table that the duty cycle of the PWM signal is a first value; When the first state information indicates that the accelerator board is in a non-in-place state, the second state information indicates that the smart network card is in a position state, and the third state information indicates that the OCP network card is in a non-in-place state, finding out from the relationship mapping table that the duty cycle of the PWM signal is a second value, wherein the second value is greater than the first value; When the first state information indicates that the accelerator board is in the on-site state, the second state information indicates that the smart network card is in the off-site state, and the third state information indicates that the OCP network card is in the off-site state, finding out from the relationship mapping table that the duty cycle of the PWM signal is a third value, wherein the third value is greater than the second value; When the first state information indicates that the accelerator board is in place and the second state information indicates that the smart network card is in place, the duty cycle of the PWM signal is found to be a fourth value from the relationship mapping table, wherein the fourth value is greater than the third value.

5. The method according to claim 1, characterized in that The step of controlling the server fan corresponding to the target partition to run at a preset speed based on the duty cycle of the PWM signal includes: Determine the position information of each group of server fans in M ​​groups of server fans through a first input signal and a second input signal on a fan board CPLD of the target server, where M is a positive integer greater than or equal to 2; Based on the status information of the N components, determine a target position of at least one group of server fans for cooling the components in the target partition, wherein the components in a partition require at least one group of server fans for cooling, and the M groups of server fans include the at least one group of server fans; Based on the duty cycle of the PWM signal, the at least one group of server fans at the target location is controlled to operate at the preset speed.

6. The method according to claim 5, characterized in that The determining, by using a first input signal and a second input signal on a fan board CPLD of the target server, position information of each group of server fans in the M groups of server fans comprises: Sequentially obtain a group of server fans from the M groups of server fans as a current group of server fans; when the first input signal is at a low level and the second input signal is at a low level, determine that the current group of server fans is located at a first position in the fan partition; When the first input signal is at a low level and the second input signal is at a high level, determining that the current group of server fans is located at a second position in the fan partition; When the first input signal is at a high level and the second input signal is at a low level, determining that the current group of server fans is located at a third position in the fan partition; When the first input signal is at a low level and the second input signal is at a low level, it is determined that the current group of server fans is located at a fourth position in the fan zone.

7. The method according to any one of claims 1 to 6, characterized in that The method further comprises: Adding an in-place status signal of the N+1th component in the relationship mapping table; Based on the in-position state signal of the N+1th component, updating the mapping relationship between the state information of the N+1th component and the duty cycle of the PWM signal to obtain an updated relationship mapping table; When the current state information of the N+1 components is obtained, searching the updated relationship mapping table for the current duty cycle of the PWM signal corresponding to the current state information; Based on the current duty cycle of the PWM signal, the current server fan corresponding to the current partition is controlled to run at the current speed, wherein the current partition includes the server partition where the components in the N+1 components that are in place at the current moment are located.

8. A control device for a server fan, characterized in that: include: A first acquisition unit is used to acquire state information of a control component of a target server in a shutdown state, wherein the state information of the control component includes an in-place state and an out-of-place state, and the control component includes N components, where N is a positive integer greater than or equal to 1; a search unit, configured to search a duty cycle of a PWM signal corresponding to the status information of the N components from a relationship mapping table when at least one of the N components is in an on-position state, wherein the higher the duty cycle of the PWM signal, the higher the rotation speed of the server fan; A first control unit is used to control a server fan corresponding to a target partition to run at a preset speed based on a duty cycle of the PWM signal, wherein the target server supports multiple partitions, each of the N components is arranged in one of the multiple partitions according to a preset arrangement rule, and a group of components in the target partition are cooled by running the server fan.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program implements the steps of the method described in any one of claims 1 to 7 when executed by a processor.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method described in any one of claims 1 to 7 are implemented.

Images