Heat dissipation control device and server

Abstract

The invention discloses a heat dissipation control device and a server, relates to the technical field of servers, and controls a heat dissipation assembly through a first controller when a substrate management controller fails. The heat dissipation control device comprises a substrate management controller, a plurality of heat dissipation controllers, a temperature information switching unit, a temperature sensor group and a heat dissipation assembly, the substrate management controller is connected with the temperature information switching unit and is also connected with the heat dissipation controller, and the temperature information switching unit is connected with the temperature sensor group and is also connected with the heat dissipation controller; the heat dissipation controller is electrically connected with the heat dissipation assembly; when the substrate management controller works normally, the substrate management controller obtains temperature information of the temperature sensor group through the temperature information switching unit and transmits the temperature information to the heat dissipation controller, and the heat dissipation assembly is adjusted; and when the substrate management controller works abnormally, the heat dissipation controller switches the temperature information of the unit temperature sensor group through the temperature information, and adjusts the heat dissipation assembly according to the temperature information.

Description

Technical Field

[0001] This application relates to the field of server technology, and in particular to a heat dissipation control device and a server. Background Technology

[0002] With the rapid development of cloud computing, artificial intelligence, and high-performance computing, the computing density and power consumption of servers have increased dramatically, with single-machine power consumption exceeding 500 watts. Against this backdrop, heat dissipation efficiency directly determines the operational stability, energy efficiency, and hardware lifespan of servers. Furthermore, approximately 40% of the energy consumption in data centers is used for heat dissipation, making the optimization of heat dissipation technology a core challenge for reducing operating costs and achieving green and low-carbon goals.

[0003] Current servers commonly employ air cooling, with the mainstream solution involving multiple fans arranged within the chassis. These fans are typically mounted on fan plates and controlled by Complex Programmable Logic Devices (CPLDs) on the plates. The Baseboard Management Controller (BMC) sends commands to the CPLDs via the I2C bus, adjusting the fan speeds to create convection cooling that draws in cool air from the front / bottom of the chassis and exhausts hot air from the rear / top.

[0004] However, when the baseboard management controller malfunctions or becomes inactive, the programmable logic devices (PLDs) cannot obtain information from the overall system temperature sensors and can only drive the fans to run at a fixed high speed, resulting in unnecessary energy consumption and noise, and they cannot make precise adjustments based on real-time temperature. If the PLD itself is damaged, the fans it controls will be completely out of control, seriously affecting heat dissipation safety. In high-density servers using multiple fan boards, the PLDs of each fan board have difficulty coordinating to read global temperature information, and when a fault occurs, it is difficult to accurately locate the specific faulty fan board, increasing the complexity and risk of operation and maintenance. Summary of the Invention

[0005] This application provides a heat dissipation control device and server to at least solve the problem that when the baseboard management controller fails or becomes inactive, the entire heat dissipation control device can only adopt a fixed, high-energy-consuming, and high-noise heat dissipation mode because it cannot obtain temperature information.

[0006] This application provides a heat dissipation control device, which includes: a substrate management controller, multiple heat dissipation controllers, a temperature information switching unit, a temperature sensor group, and a heat dissipation component; the substrate management controller is connected to the temperature information switching unit and also to the heat dissipation controllers, the temperature information switching unit is connected to the temperature sensor group and also to the heat dissipation controllers; the heat dissipation controllers are electrically connected to the heat dissipation component; when the substrate management controller is working normally, the substrate management controller obtains the temperature information of the temperature sensor group through the temperature information switching unit and transmits the temperature information to the heat dissipation controllers to adjust the heat dissipation component; when the substrate management controller is malfunctioning, the heat dissipation controller obtains the temperature information of the temperature sensor group through the temperature information switching unit and adjusts the heat dissipation component according to the temperature information.

[0007] This application also provides a server, which includes the above-described heat dissipation control device.

[0008] This application also provides an electronic device, including: a memory for storing a computer program; and a processor for executing the computer program to implement the control process of the above-mentioned heat dissipation control device when the board management controller malfunctions.

[0009] This application also provides a computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, it implements the control process of the above-mentioned heat dissipation control device when the board management controller malfunctions.

[0010] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the control process of the above-mentioned heat dissipation control device when the board management controller malfunctions.

[0011] This application introduces a temperature information switching unit to achieve intelligent switching of control over the heat dissipation control device between the baseboard management controller and the heat dissipation controller. This fundamentally avoids the problem of the heat dissipation control device failing to operate normally due to a single point of failure in the baseboard management controller, ensuring that the system can perform precise heat dissipation based on real-time temperature data under any condition. This significantly reduces unnecessary energy consumption and operating noise while ensuring the high-temperature reliability of the server, thereby improving the overall system's energy efficiency and availability. Attached Figure Description

[0012] To more clearly illustrate the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1This is a schematic diagram of a fan board circuit in the prior art; Figure 2 This is a schematic diagram of the structure of a heat dissipation control device provided in an embodiment of this application; Figure 3 This is a schematic diagram of another heat dissipation control device provided in an embodiment of this application; Figure 4 This is a schematic diagram of another heat dissipation control device provided in an embodiment of this application. Attached image description: 1. Baseboard management controller; 2. Heat dissipation controller; 3. Temperature information switching unit; 31. First multiplexer; 4. Temperature sensor group; 5. Heat dissipation component; 6. Buffer; 7. Signal selector; 8. Fan board; 9. First switch; 10. Second switch; 100. Heat dissipation control device. Detailed Implementation

[0015] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of this application.

[0016] It should be noted that, in the description of this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. The terms "first," "second," etc., in this application are used to distinguish similar objects and are not used to describe a specific order or sequence.

[0017] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0018] Current servers commonly employ air cooling, with the mainstream solution involving multiple fans arranged within the chassis. These fans are typically mounted on fan plates and controlled by Complex Programmable Logic Devices (CPLDs) on the plates. The Baseboard Management Controller (BMC) sends commands to the CPLDs via the I2C bus, adjusting the fan speeds to create convection cooling that draws in cool air from the front / bottom of the chassis and exhausts hot air from the rear / top.

[0019] For example, refer to Figure 1 , Figure 1 This is a schematic diagram of an existing fan board circuit.

[0020] Among them, PRSNT is the fan presence signal, used by the CPLD to detect whether the fan is in position; TACH is the fan speed signal. The CPLD can obtain the fan speed in real time by reading the frequency of the TACH signal, thereby understanding the fan's operating status. If the fan malfunctions, such as stopping or abnormal speed, the TACH signal will change accordingly. The system can detect these anomalies and issue alarms in a timely manner to remind maintenance personnel to handle the situation and avoid server hardware damage due to poor heat dissipation; FAN_NORMAL / FAULT signal: a fan normal / abnormal signal issued by the CPLD, controlling the green / red LEDs on the fan board. When the fan malfunctions, the CPLD informs the user through the LED changes on the control board; PWM refers to pulse width modulation speed control signal, which linearly controls the fan speed by adjusting the duty cycle of the PWM signal. The frequency of the PWM signal is generally 25K to 50K, and the duty cycle is adjustable from 10% to 100%. This method can precisely control the fan speed, achieving energy saving and noise reduction. The programmable logic device (CPLD / MCU) on the fan board can be programmed to precisely control the fan speed, achieving the purpose of server heat dissipation. FM_MB_S0_DETECT_R is the power-on signal detected by the fan board. FM_BMC_WDT_FB_R is the watchdog signal of the BMC. Before the BMC completes initialization, the CPLD controls the fan speed. After the BMC completes initialization, the BMC controls the fan according to the heat dissipation strategy. FM_FAN_PWR_PLD_R_EN is the enable signal for the fan to work when the power is off. Some components also need heat dissipation when the power is off. The upstream device of the fan board will detect the presence of these devices. When the power is off, heat dissipation is required, so this signal is transmitted to the CPLD to control the fan speed.

[0021] This fan board uses a programmable logic device as its core control unit: the left side interacts with the outside (such as a server motherboard) with various signals through the Sideband interface, including I2C bus, power-on detection, BMC watchdog feedback (distinguishing the fan control stage between the CPLD and BMC), power-off thermal enable, etc.; at the same time, the CPLD is connected to a thermal sensor and EEPROM, and can also be debugged through the JTAG interface, while the power supply is provided by P3V3_STBY, etc.

[0022] The CPLD is responsible for precise control and status feedback of the fans: it outputs a pulse width modulation signal to the fan interface through a PWM isolation circuit to adjust the fan speed, reads the fan speed and detects the fan presence signal through a TACH isolation circuit, and outputs FAN_NORMAL / FAULT signals based on this information to control the green / red LEDs to display the normal / abnormal status of the fans. At the same time, it works with a hot-swap module to support hot-swapping of fans and a power monitoring module to ensure power supply, ultimately realizing intelligent control and visual monitoring of server heat dissipation.

[0023] However, when the baseboard management controller malfunctions or becomes inactive, the programmable logic devices (PLDs) cannot obtain information from the overall system temperature sensors and can only drive the fans to run at a fixed high speed, resulting in unnecessary energy consumption and noise, and they cannot make precise adjustments based on real-time temperature. If the PLD itself is damaged, the fans it controls will be completely out of control, seriously affecting heat dissipation safety. In high-density servers using multiple fan boards, the PLDs of each fan board have difficulty coordinating to read global temperature information, and when a fault occurs, it is difficult to accurately locate the specific faulty fan board, increasing the complexity and risk of operation and maintenance.

[0024] For example, a high-performance server contains eight high-power GPUs and multiple CPUs, and is equipped with four independent heat dissipation components (fan modules) for cooling.

[0025] In the traditional approach: when the baseboard management controller is working properly, it directly reads data from all temperature sensor groups via the I2C bus, calculates the optimal fan speed, and executes it by controlling the heat dissipation controller on each fan module.

[0026] In case of abnormal operation of the baseboard management controller: At this time, the heat dissipation controller on each fan module loses the communication path with the temperature sensor. In order to prevent overheating, it can only execute the preset safety mode, which makes all fans run continuously at 100% maximum speed.

[0027] This leads to two serious problems: low energy efficiency and noise pollution. Even when the server is under low load and the temperature is not high, the fans still run at full speed, resulting in huge power waste and noise. The heat dissipation is not precise and cannot be dynamically adjusted according to the real-time hotspots of the GPU / CPU, which may lead to the risk of localized overheating or overcooling.

[0028] Based on this, this application provides a heat dissipation control device, such as... Figure 2 As shown, the heat dissipation control device 100 includes: a substrate management controller 1, multiple heat dissipation controllers 2, a temperature information switching unit 3, a temperature sensor group 4, and a heat dissipation component 5.

[0029] The substrate management controller 1 is connected to the temperature information switching unit 3 and also to the heat dissipation controller 2. The temperature information switching unit 3 is connected to the temperature sensor group 4 and also to the heat dissipation controller 2. The heat dissipation controller 2 is electrically connected to the heat dissipation component 5.

[0030] When the substrate management controller 1 is working normally, the substrate management controller 1 obtains the temperature information of the temperature sensor group 4 through the temperature information switching unit 3 and transmits the temperature information to the heat dissipation controller 2 to adjust the heat dissipation component 5; when the substrate management controller 1 is malfunctioning, the heat dissipation controller 2 obtains the temperature information of the temperature sensor group 4 through the temperature information switching unit 3 and adjusts the heat dissipation component 5 according to the temperature information.

[0031] In some embodiments, the heat dissipation component 5 is a fan.

[0032] In other words, the temperature information switching unit 3 is used to establish the first communication line between the substrate management controller 1 and the temperature sensor group 4 when the substrate management controller 1 is working normally.

[0033] When the substrate management controller 1 malfunctions, in response to a switching command issued by at least one heat dissipation controller 2, the first communication line is disconnected and a second communication line is established between the heat dissipation controller 2 that issued the switching command and the temperature sensor group 4.

[0034] For example, a high-performance server contains eight high-power GPUs and multiple CPUs, and is equipped with four independent heat dissipation components (fan modules) for cooling.

[0035] When the substrate management controller 1 is working normally: This unit switches the path of the temperature sensor group 4 to the substrate management controller 1, and the substrate management controller 1 performs global intelligent control. The working mode is the same as the traditional solution.

[0036] When the baseboard management controller 1 malfunctions: any one or more heat dissipation controllers 2 can issue a command to control the temperature information switching unit 3 to switch the path of the temperature sensor group 4 to itself; the heat dissipation controller 2 that takes over the control can then read the real-time temperature and adjust the fan speed precisely and variably, instead of running it at a fixed full speed.

[0037] In summary, this application provides a heat dissipation control device 100. By introducing a temperature information switching unit 3, it achieves intelligent switching of control between the baseboard management controller 1 and the heat dissipation controller 2. This fundamentally avoids the problem of the heat dissipation control device 100 failing to operate normally due to a single point of failure in the baseboard management controller 1, ensuring that the system can perform precise heat dissipation based on real-time temperature data under any condition. Thus, while ensuring the high-temperature reliability of the server, it significantly reduces unnecessary energy consumption and operating noise, and improves the overall system's energy efficiency and availability.

[0038] like Figure 2 and Figure 3 As shown, in some embodiments, the heat dissipation control device 100 further includes a first switch 9. The first switch 9 is connected to the temperature sensor group 4 and also to the temperature information switching unit 3.

[0039] The first switch 9 is configured to transmit the temperature information of the temperature sensor group 4 to the baseboard management controller 1 or the heat dissipation controller 2 via the temperature information switching unit 3.

[0040] The first switch mentioned above usually refers to an I2C switch. The temperature sensor group includes multiple temperature sensors. If they are all directly connected to an I2C bus, it will lead to excessive bus load, address conflicts, and wiring difficulties. The first switch acts as the only intermediate node, connecting all the scattered temperature sensors as slave devices to the downstream ports of the first switch. Regardless of the physical number of sensors, the upstream master control device, board management controller, or thermal controller only needs to communicate with this first switch to access all temperature sensors.

[0041] like Figure 3 As shown, in some embodiments, the temperature information switching unit 3 includes a first multiplexer 31.

[0042] The first data input terminal of the first multiplexer 31 is electrically connected to the communication terminal of the substrate management controller 1; the second data input terminal of the first multiplexer 31 is electrically connected to the communication terminal of the heat dissipation controller 2; the data output terminal of the first multiplexer 31 is electrically connected to the temperature sensor group 4; and the selection control terminal of the first multiplexer 31 is electrically connected to the general input / output terminal of the heat dissipation controller 2.

[0043] The thermal controller 2 is configured to send a switching command to the multiplexer via a general-purpose input / output terminal when the baseboard management controller 1 malfunctions, so as to control the first multiplexer 31 to connect the thermal controller 2 and the temperature sensor group 4.

[0044] For example, a server equipped with multiple high-performance GPUs includes a baseboard management controller 1, four thermal controllers 2, a group of temperature sensors 4 distributed near the GPU and CPU, and a first multiplexer 31.

[0045] When the server is running normally and the baseboard management controller 1 is working properly, the first multiplexer 31 connects the data path of the temperature sensor group 4 to the baseboard management controller 1. The baseboard management controller 1 intelligently adjusts the speed of each fan according to the global temperature. When the server performs a firmware upgrade on the baseboard management controller 1, causing the baseboard management controller 1 to temporarily become inactive, the thermal controller 2 on one of the fan boards detects an abnormality in the watchdog signal of the baseboard management controller 1, and then sends a switching command to the first multiplexer 31 through its general purpose input / output terminal, i.e., the GPIO pin.

[0046] The first multiplexer 31 then switches the data path from the baseboard management controller 1 to the thermal controller 2 that issued the command. This allows the thermal controller 2 to directly read the real-time data from all temperature sensors and calculate the optimal fan speed strategy based on this data. It then notifies other thermal controllers 2 to execute the strategy through a collaborative mechanism. This maintains precise and quiet thermal control throughout the entire upgrade process when the baseboard management controller 1 is unavailable, preventing the GPU from throttling due to overheating.

[0047] The first multiplexer 31 completely solves the problem of blind operation of the heat dissipation system caused by the failure of the baseboard management controller 1 in traditional solutions. Specifically, in the scenario of firmware upgrade of the baseboard management controller 1 (equivalent to the baseboard management controller 1 malfunctioning), the system no longer needs to run all fans at 100% maximum speed continuously to prevent overheating. Instead, it can smoothly and accurately adjust the fan speed according to the actual temperature of core components such as the GPU. This not only reduces the fan noise during the upgrade process from a continuous whistling sound to an acceptable background noise level, but more importantly, it can dynamically adjust the heat dissipation intensity according to the thermal load. While ensuring that the GPU can run at full capacity without overheating and crashing, it avoids the significant additional energy consumption caused by fixed high-speed rotation, achieving a balance between high availability and high energy efficiency.

[0048] In some embodiments, each heat dissipation controller 2 has unique identification information; the heat dissipation component 5 includes a fan.

[0049] Multiple heat dissipation controllers 2 are interconnected via an interconnect bus and are configured to exchange identification information via the interconnect bus when the baseboard management controller 1 malfunctions, and determine a master heat dissipation controller 2 from among the multiple heat dissipation controllers 2 according to preset rules.

[0050] The main heat dissipation controller 2 is used to acquire temperature information and calculate the fan speed control strategy based on the temperature information.

[0051] The thermal controller 2 is also configured to receive and execute fan speed control strategies via the interconnect bus when acting as a slave thermal controller 2.

[0052] In some embodiments, the identification information is the number of the heat dissipation module, and the preset rule is: when it is necessary to select the main heat dissipation controller 2, the heat dissipation controller 2 with the smallest or largest number among all heat dissipation controllers 2 is selected as the main heat dissipation controller 2.

[0053] In one specific embodiment, a server is equipped with four independent fan boards, each board carrying a heat dissipation controller 2, and each board is pre-assigned a unique board number as identification information, such as FAN_ID 1, 2, 3, and 4 respectively; these four heat dissipation controllers 2 are interconnected through a shared I2C bus.

[0054] When the server loses primary cooling control due to a hardware failure of the baseboard management controller 1, the four cooling controllers 2 exchange their respective FAN_IDs via the I2C bus. Based on a preset rule that the smallest number is primary, the cooling controller 2 with FAN_ID 1 is quickly identified as the primary cooling controller 2. This primary cooling controller 2 then obtains the overall system temperature through the temperature information switching unit 3, calculates the required fan speed strategy for each area, and issues the specific speed commands to the slave cooling controllers 2 (numbered 2, 3, and 4) via the same I2C bus. All cooling controllers 2 execute synchronously, thus achieving rapid and orderly takeover of the cooling system in the event of a failure.

[0055] This application fundamentally solves the resource conflict problem that may occur when multiple controllers take over in parallel by adopting a set of deterministic preset rules based on hardware identifiers. This solution ensures that in the instant the baseboard management controller 1 fails, the system can elect a unique master node without conflict within milliseconds and establish an efficient master-slave command system. This allows the heat dissipation takeover process to be carried out in an orderly and reliable manner, just like a single controller, greatly improving the stability and determinism of complex multi-fan board systems in emergency situations.

[0056] like Figure 3 As shown, in some embodiments, the heat dissipation control device 100 further includes a redundant control channel; the redundant control channel is connected between at least two heat dissipation controllers 2.

[0057] The signal output terminal of each heat dissipation controller 2 is electrically connected to the signal input terminal of the other heat dissipation controller 2.

[0058] In other words, the PWM signal output terminal of the thermal controller 2 is cross-connected to a backup signal input terminal on the other fan board through a preset PCB trace (i.e., redundant control channel). When the thermal controller 2 of one fan board completely shuts down due to a sudden component failure and is unable to output a PWM signal, the normal PWM signal generated by the thermal controller 2 of the other fan board is transmitted to the drive circuit of the first fan board through this physical channel, thereby maintaining the continued operation of the first fan and preventing the area it is responsible for from overheating due to the interruption of heat dissipation.

[0059] This physical connectivity-based redundancy mechanism does not rely on any complex software diagnostics or switching protocols, achieving a response speed in the microsecond range and realizing true seamless fault tolerance. It effectively prevents catastrophic consequences caused by a single point of hardware failure leading to complete failure of local heat dissipation and subsequent high-temperature damage to critical server components, elevating system reliability from logical fault tolerance to a new level of hardware physical fault tolerance.

[0060] like Figure 4 As shown, in some embodiments, the heat dissipation control device 100 further includes a plurality of fan plates 8; each fan plate 8 is provided with a buffer 6; a buffer 6 and a heat dissipation controller 2 are disposed on the same fan plate 8.

[0061] The input terminal of buffer 6 is electrically connected to the signal output terminal of other heat dissipation controller 2; the output terminal of buffer 6 is electrically connected to heat dissipation component 5.

[0062] The buffer 6 is used to receive fan control signals generated by other heat dissipation controllers 2 and drive the heat dissipation component 5 through the output terminal.

[0063] In one specific embodiment, a server is configured with four cooling fan boards. After generating a PWM control signal, the cooling controller 2 on each board first inputs it to a local buffer 6. The buffer 6 amplifies the drive signal and directly drives the four fans on the board through the first output terminal. On the other hand, through the second output terminal, it sends an identical copy of the PWM signal to the signal selector 7 on the other three fan boards in real time as a backup via a redundant control channel.

[0064] In the aforementioned server, when fan board number 2 loses control due to a sudden failure of its thermal controller 2, buffers 6 on fan boards numbered 1, 3, and 4 have already continuously provided three healthy backup PWM signals through redundant channels. At this time, the signal selector 7 of board 2 can instantly switch from the failed local signal to a backup signal from, for example, board 1, thereby maintaining the normal operation of its fan and preventing the GPU core area from triggering high-temperature protection due to localized heat dissipation interruption, which could lead to training task failure. This mechanism completely isolates the impact of single-point controller failure, ensuring continuous heat dissipation capability at the hardware level and greatly improving the long-term operational reliability of critical computing services.

[0065] In some embodiments, the heat dissipation control device 100 further includes a plurality of signal selectors 7, at least one of which is disposed on the same fan plate 8 as a heat dissipation controller 2.

[0066] The first input terminal of the signal selector 7 is electrically connected to the signal output terminal of the heat dissipation controller 2, and is used to receive the fan control signal generated by the heat dissipation controller 2.

[0067] The second input terminal of the signal selector 7 is electrically connected to the output terminal of the buffer 6, and is used to receive backup control signals from other heat dissipation controllers 2; the output terminal of the signal selector 7 is electrically connected to the heat dissipation assembly 5.

[0068] The signal selector 7 is configured to connect its output to the first input when the thermal controller 2 is detected to be working properly, so as to output a fan control signal.

[0069] When a malfunction is detected in the heat dissipation controller 2, the output terminal is connected to the second input terminal to output a backup control signal, thereby allowing the heat dissipation component 5 to be controlled by another heat dissipation controller 2.

[0070] For example, in a server configured with six hard drives and three independent cooling zones, each cooling zone's fan board is equipped with a signal selector 7. Its first input is directly connected to the PWM output of the local thermal controller 2, while its second input is cross-connected to the backup signal output of the buffer 6 on the fan boards of the other two zones via an inter-board redundancy channel. When the thermal controller 2 in the middle zone locks up due to a momentary voltage surge and cannot generate a valid PWM signal, the signal selector 7 on this board detects the local signal abnormality within microseconds and automatically switches its output from the first input to the second input connected to the backup signal of the fan board in the left zone. This causes the fans in the middle zone to immediately switch to being controlled by the normally functioning thermal controller 2 on the left, thereby maintaining continuous cooling of the hard drives in that zone.

[0071] This solution utilizes real-time monitoring and automatic switching by signal selector 7 to ensure that the cooling function continues unnoticed by users and upper-layer systems, completely shielding the impact of hardware failures at the physical layer. This not only prevents data loss and hardware damage caused by localized overheating, meeting the stringent business continuity requirements of data centers, but also significantly reduces the operational intervention required for emergency cooling failures and potential business interruption losses.

[0072] In some embodiments, the signal selector 7 described above is a second multiplexer.

[0073] In some embodiments, the heat dissipation controller 2 is further configured to issue a fault alarm signal containing its identification information when the heat dissipation controller 2 switches to use a backup control signal from another heat dissipation controller 2, or when its own state is abnormal.

[0074] In one specific embodiment, a server is equipped with three hot-swappable fan boards, each with a thermal controller 2 that has a unique FAN_ID (e.g., 1, 2, 3). When the thermal controller 2 in the middle position (FAN_ID=2) fails due to an internal logic error, the signal selector 7 on that board automatically switches to receiving the backup PWM signal from the fan board on the left (FAN_ID=1) to maintain fan operation. Simultaneously, the faulty thermal controller 2 on board 2 detects its own abnormal state and has switched to the backup signal, and immediately sends a clear fault alarm signal to the upper-level management software via the system management bus, containing its identifier FAN_ID=2 and an error code.

[0075] For details, please refer to Figure 4 , Figure 4 The buffer in the buffer is used to receive the OTHER_CPLD_PWM signal, which is the signal from other heat dissipation controllers. CPLD_PWM is the signal issued by the main heat dissipation controller, which can also be understood as the signal issued by the heat dissipation controller on a fan board. The signal from the heatsink on other fan boards is OTHER_CPLD_PWM. Figures 2-4 CPLD_NORMAL is a signal indicating whether the thermal controller is malfunctioning; for example, CPLD0_NORMAL is a signal indicating whether the first thermal controller is malfunctioning, and CPLD1_NORMAL is a signal indicating whether the second thermal controller is malfunctioning; CPLD_MASTAR_I2C is a signal that switches the thermal controller from an I2C slave device to an I2C master device; FAN_ID is the fan identification signal; CPLD_MASTAR_PWM is a signal received by the thermal controller when it acts as the master thermal controller.

[0076] In some embodiments, refer to Figure 3The heat dissipation control device 100 also includes a second switch 10.

[0077] The second switch 10 is connected to the baseboard management controller and also to at least one signal selector 7.

[0078] The baseboard management controller can read the current selection status inside any signal selector 7 through the second switch 10; this allows the baseboard management controller to grasp the health status of the entire system in real time and accurately, and know which board has been switched.

[0079] Simultaneously, by combining the FAN_ID and the status of signal selector 7, the baseboard management controller or upper-level management software can generate extremely accurate alarms. For example, for a fan board with FAN_ID=2, its thermal controller has failed, and its fan is currently being driven by the thermal controller with FAN_ID=1 through a redundant channel. This is far superior to simple fan board fault alarms, guiding maintenance personnel to directly replace the correct component.

[0080] This solution reports alarms with a clear FAN_ID, enabling the operation and maintenance monitoring platform to immediately highlight the fault of the No. 2 fan board controller on the topology map. This greatly shortens the fault location time, improves the efficiency of data center operation and maintenance and the long-term reliability of the system, and avoids the risk of causing major failures.

[0081] Embodiments of this application also provide a server including the aforementioned heat dissipation control device 100. This server, through the heat dissipation control device 100, can solve the problems of excessive energy consumption and noise caused by the failure of the baseboard management controller 1, and the potential for conflicts among controllers in multi-fan board scenarios due to a lack of coordination mechanisms. Furthermore, it cannot cope with the complete interruption of localized heat dissipation caused by the hardware failure of a single controller. It ensures that the system can perform precise heat dissipation based on real-time temperature data under any condition, thereby significantly reducing unnecessary energy consumption and operating noise while ensuring the server's high-temperature reliability, and improving the overall system's energy efficiency ratio and availability.

[0082] Embodiments of this application also provide an electronic device, 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 control process of any of the above-described heat dissipation control devices during operation.

[0083] Embodiments of this application also provide a computer-readable storage medium storing a computer program, wherein the computer program is configured to execute the control process of any of the above-described heat dissipation control devices during operation.

[0084] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as a USB flash drive, read-only memory (ROM), random access memory (RAM), portable hard disk, magnetic disk, or optical disk.

[0085] An embodiment of this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the control process for the operation of any of the above-mentioned heat dissipation control devices.

[0086] Embodiments of this application also provide another computer program product, including a non-volatile computer-readable storage medium storing a computer program, which, when executed by a processor, implements the control process of any of the above-described heat dissipation control devices.

[0087] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0088] The above provides a detailed description of a heat dissipation control device, server, electronic device, storage medium, and computer program product provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are only intended to help understand the methods and core ideas of this application. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of the claims of this application.

Claims

1. A heat dissipation control device, characterized in that, include: The baseboard management controller, multiple thermal controllers, temperature information switching unit, temperature sensor group and heat dissipation component; The substrate management controller is connected to the temperature information switching unit and also to the heat dissipation controller; the temperature information switching unit is connected to the temperature sensor group and also to the heat dissipation controller; the heat dissipation controller is electrically connected to the heat dissipation component. When the substrate management controller is working normally, the substrate management controller obtains the temperature information of the temperature sensor group through the temperature information switching unit, and transmits the temperature information to the heat dissipation controller to adjust the heat dissipation component; When the substrate management controller malfunctions, the heat dissipation controller uses the temperature information of the temperature sensor group in the temperature information switching unit to adjust the heat dissipation component according to the temperature information.

2. The heat dissipation control device according to claim 1, characterized in that, The temperature information switching unit includes: a first multiplexer; The first data input terminal of the first multiplexer is electrically connected to the communication terminal of the substrate management controller, and the second data input terminal of the first multiplexer is electrically connected to the communication terminal of the heat dissipation controller; the data output terminal of the first multiplexer is electrically connected to the temperature sensor group; and the selection control terminal of the first multiplexer is electrically connected to the general input / output terminal of the heat dissipation controller. The heat dissipation controller is configured to send a switching command to the multiplexer via the general-purpose input / output terminal when the substrate management controller malfunctions, so as to control the first multiplexer to connect the heat dissipation controller and the temperature sensor group.

3. The heat dissipation control device according to claim 1, characterized in that, Each of the heat dissipation controllers has unique identification information; the heat dissipation component includes: a fan; The multiple heat dissipation controllers are interconnected via an interconnect bus and are configured to exchange identification information via the interconnect bus when the baseboard management controller malfunctions, and determine a master heat dissipation controller from the multiple heat dissipation controllers according to preset rules. The main control heat dissipation controller is used to acquire the temperature information and calculate the fan speed control strategy based on the temperature information; The heat dissipation controller is also configured to receive and execute the fan speed control strategy via the interconnect bus when acting as a slave heat dissipation controller.

4. The heat dissipation control device according to claim 3, characterized in that, The identification information is the number of the heat dissipation module, and the preset rule is: when it is necessary to select the main heat dissipation controller, select the heat dissipation controller with the smallest or largest number among all heat dissipation controllers as the main heat dissipation controller.

5. The heat dissipation control device according to claim 1, characterized in that, It also includes multiple fan plates, each of which is provided with a buffer; one buffer and one heat dissipation controller are disposed on the same fan plate; The input terminal of the buffer is electrically connected to the signal output terminal of the other heat dissipation controllers; the output terminal of the buffer is electrically connected to the heat dissipation assembly. The buffer is used to receive and generate fan control signals from the heat dissipation controller, and drive the heat dissipation component through the output terminal.

6. The heat dissipation control device according to claim 5, characterized in that, It also includes multiple signal selectors; at least one of the signal selectors is mounted on the same fan plate as one of the heat dissipation controllers; The first input terminal of the signal selector is electrically connected to the signal output terminal of the heat dissipation controller, and is used to receive the fan control signal generated by the heat dissipation controller. The second input terminal of the signal selector is electrically connected to the output terminal of the buffer, and is used to receive backup control signals from other heat dissipation controllers; the output terminal of the signal selector is electrically connected to the heat dissipation assembly. The signal selector is configured to connect the output terminal to the first input terminal when the heat dissipation controller is detected to be working properly, so as to output the fan control signal; When a malfunction is detected in the heat dissipation controller, the output terminal is connected to the second input terminal to output the backup control signal, thereby allowing the heat dissipation component to be controlled by another heat dissipation controller.

7. The heat dissipation control device according to claim 6, characterized in that, The signal selector is a second multiplexer.

8. The heat dissipation control device according to claim 3, characterized in that, The heat dissipation controller is also configured to issue a fault alarm signal containing its identification information when the heat dissipation controller switches to use a backup control signal from another heat dissipation controller or when its own state is abnormal.

9. The heat dissipation control device according to claim 1, characterized in that, Also includes: The first switch is connected to the temperature sensor group and also to the temperature information switching unit. The first switch is configured to transmit temperature information from the temperature sensor group to the substrate management controller or heat dissipation controller via the temperature information switching unit.

10. A server, characterized in that, Includes the heat dissipation control device as described in any one of claims 1 to 9.

Images