Computer board card system

Abstract

The invention discloses a computer board card system, and belongs to the technical field of BMC management. Comprising a mainboard circuit, an upper computer and a plurality of BMC management circuits connected with the mainboard circuit and the upper computer. The mainboard circuit comprises a plurality of function modules which are connected with one another, and the plurality of BMC management circuits are respectively in one-to-one correspondence with the plurality of function modules; the function module is used for acquiring state information of the function module based on the driving information of the function module; in response to the adjustment instruction, adjusting the operation mode of the function module; the BMC management circuit is used for acquiring state information of the corresponding function module, analyzing and diagnosing the state information to obtain a diagnosis result and determining an adjusting instruction corresponding to the diagnosis result; receiving configuration information sent by the upper computer and performing configuration; and the upper computer is used for analyzing the state information and the diagnosis results of the plurality of BMC management circuits and generating configuration information of each BMC management circuit. According to the system, the failure rate of the BMC management circuit is greatly reduced, and the stability of a computer board card circuit system is improved.

Description

Technical Field

[0001] The present application belongs to the field of BMC management technology, and in particular, relates to a computer board system. Background Art

[0002] As a key component in modern servers and computing devices, the management and monitoring functions of BMC (Baseboard Management Controller) management circuits have been significantly developed in recent years.

[0003] Modern computer systems need to monitor hardware status in real time. The BMC management circuit needs to process data from multiple sensors, such as temperature, fan speed, and voltage, to ensure that the system runs in the best condition. With the rapid iteration of computer equipment, especially in the development of cloud computing, big data, artificial intelligence and other fields, the requirements for BMC management circuit to process data have increased significantly.

[0004] In related technologies, the processing speed of the BMC management circuit is improved by continuously developing a new generation of processing chips. However, in high-load and high-speed operation scenarios, the chip itself will inevitably have performance bottlenecks, which makes the BMC management circuit prone to failures and low stability when processing complex monitoring data or management requests. Summary of the invention

[0005] The present application aims to solve at least one of the technical problems existing in the prior art. To this end, the present application proposes a computer board system, which greatly reduces the failure rate of the BMC management circuit and improves the stability of the computer board circuit system.

[0006] In a first aspect, the present application provides a computer board system, comprising:

[0007] A mainboard circuit and a host computer, and a plurality of BMC management circuits respectively connected to the board circuit and the host computer;

[0008] The mainboard circuit includes a plurality of interconnected functional modules, and the plurality of BMC management circuits correspond one-to-one to the plurality of functional modules respectively;

[0009] The functional module is used to obtain the state information of the functional module based on the driving information of the functional module; and adjust the operation mode of the functional module in response to the adjustment instruction;

[0010] The BMC management circuit is used to obtain the status information of the corresponding functional module, analyze and diagnose the status information, obtain the diagnosis result, and determine the adjustment instruction corresponding to the diagnosis result; receive the configuration information sent by the host computer and perform configuration;

[0011] The host computer is used to analyze the status information and the diagnosis results of the multiple BMC management circuits to generate the configuration information of each BMC management circuit.

[0012] According to one embodiment of the present application, the host computer is further used to:

[0013] Extracting the current temperature, current voltage, current load and current fault occurrence frequency from the status information of each of the functional modules;

[0014] Analyze and diagnose the current temperature, the current voltage, the current load, and the current fault occurrence frequency respectively to determine a temperature score, a voltage score, a load score, and a historical performance score, wherein the historical performance score is obtained through time decay;

[0015] Obtaining a comprehensive status score of each of the functional modules according to the temperature score, voltage score, load score and historical performance score;

[0016] The configuration information of each of the BMC management circuits is generated according to the comprehensive status score and the diagnosis result.

[0017] According to one embodiment of the present application,

[0018] The comprehensive status score C of the functional module is:

[0019]

[0020] Among them, S min is the minimum nonlinear score, S max is the maximum nonlinear score, K is the scaling factor;

[0021] The nonlinear score S of the functional module is:

[0022]

[0023] Among them, α, β, γ, and δ are the weight coefficients of temperature score T, voltage score V, load score L, and historical performance score H respectively;

[0024] The temperature rating T is:

[0025]

[0026] Among them, P T is the penalty factor of the temperature score T, T current is the current temperature, T normal is the normal temperature, T max is the maximum allowable temperature;

[0027] The voltage rating V is:

[0028]

[0029] Among them, P V is the penalty factor for the voltage score V, V current is the current voltage, V max is the maximum allowable voltage, V min is the minimum allowable voltage, V critical It is a critical voltage;

[0030] The load rating L is:

[0031]

[0032] Among them, P L is the penalty factor of load rating L, L current is the current load, L max is the normal maximum load, L critical For critical loads;

[0033] The historical performance score H is:

[0034]

[0035] Where μ is the time decay factor, t is the time, P H is the penalty factor for the historical performance score H, F is the current failure frequency, and F max is the maximum fault frequency allowed by the functional module.

[0036] According to one embodiment of the present application, the host computer is further used to:

[0037] Extract module status features from the status information of each functional module;

[0038] constructing a state vector of each functional module based on the comprehensive state score, the module state characteristics and the diagnosis result;

[0039] The state vectors are sequentially input into the information prediction model to obtain the configuration information of each of the BMC management circuits output by the information prediction model.

[0040] According to an embodiment of the present application, the information prediction model is obtained by training samples based on time series, the training samples include sample module state features and configuration information tags, and the sample module state features have timestamps.

[0041] According to an embodiment of the present application, the plurality of functional modules include: a computer module, a serial interface module, a reflective memory module, a power module, and a network module, and the processing chip of the BMC management circuit is determined based on the management requirements of the functional modules corresponding to the BMC management circuit;

[0042] The processing chip corresponding to the computer module is an Intel BMC series chip or an ASPEED AST2400 / AST2500 chip;

[0043] The processing chip corresponding to the serial interface module is a Texas Instruments Tiva C series chip or an STMicroelectronics STM32 series chip;

[0044] The processing chip corresponding to the reflective memory module is a Texas Instruments TMS320C67x series chip or a Xilinx Zynq series chip;

[0045] The processing chip corresponding to the power module is NXP LPC series chip, Microchip MCP series chip or Renesas ISL series chip;

[0046] The processing chip corresponding to the network module is the i.MX series chip of NXP / Freescale or the AST series chip of ASPEED.

[0047] According to an embodiment of the present application, the functional module is further used to generate warning information according to the diagnosis result;

[0048] The BMC management circuit is also used for:

[0049] Based on the polling cycle, poll the functional module corresponding to the BMC management circuit, and when it is determined that the functional module generates the alarm information within three consecutive polling cycles, query the alarm information in the event log;

[0050] When the query result is empty, the alarm information is recorded in the event log.

[0051] According to one embodiment of the present application, the BMC management circuit is further used for:

[0052] Based on the state information, determining a current load of the functional module;

[0053] When the load is greater than a preset load threshold, determining the polling cycle to be a first polling cycle;

[0054] When the load is less than or equal to a preset load threshold, the polling cycle is determined to be a second polling cycle, and the second polling cycle is greater than the first polling cycle.

[0055] According to one embodiment of the present application, the BMC management circuit is further used for:

[0056] Based on non-blocking waiting and the polling cycle, an asynchronous function is called to perform concurrent asynchronous polling on the functional module to obtain the status information.

[0057] According to one embodiment of the present application, the functional module is further used to:

[0058] parsing the adjustment instruction to identify the target operation mode of the functional module;

[0059] Execute a status self-check process, and if the self-check passes, configure the operating parameters of the functional module based on the target operating mode to adjust the operating mode;

[0060] The status self-check process is executed, and if the self-check passes, confirmation information is sent to the BMC management circuit.

[0061] Additional aspects and advantages of the present application will be given in part in the description below, and in part will become apparent from the description below, or will be learned through the practice of the present application.

[0062] The computer board system provided by the present invention has the following beneficial effects compared with the prior art:

[0063] (1) By deploying the BMC management circuit in correspondence with different functional modules in the motherboard circuit, monitoring data can be effectively collected and analyzed in the corresponding functional modules to achieve separate management, reduce the processing pressure of the chip in the BMC management circuit, greatly reduce the failure rate of the BMC management circuit, and improve the stability, flexibility and scalability of the computer board circuit system. It can monitor and diagnose the status and performance of each functional module in real time to achieve better management and monitoring. It is suitable for complex system architectures such as data centers or high-performance computing clusters. When a large amount of status information or control tasks need to be processed, multiple BMCs can share the workload and improve the response speed and processing capability. It is suitable for computing devices such as servers and supercomputers that need to process massive amounts of data. The functional modules adjust the operating mode accordingly and can respond quickly to faults. The host computer combines the status information, analyzes the diagnosis results and generates the configuration information of the BMC management circuit, so as to achieve accurate management and rapid response to the hardware status of the motherboard circuit, improve the maintainability and stability of the computer board circuit system, and adapt to the increasingly complex computing environment requirements.

[0064] (2) The system adopts a modular structure, separating the mainboard circuit into different functional modules and configuring a dedicated BMC management circuit separately to facilitate separate testing and maintenance. By selecting the corresponding processing chip according to the management requirements of each functional module, the system can achieve the best balance in terms of performance, flexibility, reliability and cost.

[0065] (3) By using the current load as the basis for whether to shorten the polling cycle, the polling frequency can be increased under high load conditions to more closely monitor the status of functional modules, thereby responding to possible problems more quickly and improving real-time performance. When the load is low, extending the polling cycle can reduce unnecessary monitoring operations, thereby reducing the energy consumption of the system.

[0066] (4) By performing self-checks before and after adjusting the operating mode, the functional module can respond to adjustment instructions efficiently and reliably, and switch to the specified operating mode on the basis of ensuring its own health.

[0067] (5) Consider the impact of time factors on the comprehensive status score C, especially by introducing a time decay factor into the fault frequency and historical performance score, so that the historical performance score decreases over time, reflecting that recent performance is more important. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] The above and / or additional aspects and advantages of the present application will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0069] Figure 1 It is a structural schematic diagram of a computer board system provided in an embodiment of the present application;

[0070] Figure 2 It is a flowchart of the operation of the computer board system provided in the embodiment of the present application;

[0071] Figure 3 It is a schematic diagram of the polling process of the BMC management circuit provided in an embodiment of the present application. DETAILED DESCRIPTION

[0072] The following will be combined with the drawings in the embodiments of the present application to clearly describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments in the present application belong to the scope of protection of this application.

[0073] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of one type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" in the specification and claims represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0074] The computer board system provided in the embodiment of the present application is described in detail below through specific embodiments and their application scenarios in conjunction with the accompanying drawings.

[0075] like Figure 1 As shown, the computer board system (hereinafter referred to as the system) includes: a mainboard circuit and a host computer, and a plurality of BMC management circuits respectively connected to the board circuit and the host computer;

[0076] The mainboard circuit includes a plurality of interconnected functional modules, and the plurality of BMC management circuits correspond one-to-one to the plurality of functional modules respectively;

[0077] The functional module is used to obtain the state information of the functional module based on the driving information of the functional module; and adjust the operation mode of the functional module in response to the adjustment instruction;

[0078] The BMC management circuit is used to obtain the status information of the corresponding functional module, analyze and diagnose the status information, obtain the diagnosis result, and determine the adjustment instruction corresponding to the diagnosis result; receive the configuration information sent by the host computer and perform configuration;

[0079] The host computer is used to analyze the status information and the diagnosis results of the multiple BMC management circuits to generate the configuration information of each BMC management circuit.

[0080] Among them, the BMC management circuit is connected to each functional module through the I2C bus or the LPC bus respectively. The processing chip of the BMC management circuit can be an AST2500 chip. The BMC management circuit includes: a microcontroller (AST2500 chip), a memory, a sensor interface, a communication interface and a user interface, etc., which can realize the control of the cooling fan of the functional module and realize PCIe (Peripheral Component Interconnect Express) data collection.

[0081] It should be noted that during the data transmission and storage process in the system, an error detection and correction (EDC) algorithm is used to ensure the integrity and accuracy of information, improve the reliability of data transmission, and reduce failures caused by data errors.

[0082] The adjustment instructions are used to adjust the working mode of the functional modules to reduce the damage caused by equipment failure and achieve a rapid response to the failure.

[0083] It can be understood that the computer module is responsible for processing data and performing computing tasks, and usually contains core components such as CPU and memory. The CPU can be a QM77 bridge controller. The serial interface module provides a communication interface between the system and external devices, such as RS-232, USB, etc., and is responsible for sending and receiving data. The reflective memory module is used to achieve high-speed data transmission and allows multiple processing units to share memory to improve system performance. The power module provides a stable power supply for the motherboard circuit and all its components to ensure the normal operation of the system.

[0084] Driver information is the data or parameters used by the driver program corresponding to the functional module to configure the functional module to obtain the status information of the functional module. The driver information may include configuration parameters, status reports, error codes, device characteristics, etc. of the functional module.

[0085] Status information is data on the current status of each module, which may include temperature, current, voltage, and signal, etc. Fault information may include the fault type and occurrence time of the system abnormality or fault. The operating mode of the functional module may be the working mode under a specific state, such as normal working mode, fault protection mode, and maintenance mode, etc.

[0086] In actual implementation, Figure 2 As shown, after the system is powered on, the operating system of the mainboard circuit is started, each functional module in the mainboard will be initialized, the corresponding driver will be loaded, and the status of each component will be automatically checked to ensure that all functional modules are working normally and are ready to communicate with the BMC management circuit.

[0087] Each functional module of the mainboard circuit obtains its own temperature, operating voltage, operating load and other status information according to the driving information, and packages the status information and sends it to the BMC management circuit corresponding to each functional module.

[0088] Each BMC management circuit receives the status information of the corresponding functional module, parses the received status information, extracts the key parameters and error logs that affect the working status of the corresponding functional module, and makes abnormal judgments on the key parameters according to the preset range or normal status corresponding to the key parameters, analyzes the error logs to identify faults, obtains diagnostic results, and sends the adjustment instructions corresponding to the diagnostic results to the various functional modules of the mainboard circuit respectively, and sends the diagnostic results and status information to the host computer for summary.

[0089] For example, the key parameters of the status information corresponding to the computer module include voltage, temperature and load, and may also include CPU usage, processor temperature and computer error log; the key parameters of the status information corresponding to the serial interface module include serial port status, data transmission rate, packet loss rate and interface temperature; the key parameters of the status information corresponding to the reflective memory module include memory usage, delay time, data integrity and memory error log; the key parameters of the status information corresponding to the power module include input voltage, output voltage, output current and power supply temperature.

[0090] Among them, the preset ranges of CPU usage and processor temperature are 0%-85% and 0℃-85℃ respectively, the normal state of the serial port status is normal or idle, the preset range of data transmission rate is higher than the minimum rate threshold, the normal state of packet loss rate is 0%, and the preset range of interface temperature is 0℃-85℃.

[0091] In addition, for computer modules, computer error logs are checked to identify potential hardware or software failures; for reflective memory modules, memory error logs are analyzed to identify memory failures, and the combination of the abnormality judgment result and the identified failure is used as the diagnosis result.

[0092] On the one hand, the host computer analyzes the diagnosis results and status information to obtain the configuration information of each BMC management circuit. The host computer sends the configuration information of each BMC management circuit to the corresponding BMC management circuit, and the BMC management circuit configures itself according to the received configuration information.

[0093] On the other hand, each functional module of the mainboard circuit responds to the adjustment instruction and adjusts the operation mode of the functional module.

[0094] It is understandable that each functional module corresponds to a different working mode, which helps each functional module to perform at its best in different usage scenarios while ensuring the security and reliability of the system.

[0095] For example, the computer module includes normal mode, power saving mode and fail-safe mode. In normal mode, the computer executes various tasks of user programs and operating system, and the CPU, memory and other hardware modules are actively working; in power saving mode, the CPU frequency is reduced and some hardware is turned off to reduce energy consumption; in fail-safe mode, the computer module enters safe mode, limiting functions to protect hardware and allow troubleshooting.

[0096] The serial interface module includes data transmission mode, monitoring mode and idle mode. In data transmission mode, the serial interface module sends and receives data in real time through the serial port; in monitoring mode, it monitors the serial port status and traffic, including error detection and data packet monitoring. In idle mode, there is no data transmission, and it maintains a low power consumption state, ready to respond to data requests at any time.

[0097] The reflective memory module includes read-write mode, synchronization mode and fault detection mode. In read-write mode, data can be quickly read and written between memories to support high-performance real-time applications. In synchronization mode, data consistency between different computing nodes is ensured, allowing multiple nodes to share data. In fault detection mode, memory status is monitored, error logs are recorded and fault conditions are reported.

[0098] The power module includes voltage regulation mode, power saving mode and protection mode. In voltage regulation mode, it maintains stable output voltage and current to ensure reliable power supply to the system; in power saving mode, it reduces output power to improve energy efficiency, which is suitable for low load conditions; in protection mode, it automatically enters protection state to prevent damage, which is suitable for detecting abnormal conditions such as overvoltage, overcurrent or overheating.

[0099] According to the computer board system provided by the embodiment of the present application, by deploying the BMC management circuit corresponding to different functional modules in the mainboard circuit, the monitoring data is effectively collected and analyzed in the corresponding functional modules, and separate management is realized, which reduces the processing pressure of the chip in the BMC management circuit, greatly reduces the failure rate of the BMC management circuit, and improves the stability, flexibility and scalability of the computer board circuit system. It can monitor and diagnose faults of the status and performance of each functional module in real time, and achieve better management and monitoring. It is suitable for complex system architectures such as data centers or high-performance computing clusters. When a large amount of status information or control tasks need to be processed, multiple BMCs can share the workload, improve the response speed and processing capability, and are suitable for computing devices such as servers and supercomputers that need to process massive amounts of data. The functional modules adjust the operating mode accordingly and can respond quickly to faults. The host computer analyzes the diagnosis results in combination with the status information and generates the configuration information of the BMC management circuit, so as to achieve accurate management and rapid response to the hardware status of the mainboard circuit, improve the maintainability and stability of the computer board circuit system, and can adapt to the increasingly complex computing environment requirements.

[0100] In some embodiments, the host computer is further used to:

[0101] Extracting the current temperature, current voltage, current load and current fault occurrence frequency from the status information of each of the functional modules;

[0102] Analyze and diagnose the current temperature, the current voltage, the current load, and the current fault occurrence frequency respectively to determine a temperature score, a voltage score, a load score, and a historical performance score, wherein the historical performance score is obtained through time decay;

[0103] Obtaining a comprehensive status score of each of the functional modules according to the temperature score, voltage score, load score and historical performance score;

[0104] The configuration information of each of the BMC management circuits is generated according to the comprehensive status score and the diagnosis result.

[0105] Among them, the historical performance score reflects the operating status of the functional module in the past period of time. The historical performance score is adjusted in time through time decay to avoid over-reliance on outdated data.

[0106] Historical performance scores can be exponentially decayed over time.

[0107] In some embodiments, the comprehensive status score C of the functional module is:

[0108]

[0109] Among them, S min is the minimum nonlinear score, S max is the maximum nonlinear score, K is the scaling factor;

[0110] The nonlinear score S of the functional module is:

[0111]

[0112] Among them, α, β, γ, and δ are the weight coefficients of temperature score T, voltage score V, load score L, and historical performance score H respectively;

[0113] The temperature rating T is:

[0114]

[0115] Among them, P T is the penalty factor of the temperature score T, T current is the current temperature, T normal is the normal temperature, T max is the maximum allowable temperature;

[0116] The voltage rating V is:

[0117]

[0118] Among them, P V is the penalty factor for the voltage score V, V current is the current voltage, V max is the maximum allowable voltage, V min is the minimum allowable voltage, V critical It is a critical voltage;

[0119] The load rating L is:

[0120]

[0121] Among them, P L is the penalty factor of load rating L, L current is the current load, L max is the normal maximum load, L critical For critical loads;

[0122] The historical performance score H is:

[0123]

[0124] Where μ is the time decay factor, t is the time, P H is the penalty factor for the historical performance score H, T is the current failure frequency, and F max is the maximum fault frequency allowed by the functional module.

[0125] It can be understood that the higher the comprehensive status score C is, the better the status of the mainboard circuit is, and the maximum failure rate is used to evaluate the reliability of the functional module under the worst case scenario.

[0126] The weight coefficients α, β, γ, and δ are all no less than 1 and can be adjusted according to the importance of the temperature score T, the voltage score V, the load score L, and the historical performance score H.

[0127] The lower the temperature value, the higher the temperature score T; the lower the voltage fluctuation, the higher the voltage score V; the lower the load value, the higher the load score L; the historical performance score H is based on the historical failure rate or performance stability score.

[0128] By introducing the penalty factor P T , P V , P L , P L , which can reduce the scores caused by faults, errors or non-compliant behaviors, and ensure that the temperature score T, voltage score V, load score L and historical performance score H can reflect potential risks in a timely manner.

[0129] The introduction of Normalize processing and scaling factor K can ensure that the value of the comprehensive status score C is within a specific range and can reflect the relative status of each functional module.

[0130] The influence of time factors on the comprehensive status score C is considered, especially by introducing a time decay factor into the fault frequency and historical performance score, so that the historical performance score decreases over time, reflecting that recent performance is more important.

[0131] For example, the weight coefficient α of the temperature score T is 2.0, the weight coefficient β of the voltage score V is 1.5, the weight coefficient γ of the load score L is 1.0, the weight coefficient δ of the historical performance score H is 2.5, the time decay factor μ is 0.1, and the scaling factor K is 100.

[0132] The higher historical performance score weight δ=2.5 can improve the sensitivity to the frequency of failures, help to detect potential problems in a timely manner, and enhance the fault warning capability.

[0133] The weight coefficient α=2.0 of the temperature score T and the weight coefficient β=1.5 of the voltage score V emphasize the importance of temperature and voltage to system safety, which is especially important for equipment in high temperature and high pressure environments, ensuring early response under dangerous thresholds.

[0134] The historical performance score reflects the operating status of the system over a period of time. It should be adjusted in time to avoid over-reliance on outdated data. The introduction of the time decay factor μ=0.1 ensures that the historical performance score H decays over time, allowing the system to focus more on the current status rather than historical records, thereby adapting to a dynamically changing environment.

[0135] By using a scaling factor K = 100, the score is converted into a more intuitive range, making it easier for users to understand the system status and facilitate subsequent decision-making.

[0136] The use of nonlinear score S makes the improvement of a certain score have a significant impact on the overall score, which can capture the interaction between the scores and improve the accuracy of the score.

[0137] In some embodiments, the host computer is further used to:

[0138] Extract module status features from the status information of each functional module;

[0139] constructing a state vector of each functional module based on the comprehensive state score, the module state characteristics and the diagnosis result;

[0140] The state vectors are sequentially input into the information prediction model to obtain the configuration information of each of the BMC management circuits output by the information prediction model.

[0141] In actual execution, multiple training samples are constructed using historical configuration information and state vectors to train the information prediction model so that the information prediction model can learn the relationship between the configuration information and the state vector of each module.

[0142] In some embodiments, the information prediction model is obtained by training samples based on time series, and the training samples include sample module state features and configuration information tags, and the sample module state features have timestamps.

[0143] The time series is formed by sorting the training samples according to timestamps. According to the time series, the training samples are sequentially input into the information prediction model to train the information prediction model.

[0144] In this embodiment, the introduction of time series features can enhance the adaptability of the model to dynamic changes, effectively capture the time dependency of data, and improve the accuracy and reliability of fault prediction.

[0145] It is understandable that the host computer can also make a comprehensive fault judgment on each functional module based on the status information, reduce false alarms of faults, and more easily identify the root cause of the fault.

[0146] For example, when the CPU usage rate of the computer module is low and the processor temperature is low, but the power supply temperature is high, the upper computer will make a comprehensive judgment based on multiple status information that the power supply module is faulty and needs to increase the fan power of the power supply module and report an error. When the CPU usage rate of the computer module is high and the processor temperature is high, but the power supply temperature is high, the upper computer will make a comprehensive judgment based on multiple status information that the power supply module is not faulty and does not need to report an error. Instead, the upper computer only needs to increase the output power of the fans at the computer module and the power supply module.

[0147] In some embodiments, the processing chip of the BMC management circuit is determined based on the management requirements of the computer module, the serial interface module, the reflective memory module, the power module, and the network module respectively;

[0148] The processing chip corresponding to the computer module is an Intel BMC series chip or an ASPEED AST2400 / AST2500 chip;

[0149] The processing chip corresponding to the serial interface module is a Texas Instruments Tiva C series chip or an STMicroelectronics STM32 series chip;

[0150] The processing chip corresponding to the reflective memory module is a Texas Instruments TMS320C67x series chip or a Xilinx Zynq series chip;

[0151] The processing chip corresponding to the power module is NXP LPC series chip, Microchip MCP series chip or Renesas ISL series chip;

[0152] The processing chip corresponding to the network module is the i.MX series chip of NXP / Freescale or the AST series chip of ASPEED.

[0153] Among them, the serial interface module is an eight-way asynchronous serial port module, and the network module is a dual redundant gigabit network module.

[0154] The computer module, serial interface module, reflective memory module, power module and network module have different requirements for the BMC management circuit, so different processing chips need to be selected.

[0155] The computer module needs to monitor the status of the CPU, memory and other key hardware in real time and handle complex management tasks. The corresponding BMC management circuit processing chip can choose Intel BMC series chips or ASPEED AST2400 / AST2500 chips; the serial interface module requires efficient data transmission and protocol processing and monitors the health status of serial communication. The corresponding BMC management circuit processing chip can choose Texas Instruments Tiva C series chips or STMicroelectronics STM32 series chips; the reflective memory module needs to process high-speed data streams and synchronous operations, requiring low latency and high bandwidth. The corresponding BMC management circuit processing chip can choose Texas InstrumentsTMS320C67x series chips or Xilinx Zynq series chips; the power module needs to monitor the power status, perform power management operations, and ensure the stable operation of the system. The corresponding BMC management circuit processing chip can choose NXP LPC series chips, Microchip MCP series chips or Renesas ISL series chips; the network module needs to process network traffic. According to the complexity of the task, the corresponding BMC management circuit processing chip is selected as NXP / Freescale's i.MX series chips or ASPEED's AST series chips.

[0156] It can be understood that by setting up multiple BMC management circuits to connect the board circuits, a redundant design is implemented to ensure that when one BMC management circuit fails, other BMC management circuits can still manage the board circuits, thereby improving the reliability of the computer board circuit system.

[0157] In this embodiment, the system adopts a modular structure, separates the mainboard circuit into different functional modules, and separately configures a dedicated BMC management circuit to facilitate separate testing and maintenance. By selecting the corresponding processing chip according to the management requirements of each functional module, the system can achieve the best balance in terms of performance, flexibility, reliability and cost.

[0158] In some embodiments, the functional module is further used to generate warning information according to the diagnosis result;

[0159] The BMC management circuit is also used for:

[0160] Based on the polling cycle, poll the functional module corresponding to the BMC management circuit, and when it is determined that the functional module generates the alarm information within three consecutive polling cycles, query the alarm information in the event log;

[0161] When the query result is empty, the alarm information is recorded in the event log.

[0162] like Figure 3 As shown, after the system is powered on, the BMC management circuit periodically polls the alarm information of each functional module through the I2C bus, and receives the polling message of the alarm information to collect the real-time data of the sensor of the functional module. The polling period is adjustable, and the minimum polling period is not less than 5 seconds. When it is judged according to the real-time data of the sensor that the same functional module generates alarm information in three consecutive polling periods, if the alarm information exists in the event log of the BMC management circuit, the BMC management circuit does not process the alarm information. If the alarm information does not exist in the event log of the BMC management circuit, the BMC management circuit records the alarm information in the event log and lights up the fault indicator light through GPIO.

[0163] The BMC management circuit obtains the FRU information of the unit through the I2C bus and determines whether the unit is an air-cooled unit. If the unit is an air-cooled unit, the BMC management circuit obtains the fan speed information of the unit through the PWM wave signal and determines whether there is an alarm information according to the PWM wave signal. If the alarm information exists, the BMC management circuit marks the alarm information, records the alarm information to the event log, marks the fan fault information, and lights the fault indicator light through the GPIO.

[0164] In this embodiment, by querying the alarm information in the event log, it is avoided to repeatedly record the same alarm information in a short period of time, thereby reducing the redundancy of the event log.

[0165] In some embodiments, the BMC management circuit is further used for:

[0166] Based on the state information, determining a current load of the functional module;

[0167] When the load is greater than a preset load threshold, determining the polling cycle to be a first polling cycle;

[0168] When the load is less than or equal to a preset load threshold, the polling cycle is determined to be a second polling cycle, and the second polling cycle is greater than the first polling cycle.

[0169] In this embodiment, the current load is used as the basis for whether to shorten the polling cycle. The polling frequency can be increased under high load conditions to more closely monitor the status of functional modules, thereby responding to possible problems more quickly and improving real-time performance. Extending the polling cycle when the load is low can reduce unnecessary monitoring operations, thereby reducing the energy consumption of the system.

[0170] In some embodiments, the BMC management circuit is further used for:

[0171] Based on non-blocking waiting and the polling cycle, an asynchronous function is called to perform concurrent asynchronous polling on the functional module to obtain the status information.

[0172] In this embodiment, the BMC management circuit can initiate multiple rounds of asynchronous polling to the functional modules without waiting for the response of each functional module, so that resources can be managed more efficiently, waiting time can be reduced, and overall performance can be improved.

[0173] In some embodiments, the functional module is further used to:

[0174] parsing the adjustment instruction to identify the target operation mode of the functional module;

[0175] Execute a status self-check process, and if the self-check passes, configure the operating parameters of the functional module based on the target operating mode to adjust the operating mode;

[0176] The status self-check process is executed, and if the self-check passes, confirmation information is sent to the BMC management circuit.

[0177] In actual execution, the functional module first receives an adjustment instruction including identification information of a target operating mode. The functional module parses the instruction content and extracts parameters of the target operating mode, such as speed, power or other related settings.

[0178] After parsing the instructions, the functional module starts the self-check process to check its own operating status and health. During the self-check process, the module will check the status of key components, including hardware connection, software operation, temperature, power, etc.

[0179] After completing the self-test, the functional module evaluates the self-test results. If all check items pass, it proceeds to the next step; if any check item fails, it returns an error message and stops subsequent operations.

[0180] If the self-test passes, the functional module extracts the required operating parameters according to the target operating mode, and the module adjusts the internal settings and configures the required operating parameters to ensure that it can operate effectively in the target mode.

[0181] After the parameter configuration is completed, execute the status self-check process again to ensure that the functional modules under the new configuration are still in normal operation.

[0182] If the self-test passes, the functional module sends a confirmation message to the BMC management circuit, indicating that it has successfully switched to the target operating mode.

[0183] In this embodiment, by performing self-checks before and after adjusting the operating mode, the functional module can respond to the adjustment instructions efficiently and reliably, and switch to the designated operating mode on the basis of ensuring its own health.

[0184] In the description of this application, "first feature" or "second feature" may include one or more of the features.

[0185] In the description of the present application, “plurality” means two or more.

[0186] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the claims, all of which are within the protection of the present application.

[0187] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0188] Although the embodiments of the present application have been shown and described, those skilled in the art will appreciate that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present application, and that the scope of the present application is defined by the claims and their equivalents.

Claims

1. A computer board system, characterized in that: include: A mainboard circuit and a host computer, and a plurality of BMC management circuits respectively connected to the board circuit and the host computer; The mainboard circuit includes a plurality of interconnected functional modules, and the plurality of BMC management circuits correspond to the plurality of functional modules one by one respectively; The functional module is used to obtain the state information of the functional module based on the driving information of the functional module; and adjust the operation mode of the functional module in response to the adjustment instruction; The BMC management circuit is used to obtain the status information of the corresponding functional module, analyze and diagnose the status information, obtain the diagnosis result, and determine the adjustment instruction corresponding to the diagnosis result; receive the configuration information sent by the host computer and perform configuration; The host computer is used to analyze the status information and the diagnosis results of the multiple BMC management circuits to generate the configuration information of each BMC management circuit.

2. The computer board system according to claim 1, characterized in that: The host computer is further used for: Extracting the current temperature, current voltage, current load and current fault occurrence frequency from the status information of each of the functional modules; Analyze and diagnose the current temperature, the current voltage, the current load, and the current fault occurrence frequency respectively to determine a temperature score, a voltage score, a load score, and a historical performance score, wherein the historical performance score is obtained through time decay; Obtaining a comprehensive status score of each of the functional modules according to the temperature score, voltage score, load score and historical performance score; The configuration information of each of the BMC management circuits is generated according to the comprehensive status score and the diagnosis result.

3. The computer board system according to claim 2, characterized in that: The comprehensive status score C of the functional module is: Among them, S min is the minimum nonlinear score, S max is the maximum nonlinear score, K is the scaling factor; The nonlinear score S of the functional module is: Among them, α, β, γ, and δ are the weight coefficients of temperature score T, voltage score V, load score L, and historical performance score H respectively; The temperature rating T is: Among them, P T is the penalty factor of the temperature score T, T current is the current temperature, T normal is the normal temperature, T max is the maximum allowable temperature; The voltage rating V is: Among them, P V is the penalty factor for the voltage score V, V current is the current voltage, V max is the maximum allowable voltage, V min is the minimum allowable voltage, V critical It is a critical voltage; The load rating L is: Among them, P L is the penalty factor of load rating L, L current is the current load, L max is the normal maximum load, L critical For critical loads; The historical performance score H is: Where μ is the time decay factor, t is the time, P H is the penalty factor for the historical performance score H, F is the current fault frequency, and F max is the maximum fault frequency allowed by the functional module.

4. The computer board system according to claim 2, characterized in that: The host computer is further used for: Extract module status features from the status information of each functional module; constructing a state vector of each functional module based on the comprehensive state score, the module state characteristics and the diagnosis result; The state vectors are sequentially input into the information prediction model to obtain the configuration information of each of the BMC management circuits output by the information prediction model.

5. The computer board system according to claim 4, characterized in that: The information prediction model is obtained by training samples based on time series, wherein the training samples include sample module state features and configuration information labels, and the sample module state features are provided with timestamps.

6. The computer board system according to claim 1, characterized in that: The multiple functional modules include: a computer module, a serial interface module, a reflective memory module, a power module and a network module, and the processing chip of the BMC management circuit is determined based on the management requirements of the functional modules corresponding to the BMC management circuit; The processing chip corresponding to the computer module is an Intel BMC series chip or an ASPEED AST2400 / AST2500 chip; The processing chip corresponding to the serial interface module is a Texas Instruments Tiva C series chip or an STMicroelectronics STM32 series chip; The processing chip corresponding to the reflective memory module is a Texas Instruments TMS320C67x series chip or a Xilinx Zynq series chip; The processing chip corresponding to the power module is NXP LPC series chip, Microchip MCP series chip or RenesasISL series chip; The processing chip corresponding to the network module is the i.MX series chip of NXP / Freescale or the AST series chip of ASPEED.

7. The computer board system according to claim 1, characterized in that: The functional module is further used to generate warning information according to the diagnosis result; The BMC management circuit is also used for: Based on the polling cycle, poll the functional module corresponding to the BMC management circuit, and when it is determined that the functional module generates the alarm information within three consecutive polling cycles, query the alarm information in the event log; When the query result is empty, the alarm information is recorded in the event log.

8. The computer board system according to claim 7, characterized in that: The BMC management circuit is also used for: Based on the state information, determining a current load of the functional module; When the load is greater than a preset load threshold, determining the polling cycle to be a first polling cycle; When the load is less than or equal to a preset load threshold, the polling cycle is determined to be a second polling cycle, and the second polling cycle is greater than the first polling cycle.

9. The computer board system according to claim 7, characterized in that: The BMC management circuit is also used for: Based on non-blocking waiting and the polling cycle, an asynchronous function is called to perform concurrent asynchronous polling on the functional module to obtain the status information.

10. The computer board system according to claim 1, characterized in that: The functional module is further used for: parsing the adjustment instruction to identify the target operation mode of the functional module; Execute a status self-check process, and if the self-check passes, configure the operating parameters of the functional module based on the target operating mode to adjust the operating mode; The status self-check process is executed, and if the self-check passes, confirmation information is sent to the BMC management circuit.

Images