A server safety power on / off method and system

Abstract

The present invention discloses a method and system for safely powering on and off a server. The method utilizes a baseboard management controller (BMC) to detect the status of the server's operating system, making startup decisions based on this information. Before startup, the health of the server's hardware is checked to determine whether the hardware is in an abnormal fault state, ensuring that the server is powered on and powered on safely and reliably. The BMC controller verifies the shutdown process status with the operating system, making shutdown decisions based on the verification results. The BMC repeatedly confirms the shutdown process status with the system OS, ensuring server service reliability.

Description

Technical Field

[0001] The present invention relates to the technical field of computer firmware, and in particular to a method and system for safely powering on and off a server. Background Art

[0002] The following problems exist when performing remote power management on existing servers:

[0003] 1) When using remote power on and off, existing servers directly use a complex programmable logic device (CPLD) to perform direct current (DC) power on and off operations, which poses the risk of malfunction or repeated operation.

[0004] 2) When the server hardware equipment is abnormal, directly turning on the computer may cause overheating and damage to important components such as the Central Processing Unit (CPU) and Graphics Processing Unit (GPU).

[0005] 3) When the server operating system (OS) is running, if you directly shut down the server remotely to power off the DC, if the system is reading or writing disk operations, it may cause system crashes, damage or loss of files on the disk, and other problems, which is not conducive to reliable system operation. Summary of the Invention

[0006] In view of this, the present invention provides a method for safely powering on and off a server, which can safely and effectively manage the power supply of the server and improve the safety of powering on and off the server.

[0007] The specific technical solutions adopted in the present invention are as follows:

[0008] A server safe power on / off method, comprising:

[0009] Using the baseboard management controller (BMC) to detect the status information of the server's operating system and make a startup decision based on the status information;

[0010] The shutdown process status is verified by the BMC controller and the operating system, and a shutdown decision is made according to the verification result.

[0011] Furthermore, the status information includes at least: server power on or off status information, hardware device status information;

[0012] The hardware device status information includes at least: backup power supply BBU access information, central processing unit CPU temperature information, and fan redundancy information.

[0013] Furthermore, making a power-on decision based on the status information includes:

[0014] A1. The BMC controller detects the server power status. If it is off, it detects the hardware status. If it is on, it ends the process.

[0015] A2. When the server power supply is a backup power supply BBU, the BMC controller detects whether the BBU power line is connected normally. If normal, it continues to detect the hardware status; if abnormal, the process ends;

[0016] A3. The BMC controller sequentially detects the CPU temperature and fan redundancy. If both the CPU temperature information and the fan redundancy information meet the requirements, the complex programmable logic device (CPLD) is used to control the power supply to power on, thereby completing the safe startup of the server.

[0017] Furthermore, the shutdown process status verification is performed with the operating system through the BMC controller, and a shutdown decision is made based on the verification result, including:

[0018] B1, the BMC controller receives a shutdown request instruction, stores a shutdown event in an event message buffer, and sets an interrupt event EVT_ATN on the bitstream protocol interface BT;

[0019] B2. The operating system monitors and receives the interrupt event EVT_ATN through the intelligent platform management interface IPMI;

[0020] B3, the operating system sends an interrupt signal corresponding to the interrupt event EVT_ATN to the IPMI toolkit, and sends an acquisition message identifier to the BMC controller through the IPMI toolkit;

[0021] B4. The BMC controller responds to the acquisition message identifier and sends the acquisition event message identifier to the IPMI toolkit;

[0022] B5. The IPMI toolkit of the operating system sends an instruction to read the event message buffer to the BMC controller;

[0023] B6. The BMC controller sends the shutdown event in the event message buffer to the IPMI toolkit;

[0024] B7, the IPMI toolkit parses and verifies the shutdown event, first stops the service process, and then sends a forced power-off command to the BMC controller;

[0025] B8. The BMC controller performs a power-off operation through the CPLD according to the forced power-off command to complete a safe shutdown of the server.

[0026] A server safety power on / off system, comprising: an operating subsystem and a baseboard management controller (BMC) subsystem; the operating subsystem and the BMC subsystem are arranged on the server;

[0027] Detecting status information of the operating subsystem through the BMC subsystem, and making a startup decision based on the status information;

[0028] The shutdown process status is verified by the BMC subsystem and the operation subsystem, and a shutdown decision is made according to the verification result.

[0029] Furthermore, the status information includes at least: server power on or off status information, hardware device status information;

[0030] The hardware device status information includes at least: backup power supply BBU access information, central processing unit CPU temperature information, and fan redundancy information.

[0031] Furthermore, the operating subsystem includes: an application module for monitoring message events of the BMC subsystem; an IPMI driver module for implementing the Intelligent Platform Management Interface IPMI protocol specification; and an IPMI communication module for communicating with the BMC subsystem.

[0032] The BMC subsystem includes: a webpage module for providing a human-machine management interface; an IPMI module for providing a server-side response for the operating subsystem; a low-pin-count bus LPC module for providing physical information interaction between the BMC subsystem and the CPU processor; and a general-purpose input / output interface GPIO module for powering on and off the operating subsystem through a complex programmable logic device CPLD.

[0033] Furthermore, the BMC subsystem detects status information of the operating subsystem and makes a startup decision based on the status information, including:

[0034] A1. The BMC subsystem detects the power status of the operating subsystem. If it is off, the hardware status is detected; if it is on, the process ends.

[0035] A2. When the power supply of the operating subsystem is the backup power supply BBU, the BMC subsystem detects whether the BBU power line is connected normally. If normal, it continues to detect the hardware status; if abnormal, it ends the process;

[0036] A3. The BMC subsystem sequentially detects the CPU temperature and fan redundancy of the operating subsystem. If the CPU temperature information and fan redundancy information meet the requirements, the complex programmable logic device (CPLD) is used to control the power supply to power on, thereby completing the safe startup of the server.

[0037] Furthermore, the BMC subsystem and the operation subsystem perform shutdown process status verification, and make a shutdown decision based on the verification result, including:

[0038] B1. The BMC subsystem receives a shutdown request instruction, stores a shutdown event in an event message buffer, and sets an interrupt event EVT_ATN on the bitstream protocol interface BT;

[0039] B2. The operating subsystem monitors and receives the interrupt event EVT_ATN through the IPMI driver module;

[0040] B3, the operating subsystem sends an interrupt signal corresponding to the interrupt event EVT_ATN to the IPMI communication module, and sends an acquisition message identifier to the BMC subsystem through the IPMI communication module;

[0041] B4. The IPMI module of the BMC subsystem responds to the acquisition message identifier and sends the acquisition event message identifier to the IPMI communication module;

[0042] B5. The IPMI communication module of the operating subsystem sends an instruction to read the event message buffer to the BMC subsystem;

[0043] B6. The BMC subsystem sends the shutdown event in the event message buffer to the IPMI communication module;

[0044] B7, the IPMI communication module parses and verifies the shutdown event, first stops the service process, and then sends a forced power-off command to the BMC subsystem;

[0045] B8. The BMC subsystem performs a power-off operation on the GPIO module through the CPLD according to the forced power-off command, thereby completing a safe shutdown of the server.

[0046] Beneficial effects:

[0047] (1) A method for safely powering on and off a server uses a baseboard management controller (BMC) to detect the status of the server's operating system, making a power-on decision based on the status information. Before powering on, the health status of the server's hardware is detected to determine whether the hardware is in an abnormal fault state, ensuring that the server is powered on and powered on safely and reliably. The shutdown process status is verified between the BMC controller and the operating system, and a shutdown decision is made based on the verification result. The shutdown process status is repeatedly confirmed between the BMC and the system OS, ensuring the reliability of the server's services.

[0048] (2) The BMC subsystem detects whether the BBU power cable is connected normally. If normal, it continues to detect the hardware status; if not, it ends the process, which can prevent the BBU power supply from starting up in battery standby mode, causing the BBU battery power to be consumed quickly.

[0049] (3) The BMC subsystem detects the CPU temperature and fan redundancy of the operating subsystem in turn. When the CPU temperature information and fan redundancy information meet the requirements, the complex programmable logic device (CPLD) is used to control the power supply to prevent the CPU from overheating again and causing damage to the CPU chip. It also prevents the server from starting up when the number of cooling fans is insufficient, thereby avoiding the occurrence of internal temperature rise failures in the server.

[0050] (4) The IPMI communication module parses and verifies the shutdown event, stops the service process first, and then sends a forced power-off command to the BMC subsystem. When the BMC subsystem executes the shutdown, it can interact with the application layer of the operating subsystem to exchange shutdown event message information, thus realizing the shutdown of system services before shutting down and powering off, ensuring the reliability of application layer service execution. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 This is a structural block diagram of the server safety power on / off system of the present invention;

[0052] Figure 2 This is a flow chart of a server secure startup method of the present invention;

[0053] Figure 3 This is a flow chart of a server safe shutdown method of the present invention. DETAILED DESCRIPTION

[0054] The present invention provides a method and system for safely powering on and off a server. This method uses a baseboard management controller (BMC) to detect the status of the server's operating system and make startup decisions based on this information. Before startup, the health of the server's hardware is checked to determine whether the hardware is in an abnormal fault state, ensuring that the server is powered on and powered on safely and reliably. The BMC controller verifies the shutdown process status with the operating system, making shutdown decisions based on the verification results. This process repeatedly confirms the shutdown process status with the operating system, ensuring server service reliability.

[0055] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0056] The Baseboard Management Controller (BMC) and Intelligent Platform Management Interface (IPMI) are the basic core functional subsystems of the server, responsible for core functions such as hardware status management, operating system management, health status management, and power management.

[0057] A BMC is a small operating system independent of the server system and is a chip integrated into the motherboard. Server clusters typically use BMC commands or perform large-scale unattended operations, including remote management, monitoring, installation, and power cycling of servers.

[0058] An embodiment of the present invention provides a method for safely powering on and off a server, comprising:

[0059] Step 1: Use the baseboard management controller (BMC) to detect the status information of the server's operating system and make a startup decision based on the status information;

[0060] In a specific embodiment, the status information includes at least: server power on or off status information, hardware device status information; wherein, the hardware device status information includes at least: backup power supply (Battery Backup Unit, BBU) access information, central processing unit CPU temperature information, and fan redundancy information.

[0061] Step 2: Verify the shutdown process status through the BMC controller and the operating system, and make a shutdown decision based on the verification result.

[0062] In a specific embodiment, in step 1, making a power-on decision based on the status information includes:

[0063] A1. The BMC controller checks the server power status. If it is powered off, it checks the hardware status. If it is powered on, the process ends.

[0064] A2. When the server power supply is the backup power supply BBU, the BMC controller checks whether the BBU power cable is connected normally. If normal, it continues to check the hardware status. If not, the process ends.

[0065] A3. The BMC controller sequentially detects the CPU temperature and fan redundancy. If both the CPU temperature information and the fan redundancy information meet the requirements, the complex programmable logic device (CPLD) controls the power supply to power on, completing the safe startup of the server.

[0066] In a specific embodiment, in step 2, the shutdown process status is verified by the BMC controller and the operating system, and a shutdown decision is made based on the verification result, including:

[0067] B1. The BMC controller receives the shutdown request instruction, stores the shutdown event in the event message buffer, and sets the interrupt event EVT_ATN on the Bit Torrent protocol interface (BT);

[0068] B2. The operating system monitors and receives the interrupt event EVT_ATN through the intelligent platform management interface IPMI;

[0069] B3. The operating system sends an interrupt signal corresponding to the interrupt event EVT_ATN to the IPMI toolkit, and sends an acquisition message identifier to the BMC controller through the IPMI toolkit;

[0070] B4. The BMC controller responds to the acquisition message identifier and sends the acquisition event message identifier to the IPMI toolkit;

[0071] B5. The IPMI toolkit of the operating system sends a command to read the event message buffer to the BMC controller;

[0072] B6. The BMC controller sends the shutdown event in the event message buffer to the IPMI toolkit;

[0073] B7. The IPMI toolkit parses and verifies the shutdown event, first stopping the service process, and then sending a forced power-off command to the BMC controller;

[0074] B8. The BMC controller performs a power-off operation through the CPLD according to the forced power-off command, completing a safe shutdown of the server.

[0075] An embodiment of the present invention also provides a server safety power on / off system, including: an operating subsystem and a baseboard management controller (BMC) subsystem; the operating subsystem and the BMC subsystem are arranged on a server; the BMC subsystem detects status information of the operating subsystem and makes a power on decision based on the status information; the BMC subsystem and the operating subsystem verify the shutdown process status and make a shutdown decision based on the verification result.

[0076] In a specific embodiment, the status information includes at least: server power on or off status information, hardware device status information;

[0077] The hardware device status information includes at least: backup power supply BBU access information, central processing unit CPU temperature information, and fan redundancy information.

[0078] In a specific embodiment, the operating subsystem includes: an application module for monitoring message events of the BMC subsystem; an IPMI driver module for implementing the Intelligent Platform Management Interface IPMI protocol specification; an IPMI communication module for communicating with the BMC subsystem;

[0079] The BMC subsystem includes: a webpage module for providing a human-machine management interface; an IPMI module for providing server-side responses to the operating subsystem; a low pin count (LPC) bus module for providing physical information exchange between the BMC subsystem and the CPU processor; and a general purpose input / output (GPIO) module for powering on and off the operating subsystem via a complex programmable logic device (CPLD).

[0080] In a specific embodiment, the BMC subsystem detects status information of the operating subsystem and makes a startup decision based on the status information, including:

[0081] A1. The BMC subsystem checks the power status of the operating subsystem. If it is powered off, the hardware status is checked. If it is powered on, the process ends.

[0082] A2. When the operating subsystem is powered by the backup power supply BBU, the BMC subsystem checks whether the BBU power cable is connected properly. If so, it continues to check the hardware status. If not, the process ends.

[0083] A3. The BMC subsystem sequentially detects the CPU temperature and fan redundancy of the operating subsystem. If the CPU temperature information and fan redundancy information meet the requirements, the complex programmable logic device (CPLD) controls the power supply to power on, completing the safe startup of the server.

[0084] In a specific embodiment, the BMC subsystem and the operation subsystem verify the shutdown process status and make a shutdown decision based on the verification result, including:

[0085] B1. The BMC subsystem receives the shutdown request instruction, stores the shutdown event in the event message buffer, and sets the interrupt event EVT_ATN on the bitstream protocol interface BT;

[0086] B2. The operating subsystem monitors and receives the interrupt event EVT_ATN through the IPMI driver module;

[0087] B3. The operating subsystem sends an interrupt signal corresponding to the interrupt event EVT_ATN to the IPMI communication module, and sends an acquisition message identifier to the BMC subsystem through the IPMI communication module;

[0088] B4. The IPMI module of the BMC subsystem responds to the acquisition message identifier and sends the acquisition event message identifier to the IPMI communication module;

[0089] B5. The IPMI communication module of the operating subsystem sends a command to read the event message buffer to the BMC subsystem;

[0090] B6. The BMC subsystem sends the shutdown event in the event message buffer to the IPMI communication module;

[0091] B7. The IPMI communication module parses and verifies the shutdown event, first stops the service process, and then sends a forced power-off command to the BMC subsystem;

[0092] B8. The BMC subsystem responds to the forced power-off command, and the GPIO module performs the power-off operation through the CPLD, completing the safe shutdown of the server.

[0093] In order to enable those skilled in the art to better understand the technical solution of the present invention, it is described below in conjunction with specific scenario embodiments.

[0094] Scenario Example 1

[0095] In this scenario embodiment, the system and method provided by the present invention are introduced in detail based on the specific system framework figure and the power on / off method flow chart.

[0096] Figure 1 This is a structural diagram of the server safety power on / off system of the present invention. Figure 1As shown, at the hardware level, the hardware layer mainly provides the physical interface for communication between the server subsystem modules. Its physical interface includes the CPU processor responsible for executing program tasks, the PC power supply unit (PSU) responsible for providing server power input, the CPLD device responsible for processing hardware logic, and the GPIO / I2C interface responsible for transmitting level signals or data between the BMC and the CPLD (i.e., the GPIO module in the above embodiment). The operating system layer provides an application program that monitors BMC message events under the OS, and the kernel is used to implement the IPMI driver of the IPMI protocol specification. The communication layer mainly includes the ipmiutil communication module (i.e., the IPMI communication module in the above embodiment) that monitors and parses IPMI messages, the WEB module that provides the user human-machine interface management unit, the IPMI module that provides the IPMI server response, the Simple Network Management Protocol (SNMP) module for out-of-band control, the network module that provides the Ethernet interface, and the LPC module that provides the physical interface for information interaction between the BMC and the CPU.

[0097] The operating system implements information interaction with the BMC through the IPMI driver; through the LPC communication bus, the BMC can send the Evt_ATN interrupt signal of the BT interface to the operating system OS. After receiving the signal, the operating system OS triggers ipmiutil to send the Get Message Flags message using the IPMI standard protocol.

[0098] The BMC system responds to the server's Get Message Flags message through the IPMI module. When a safe shutdown is initiated through IPMI or the web, the BMC sets the Message Flags flag. The operating system then checks the Get Message Flags flag and uses the ipmiutil communication module to send a Read Event Message message using the standard IPMI protocol to retrieve detailed event information. The BMC then returns the safe shutdown event message through the IPMI module. If the operating system interprets the event message as a shutdown command, it enters the HALT state to shut down the system and then sends a power-off command to the BMC through ipmiutil.

[0099] Figure 2 This is a flow chart of a server security startup method of the present invention, such as Figure 2 As shown, the process includes the following steps:

[0100] Step 1: Safely power on the server. This can be done by sending an IPMI power-on command to the BMC or clicking the power-on button on the BMC web page.

[0101] Step 2: The BMC checks the current server power status. If it is powered off, it continues the process to check the hardware status. If it detects that the current status is powered on, it ignores the power on request and ends the process to prevent malfunction of power on.

[0102] Step 3: If the server power supply is the backup power supply BBU power supply, check whether the BBU power line is connected normally. If the AC power (Alternating Current, AC) and DC power lines are connected normally, continue the process to check the hardware status; if the BBU power supply is not connected to the AC or DC power line, ignore the power-on request to prevent the BBU power supply from starting up in battery standby mode, causing rapid consumption of the BBU battery power.

[0103] Step 4: Check whether the internal temperature of the server CPU is overheated. If not, continue the process to check the hardware status. If the CPU is in an overheated state, ignore the power-on request to prevent the CPU from overheating again and causing damage to the CPU chip.

[0104] Step 5: Check whether the fan redundancy is normal. If the number of normal cooling fans is insufficient, ignore the power-on request to prevent the server from starting when the number of cooling fans is insufficient, thereby avoiding the occurrence of internal temperature rise failures in the server.

[0105] Step 6: If all the above tests are passed, the CPLD is controlled to perform DC power-on.

[0106] Step 7: The process ends.

[0107] Figure 3 This is a flow chart of a server safety shutdown method of the present invention, such as Figure 3 As shown, the process includes the following steps:

[0108] Step 1: Safely shut down the server. You can send an IPMI shutdown command through the BMC or click the Safe Shutdown button on the BMC web page.

[0109] Step 2: After receiving the request, the BMC puts the shutdown event into the Event Message Buffer, and sets the EVT_ATN interrupt event on the BT communication interface through the Low Pin Count (LPC) bus.

[0110] Step 3: The IPMI driver of the server OS system uses the interrupt mode to monitor the BT interrupt.

[0111] Step 4: After receiving the EVT_ATN interrupt from the BT channel, the server OS passes the IPMI interrupt signal to the ipmiutil service. After receiving the interrupt, ipmiutil sends the Get Message Flags ipmi command to the BMC.

[0112] Step 5: The IPMI module of the BMC responds to the Get Message Flags command and replies that an interrupt event has occurred.

[0113] Step 6: After receiving the Get Event Message Flags status, the ipmiutil service continues to send the Read Event Message Buffer command to query detailed event information.

[0114] Step 7: The BMC responds to the shutdown event in the Event Message Buffer to the ipmiutil service.

[0115] Step 8: After receiving the safe boot message, the ipmiutil service sets an asynchronous signal to notify the application layer to process the shutdown message.

[0116] Step 9: The ipmiutil service parses the message event. If it is a shutdown event, it sets the OS system to the "HALT" state, stops the running service processes, and sends a forced power-off command to the BMC after all service processes are stopped.

[0117] Step 10: The BMC receives the power-off command and controls the CPLD to power off the DC.

[0118] Step 11: End the process.

[0119] The solution of the present invention can extract BMC web overview page information after a BMC cold reboot, including: host name, server information, BMC information; web-based WEB hardware status page information: hard disk information; extract laninfo (BMC IP information), mc info (BMC firmware version information), sensor (chassis sensor information), sel (system log information), bios ver info (BIOS firmware version information), and device GUID (server globally unique identifier information) at the operating system level, and detect the connectivity between the BMC and the Basic Input Output System (BIOS) after a cold reboot, ensuring the consistency of test information between the BMC and BIOS. The username and password for BMC and operating system logins can be configured, and switching between multiple environments is possible. It has high concurrency and supports the simultaneous launch of multiple environment stress tests.

[0120] At the method level, the present invention's safe power on / off method checks the server hardware health status before powering on to determine whether the hardware is in an abnormal fault state, ensuring that the server is powered on and powered on safely and reliably. During a soft shutdown, the present invention's safe power on / off method repeatedly confirms the shutdown process status with the BMC and the system OS, ensuring that the system OS is shut down first before powering off, thus ensuring server service reliability.

[0121] At the system level, module communication and streamlined response processes ensure the reliability of power-on and power-off events. The BMC performs a hardware self-check before the server boots up, improving security compared to traditional direct power-on methods. When executing a shutdown, the BMC interacts with the system's OS application layer to communicate shutdown event information, enabling system services to be shut down before powering off.

[0122] The above specific embodiments merely illustrate the design principles of the present invention. The shapes and names of the components described herein may vary and are not limiting. Therefore, those skilled in the art may modify or substitute equivalents for the technical solutions described in the above embodiments. Such modifications and substitutions, without departing from the inventive spirit and technical solutions of the present invention, shall fall within the scope of protection of the present invention.

Claims

1. A server safe power on / off method, characterized in that: include: Using the baseboard management controller (BMC) to detect the status information of the server's operating system and make a startup decision based on the status information; Verifying the shutdown process status through the BMC controller and the operating system, and making a shutdown decision based on the verification result, including: B1, the BMC controller receives a shutdown request instruction, stores a shutdown event in an event message buffer, and sets an interrupt event EVT_ATN on the bitstream protocol interface BT; B2. The operating system monitors and receives the interrupt event EVT_ATN through the intelligent platform management interface IPMI; B3, the operating system sends an interrupt signal corresponding to the interrupt event EVT_ATN to the IPMI toolkit, and sends an acquisition message identifier to the BMC controller through the IPMI toolkit; B4. The BMC controller responds to the acquisition message identifier and sends the acquisition event message identifier to the IPMI toolkit; B5. The IPMI toolkit of the operating system sends an instruction to read the event message buffer to the BMC controller; B6. The BMC controller sends the shutdown event in the event message buffer to the IPMI toolkit; B7, the IPMI toolkit parses and verifies the shutdown event, first stops the service process, and then sends a forced power-off command to the BMC controller; B8. The BMC controller performs a power-off operation through the CPLD according to the forced power-off command to complete a safe shutdown of the server.

2. The method according to claim 1, wherein The status information includes at least: server power on or off status information, hardware device status information; The hardware device status information includes at least: backup power supply BBU access information, central processing unit CPU temperature information, and fan redundancy information.

3. The method according to claim 1, wherein The step of making a power-on decision based on the status information includes: A1. The BMC controller detects the server power status. If it is off, it detects the hardware status. If it is on, it ends the process. A2. When the server power supply is a backup power supply BBU, the BMC controller detects whether the BBU power line is connected normally. If normal, it continues to detect the hardware status; if abnormal, the process ends; A3. The BMC controller sequentially detects the CPU temperature and fan redundancy. If both the CPU temperature information and the fan redundancy information meet the requirements, the complex programmable logic device (CPLD) is used to control the power supply to power on, thereby completing the safe startup of the server.

4. A server safety power on / off system, characterized in that: include: Operation subsystem and baseboard management controller BMC subsystem; The operating subsystem and the BMC subsystem are arranged on a server; Detecting status information of the operating subsystem through the BMC subsystem, and making a startup decision based on the status information; The shutdown process status is verified by the BMC subsystem and the operation subsystem, and a shutdown decision is made according to the verification result, specifically including: B1. The BMC subsystem receives a shutdown request instruction, stores a shutdown event in an event message buffer, and sets an interrupt event EVT_ATN on the bitstream protocol interface BT; B2. The operating subsystem monitors and receives the interrupt event EVT_ATN through the IPMI driver module; B3, the operating subsystem sends an interrupt signal corresponding to the interrupt event EVT_ATN to the IPMI communication module, and sends an acquisition message identifier to the BMC subsystem through the IPMI communication module; B4. The IPMI module of the BMC subsystem responds to the acquisition message identifier and sends the acquisition event message identifier to the IPMI communication module; B5. The IPMI communication module of the operating subsystem sends an instruction to read the event message buffer to the BMC subsystem; B6. The BMC subsystem sends the shutdown event in the event message buffer to the IPMI communication module; B7, the IPMI communication module parses and verifies the shutdown event, first stops the service process, and then sends a forced power-off command to the BMC subsystem; B8. The BMC subsystem performs a power-off operation on the GPIO module through the CPLD according to the forced power-off command, thereby completing a safe shutdown of the server.

5. The system according to claim 4, wherein: The status information includes at least: server power on or off status information, hardware device status information; The hardware device status information includes at least: backup power supply BBU access information, central processing unit CPU temperature information, and fan redundancy information.

6. The system according to claim 4, wherein: The operating subsystem includes: an application module for monitoring message events of the BMC subsystem; an IPMI driver module for implementing the Intelligent Platform Management Interface IPMI protocol specification; and an IPMI communication module for communicating with the BMC subsystem. The BMC subsystem includes: a webpage module for providing a human-machine management interface; an IPMI module for providing a server-side response for the operating subsystem; a low-pin-count bus LPC module for providing physical information interaction between the BMC subsystem and the CPU processor; and a general-purpose input / output interface GPIO module for powering on and off the operating subsystem through a complex programmable logic device CPLD.

7. The system according to claim 4, wherein: The BMC subsystem detects status information of the operating subsystem and makes a startup decision based on the status information, including: A1. The BMC subsystem detects the power status of the operating subsystem. If it is off, the hardware status is detected; if it is on, the process ends. A2. When the power supply of the operating subsystem is the backup power supply BBU, the BMC subsystem detects whether the BBU power line is connected normally. If normal, it continues to detect the hardware status; if abnormal, it ends the process; A3. The BMC subsystem sequentially detects the CPU temperature and fan redundancy of the operating subsystem. If the CPU temperature information and fan redundancy information meet the requirements, the complex programmable logic device (CPLD) is used to control the power supply to power on, thereby completing the safe startup of the server.

Images