A method for redundant backup of chassis main control management function

Abstract

This invention discloses a method for redundant backup of the main control management function of a chassis, comprising the following steps: S1, in the chassis, the IPMC components in each module are interconnected via the chassis IPMB bus; S2, the IPMCs in the main control module and other processor-equipped service modules are connected to their respective processors via a serial port, forming independent chassis management controllers; S3, the WEB management interface interacts with the chassis management controller; S4, after the chassis is powered on, the main control module switches to host working mode, and the other service modules act as backup modules, maintaining slave working mode; S5, when the main control module malfunctions, causing abnormal management function, the IPMC in the main control module determines the communication interruption via the serial port, switches to slave working mode, and switches the backup module to host working mode. This invention requires no external modules or manual operation, can automatically monitor and restore the IPMB bus status and enable the backup module to take over the main control management function, and is reliable in operation.

Description

Technical Field

[0001] This invention relates to the field of intelligent control, and in particular to a method for redundant backup of the main control management function of a VPX or LRM architecture chassis. Background Technology

[0002] VPX and LRM chassis are general terms for various chassis models designed in accordance with corresponding management specifications. VPX architecture is commonly used in fields with high requirements for data transmission speed, reliability, and stability; LRM architecture is commonly used in scenarios such as video transmission where video quality and bandwidth are critical, and this architecture features rapid maintenance and adaptability to harsh environments. Both chassis designs provide standardized management configuration architectures and signal communication interfaces according to specifications, providing hardware support for acquiring overall system health management information and diagnosing module faults.

[0003] Both architectures typically contain a main control module, power supply module, fan module, and other service modules. The main control module acquires overall system health information, and each module contains an IPMC (Intelligent Platform Management Controller). The main control module usually uses the CPU to communicate with the IPMC, processes health query commands, and exchanges health information with other modules in the chassis via the IPMB bus (Intelligent Platform Management Bus). Therefore, the overall system management functions are mainly integrated into the main control module.

[0004] If the main control module malfunctions during chassis operation, the status of devices connected to the IPMB bus cannot be viewed through the interface or backend, leading to failure of health information interaction. In previous use, if the main control module malfunctioned, it was usually resolved by resetting the module through the backend or manually unplugging and restarting it. During troubleshooting, there was no backup module in the chassis to take over the health management function.

[0005] In certain specialized work environments, where the chassis requires prolonged power supply and personnel cannot reach the site for troubleshooting, a malfunction in the main control module can lead to the failure of overall system management functions and the inability to troubleshoot module failures through the backend system. After the IPMB bus hangs, powering down individual modules is ineffective; a complete system power-down and restart is required, resulting in a complete interruption of system operations and potentially causing more serious consequences.

[0006] Currently, the redundancy backup method for VPX architecture involves adding a main control module within the chassis for functional backup. However, this method is rarely used in LRM architecture chassis designs. In many use cases, to prevent resource waste, adding a module for functional backup is not adopted. Instead, other backup modules take over the main control management functions. This approach only works when the IPMB bus is functioning normally. If the IPMB bus communication level is clamped by any module, even if a backup module takes over the main control functions, the overall system management functionality cannot be restored.

[0007] Therefore, it is necessary to design a method that can be used in VPX and LRM architecture chassis, automatically restore the IPMB bus state, and redundancy backup of the main control management function without the need for additional main control modules. Summary of the Invention

[0008] To address the above problems, this invention provides a method for redundant backup of the chassis main control management function, which can automatically monitor and restore the IPMB bus status and enable the backup module to take over the main control management function without the need for external modules or manual operation, and provides a control interface for troubleshooting faulty modules.

[0009] The technical solution of this invention is: a method for redundant backup of the main control management function of a chassis, comprising the following steps: S1. In a VPX or LRM architecture chassis, the main controller, power supply, fan and other business modules are installed in the chassis, and the IPMC components in each module are interconnected through the chassis IPMB bus. S2. Connect the main controller and other IPMCs in the processor-equipped business modules to the corresponding processor through a serial port to form their own independent chassis management controllers (i.e., CHMCs). S3. Deploy and run the WEB control and management interface on the main controller or other business modules equipped with processors, and the WEB control and management interface interacts with the chassis management controller. S4. After the chassis is powered on, the chassis management controller in each module defaults to slave mode. The IPMC in the module automatically identifies the module type and switches between master and slave modes for management functions by recognizing the current IPMB bus hardware address. The master control module switches to master mode, while the other business modules act as backup modules and remain in slave mode. S5. When the main control module fails and the management function is abnormal, the IPMC in the main control module will switch to slave mode to judge the communication interruption through the serial port. Other backup modules, including the chassis management controller, will trigger the management function redundancy backup function by judging the management query timeout time on the IPMB channel and switch the backup module to master mode to work. S6. After the health management communication of the whole machine cannot be restored after each backup module switches the host working mode, each module in the chassis calculates the maximum IPMB management timeout coefficient based on the maximum number of slots in the chassis. Once the coefficient is reached, the IPMB bus recovery mechanism is triggered. S7. All modules within the chassis redefine the IPMB bus communication control pins to a strong push-pull output mode, forcibly pulling the IPMB bus level high. After the level goes high, the module's I2C peripherals are reinitialized, restoring the corresponding peripheral communication functions of the module's IPMC. Each backup module re-performs IPMB management bus timeout monitoring and enters the preparation queue for main control management function redundancy backup.

[0010] S8. After IPMB bus communication returns to normal, the backup module with the shortest IPMB channel timeout coefficient takes over the main control management function and resumes operation. Once the overall system health functions are running normally, users can control and troubleshoot faulty modules through the web interface. This achieves redundant backup of the chassis main control management function.

[0011] In step S1, each module inside the chassis transmits the IPMI (Intelligent Platform Management Interface) communication protocol through the IPMB management bus.

[0012] In step S2, the chassis management controller communicates with the IPMC via the IPMI protocol to obtain management information for each module.

[0013] In step S2, to ensure compatibility with the module design within the VPX and LRM architecture chassis, management software can be deployed on any service module containing a processor within the chassis, excluding the main control module, to form a complete chassis management controller.

[0014] In step S3, the WEB control management interface can process and display the IPMI data of each module in the chassis on the interface, and send independent control commands to the relevant modules through the graphical buttons on the interface.

[0015] In step S3, the chassis management controller communicates with the WEB control management interface by actively reporting, using a many-to-one star communication topology to ensure that the WEB control management interface can work normally when the active reporting source changes.

[0016] In step S4, after the whole machine is powered on, each module's IPMC detects the IPMB hardware address of the current slot. If it is the hardware address of the master control module slot, it switches to master working mode, while other slot modules remain in slave working mode.

[0017] If the backup module switches to master mode, its IPMC sends a master mode switching command to the processor via serial port. After the master module switches, the chassis management controller begins to acquire overall system management information. Upon receiving a management query command for this module, the IPMC processes and replies via serial port, and forwards query commands from other modules via the IPMB bus. In slave mode, the IPMC only replies to query information received on the IPMB bus and does not communicate with the processor via serial port.

[0018] In step S5, when the main control module fails, the main control management function communication is interrupted, the IPMC performs a timeout detection of the management query function, and after the communication interruption time exceeds the preset value, the module is switched to slave working mode, and the IPMC closes the serial communication port with the processor.

[0019] In step S5, other modules, including the chassis management controller, obtain the timeout waiting coefficient based on the IPMB bus hardware address. They then determine the communication status on the IPMB bus and sequentially perform management query function timeout checks based on the calculated timeout waiting coefficient. If the timeout exceeds a preset limit, the module is switched to host mode, and the IPMC opens the communication serial port with the processor, causing the corresponding module's chassis management controller to enter host operating mode. If the IPMB bus returns to normal, other modules cancel the management query function timeout check and perform normal master control management function communication.

[0020] In step S5, the main control redundant backup module can be a single module or multiple modules. In order to distinguish the delay coefficient of host communication timeout on the monitoring IPMB channel, each module's IPMC will count the delay based on the IPMB address of its slot.

[0021] In step S6, if none of the backup modules can restore the overall health management communication after switching the host working mode, the main control, power supply, fan and other business modules in the chassis will calculate the maximum IPMB management timeout coefficient based on the maximum number of slots in the current chassis. After the cumulative IPMB communication interruption time reaches the maximum coefficient, each module containing IPMC will trigger the IPMB bus recovery mechanism.

[0022] In step S7, the module containing the IPMC within the chassis redefines the corresponding control pin I / O mode of the IPMB bus communication to push-pull output mode, forcibly pulling the IPMB bus level high. During the high-level pull, the pin I / O mode is configured to be changed to pull-up input mode, and the current bus level is detected. After the level goes high, the module's I2C peripherals are reinitialized; otherwise, the pin I / O mode is configured to be changed to strong push-pull output mode again, pulling the IPMB bus level high until the IPMB bus communication function is restored. Each backup module re-performs IPMB management bus timeout monitoring and enters the preparation queue for main control management function redundancy backup.

[0023] In step S8, after the overall IPMB bus status returns to normal, the backup module with the shortest IPMB channel timeout coefficient takes over the main control management function and resumes operation. After the redundant backup of the main control management function returns to normal, the user can view the status of the faulty module through the WEB control management interface and troubleshoot the cause of the faulty module problem through IPMI data information. The user can issue commands to the chassis management controller through the WEB control management interface based on the obtained status information, so that the IPMC of the corresponding module controls the previously faulty module to perform a power-off or reset operation.

[0024] If the health information of the original faulty module returns to normal, the original faulty module IPMC will remain in slave mode and retain the redundancy backup function of the master control management function. If the master control management function of the current backup module malfunctions, the original faulty module that has recovered can re-enter the preparation queue for the redundancy backup of the master control management function.

[0025] This invention utilizes existing VPX and LRM chassis architectures, deploying a chassis management controller within the main controller and other processor-equipped service modules to achieve redundant backup of the main control management function without requiring additional modules. In the event of an IPMB bus anomaly, each module within the chassis can automatically restore the IPMB bus level; the entire process is automated via software without manual intervention. During redundant switching of the overall system health management function, the web control management interface continues to operate normally, and faulty modules can be handled through the web control management interface without affecting the overall system operation, offering a simple and convenient solution. Attached Figure Description

[0026] Figure 1 This is a diagram showing the connection relationships between the various modules inside the chassis. Figure 2 This is a schematic diagram showing the normal operation of the main control management function. Figure 3 This is a diagram illustrating an abnormal operation of the main control management function. Figure 4 This is a schematic diagram illustrating the redundancy and backup function of the main control management system. Figure 5 This is a flowchart illustrating the redundancy backup of the chassis main control management function. Figure 6 This is a flowchart of the automatic recovery function of the chassis IPMB bus. Detailed Implementation

[0027] The present invention is as follows Figure 1-6As shown, the intelligent chassis of VPX or LRM architecture contains a main controller, power supply, fan and several service modules. Each module is equipped with an IPMC (Intelligent Platform Management Controller). Each IPMC needs to be connected to two IPMB buses (Intelligent Platform Management Buses). The IPMC can obtain a unique IPMB communication address according to the different slots in the chassis.

[0028] Each IPMC module has a reserved serial port (UART) for communication with the processor (CPU) in the module. During use, CHMC (Chassis Management Controller) software can be deployed on the main controller and other modules containing processors. The CHMC software communicates with the IPMC via the serial port. The IPMC, which communicates in host mode, not only responds to the health query commands of its own module, but is also responsible for relaying IPMI messages from other boards through the IPMB bus. Figure 1 During the operation of the entire machine, the IPMC, including the power supply, fans, and other business modules, remains in slave mode and does not participate in the management function redundancy backup.

[0029] like Figure 2 and Figure 3 As shown, the overall health management information can ultimately be presented through a separate web-based control and management interface. Data exchange with the processor in the CHMC can be achieved via a simple network management protocol (SNMP), displaying module information in host operating mode and IPMI data from other slave devices within the chassis in the graphical interface and corresponding information columns.

[0030] The chassis management controller communicates with the web control management interface via active reporting. After receiving correct reported data, the web control management interface parses and processes it before displaying it. To ensure the normal operation of the signal source channel for the web control management interface, a many-to-one star communication topology is used. Even if the active reporting source changes, the web interface can still obtain data through other channels. During operation, the information displayed on the web control management interface will remain normal, and the interface display will not be affected during redundancy switching of the main control management function.

[0031] like Figure 2 As shown, when the entire machine is powered on for the first time, all modules in the chassis default to slave mode. The IPMC in each module identifies the current module type by obtaining the IPMB bus hardware address of the corresponding slot. The master-type IPMC will open the communication serial port with the processor and send a master switching command to the processor, causing the module to enter master mode. All other modules operate in slave mode.

[0032] During normal operation of the entire machine, the modules operating in master mode will communicate data via the IPMB bus. The other modules participating in the management function redundancy backup and other slave devices will clear the communication interruption detection flag after receiving data and re-perform the communication interruption status detection.

[0033] like Figure 3 As shown, during system operation, if the main control module malfunctions, the IPMC within the module will monitor the serial port communicating with the processor and perform timeout checks for management queries. If the communication interruption time exceeds a preset value, the IPMC will switch to slave mode and close the serial port communication with the processor. The chassis management controller within this module will then interrupt communication with the web control and management interface.

[0034] When the master control module malfunctions, the IPMB bus within the chassis will not transmit any information, and all IPMCs connected to the bus will operate in slave mode. At this time, other backup modules, including the chassis management controller, will obtain the timeout waiting coefficient based on the IPMB bus hardware address and monitor the communication status on the IPMB bus for timeouts.

[0035] like Figure 4 As shown, the backup module IPMC, which includes the chassis management controller, monitors the IPMB bus communication interruption time. If the interruption exceeds a preset time, the current IPMC module is switched to host mode. The serial communication port between the IPMC and the processor is enabled, and a host mode switching command is sent to the processor, causing the corresponding module's chassis management controller to enter host mode. If normal operation resumes on the IPMB bus, other modules cancel the timeout detection for management query functions and resume normal master control management communication.

[0036] When there are multiple redundant backup modules for main control management functions in the chassis, the timeout trigger time for each backup module is calculated based on the IPMB hardware slot address and the delay coefficient, which avoids communication conflicts caused by multiple backup modules switching to host mode at the same time.

[0037] like Figure 6 As shown, if the entire machine still cannot restore health management communication after each backup module switches the host working mode, the main control, power supply, fan and other business modules in the chassis calculate the maximum IPMB management timeout coefficient (number of slots * delay coefficient) based on the current maximum number of slots in the chassis. After monitoring that the cumulative IPMB communication interruption time reaches the maximum coefficient, each module containing IPMC will trigger the IPMB bus recovery mechanism.

[0038] Each module within the chassis redefines the I / O mode of the corresponding control pins for IPMB bus communication in the IPMC, changing it to a strong push-pull output mode to force the IPMB bus level high. To detect whether the IPMB bus has returned to a high level (idle state), the pin I / O mode is configured to change to a pull-up input mode during the forced bus high period, and the current bus level is detected. After detecting a high level, the module's I2C peripherals are reinitialized. If it is still low, the pin I / O mode is configured to change to a strong push-pull output mode to force the IPMB bus level high until the IPMB bus communication function is restored.

[0039] The modification of the I / O pin mode mentioned above refers to changing the original multiplexed open-drain mode used for microcontroller I2C peripheral communication to a push-pull output mode. In push-pull output mode, current flows whether the microcontroller pulls the pin high or low, providing better drive capability. After the communication pin completes the bus pull-high operation, the microcontroller I / O port is configured to operate in input detection mode to detect the bus status. If it is still abnormal, the pull-high operation is performed again; if the bus returns to normal, the I2C peripheral is reinitialized, resetting the relevant I2C control registers and the I2C communication I / O port operating mode within the microcontroller.

[0040] After the IPMB bus returns to normal, each backup module re-performs IPMB management bus timeout monitoring and enters the preparation queue for redundancy backup of the main control management function. The backup module with the shortest IPMB channel timeout coefficient takes over the main control management function and resumes operation.

[0041] After the backup module takes over the main control functions, the chassis management controller receives and processes health information and actively reports it to the WEB control management interface. The interface can display the status information of each slot in the chassis normally. The presence status of the faulty module can be viewed through the WEB control management interface, the cause of the faulty module problem can be investigated through IPMI data information, and commands can be issued to the chassis management controller through the WEB control management interface to cause the IPMC of the corresponding module to perform a power-off or reset operation on the original faulty module.

[0042] If the health information of the previously faulty module returns to normal, the IPMC of that module will continue to operate in slave mode, retain the redundancy backup function of the master control management function, and re-enter the preparation queue for the redundancy backup of the master control management function.

[0043] In operation, the redundant backup module for the main control management function can be single or multiple, as long as it meets the composition requirements of the chassis management controller. Because the timeout detection period for each backup module's management query function is determined by the IPMB hardware slot address, communication conflicts caused by multiple backup modules simultaneously switching to host mode will not occur after a main control management function malfunction. When an IPMB bus malfunctions, the IPMB bus can be automatically restored because each module within the chassis contains an IPMC.

[0044] The embodiments described above are only for illustrating the technical solutions of the present invention, and are not intended to limit it. Although specific embodiments have been described in detail herein, those skilled in the art can still modify them or replace some technical features in an equivalent manner, and these changes do not depart from the core ideas and protection scope embodied in the embodiments of the present invention.

Claims

1. A method for redundant backup of the main control management function of a chassis, characterized in that, Includes the following steps: S1. In a VPX or LRM architecture chassis, the main controller, power supply, fan and other business modules are installed in the chassis, and the IPMC components in each module are interconnected through the chassis IPMB bus. S2. Connect the main controller and other IPMCs in the processor-equipped business modules to the corresponding processor through a serial port to form their own independent chassis management controllers; S3. Deploy and run the WEB control and management interface on the main controller or other business modules equipped with processors, and the WEB management interface interacts with the chassis management controller for data exchange. S4. After the chassis is powered on, the chassis management controller in each module defaults to slave mode. The IPMC in the module automatically identifies the module type and switches between master and slave modes for management functions by recognizing the current IPMB bus hardware address. Specifically, the master control module switches to master mode, while the other business modules act as backup modules and remain in slave mode. S5. When the main control module fails and the management function is abnormal, the IPMC in the main control module will switch to slave mode to judge the communication interruption through the serial port. Other backup modules, including the chassis management controller, will trigger the management function redundancy backup function by judging the management query timeout time on the IPMB channel and switch the backup module to master mode to work. S6. After the health management communication of the whole machine cannot be restored after each backup module switches the host working mode, each module in the chassis calculates the maximum IPMB management timeout coefficient based on the maximum number of slots in the chassis. Once the coefficient is reached, the IPMB bus recovery mechanism is triggered. S7. Each module redefines the control pin of IPMB bus communication to strong push-pull output mode, forcibly pulls the IPMB bus level high, and after the level is pulled high, the module's I2C peripherals are reinitialized to restore the communication function of the corresponding peripherals of the module's IPMC. Each backup module re-performs IPMB management bus timeout monitoring and enters the preparation queue for redundancy backup of the main control management function; S8. After the IPMB bus communication returns to normal, the backup module with the shortest IPMB channel timeout coefficient will take over the main control management function and start working again. After the overall health function is running normally, the user can control and troubleshoot the faulty module through the WEB interface.

2. The method for redundant backup of the main control management function of a chassis according to claim 1, characterized in that, In step S2, the chassis management controller communicates with the IPMC via the IPMI protocol to obtain management information for each module.

3. The method for redundant backup of the main control management function of a chassis according to claim 1, characterized in that, In step S3, the chassis management controller in host mode communication communicates with the WEB control management interface by actively reporting, and uses a many-to-one star communication topology for connection.

4. The method for redundant backup of the main control management function of a chassis according to claim 1, characterized in that, In step S5, other modules including the chassis management controller obtain the timeout waiting coefficient based on the IPMB bus hardware address, judge the communication status on the IPMB bus, and perform management query function timeout detection in sequence according to the calculated timeout waiting coefficient. If the preset time is exceeded, the module is switched to host, the IPMC opens the communication serial port with the processor, so that the chassis management controller of the corresponding module enters the host working mode.

Images