CN112905377B - A method, system and medium for avoiding BMC crash - Google Patents

Abstract

The invention discloses a method for avoiding BMC (baseboard management controller) crash, which comprises the following steps: executing a monitoring strategy, and monitoring the CPU utilization rate and an interrupt instruction; setting a flag area, a non-flag area, an interrupt parameter threshold value and a flag threshold value; when the CPU utilization rate reaches a flag area, if an instruction generator generates an interrupt instruction, executing an incremental strategy on an interrupt parameter and a flag corresponding to the instruction generator; according to the method, when the CPU utilization rate reaches the non-flag area, the decrement strategy is executed on the flag corresponding to the instruction generator, and when the server instruction generator is abnormal in factory manufacturing, the fault judgment of testers caused by the fact that the BMC cannot respond due to the fact that the BMC is crashed can be avoided.

Description

A method, system and medium for avoiding BMC crash

technical field

The present invention relates to the technical field of server detection, in particular to a method, system and medium for avoiding BMC crash.

Background technique

When the user presses the power command generator to start the server, the BMC records a SEL log to record that the server system has been pressed the power command generator. When the user presses the restart command generator, the BMC records a SEL log. It is used to record that the server system has been pressed to restart the command generator. When the user presses the interrupt command generator, the BMC records a system event log to record that the server system has been pressed to the interrupt command generator. The above BMCs are all By registering the GPIO interrupt handler, judging the change of the high and low levels of the GPIO pins, it is known that the command generator on the server system has been pressed by the user, and then the GPIO interrupt function is used to quickly execute the events corresponding to the pressed command generator. deal with.

However, when there is a problem with the instruction generator, many interrupts will be generated in a short time, which will cause the BMC to centrally process these interrupt functions and reduce the CPU utilization of the BMC, so that the upper-layer services cannot be processed immediately and effectively.

SUMMARY OF THE INVENTION

The invention mainly solves the problem that when the server instruction generator is abnormal, the BMC repeatedly processes the abnormal interruption, which leads to the reduction of the CPU utilization rate of the BMC or the collapse of the BMC.

In order to solve the above-mentioned technical problems, a technical scheme adopted by the present invention is: a method for avoiding the collapse of BMC is provided, comprising the following steps:

Execute monitoring strategy, monitor CPU utilization and interrupt instruction of instruction generator;

Set the flag area, non-flag area, interrupt parameter threshold and flag threshold;

When the CPU utilization rate reaches the flag area, if the instruction generator generates the interrupt instruction, then the corresponding interrupt parameter and the flag of the instruction generator are executed increment strategy;

When the CPU utilization reaches the non-flag area, a decrementing strategy is performed on the flag corresponding to the instruction generator.

Further, the step of setting the flag area, the non-flag area, the interrupt parameter threshold value and the flag threshold value further includes: the non-flag area includes a first threshold interval, a second threshold interval and a buffer area, and the flag area includes a first threshold interval. The parameter interval and the second parameter interval.

Further, the incremental strategy includes a first incremental strategy and a second incremental strategy;

When the CPU utilization reaches the first parameter interval, the monitoring strategy monitors whether the instruction generator generates the interrupt instruction, and if so, executes the interrupt parameter according to the interrupt instruction. The first increment strategy, when the interrupt parameter reaches the interrupt parameter threshold, execute the first increment strategy on the flag, and execute the emergency strategy when the flag is greater than the flag threshold.

Further, when the CPU utilization rate reaches the second parameter interval, the monitoring strategy monitors whether the instruction generator generates the interrupt instruction, and if so, executes the interrupt parameter according to the interrupt instruction. In the second increment strategy, when the interrupt parameter reaches the interrupt parameter threshold, the first increment strategy is executed on the flag, and when the flag is greater than the flag threshold, an emergency strategy is executed.

Further, the step of implementing the emergency strategy further comprises: closing the corresponding interrupt handler of the instruction generator, stopping the processing of the interrupt instruction of the instruction generator, and storing the information of closing the interrupt handler in the system event log, and perform alarm operations.

Further, when the described CPU utilization reaches the non-flag area, the step of executing the decreasing strategy to the flag corresponding to the instruction generator further comprises: if the CPU utilization reaches the first threshold interval, then Initializing the interrupt parameter and the flag corresponding to the instruction generator.

Further, when the described CPU utilization rate reaches the described non-flag area, the step of executing the decreasing strategy to the described flag corresponding to the instruction generator further includes:

When the CPU utilization rate reaches the second threshold interval, execute the first decreasing strategy on the flag corresponding to the instruction generator, and determine whether the interrupt handler corresponding to the instruction generator is closed State, if the described interrupt handler of the described instruction generator is in the closed state, then judge whether the described flag of the described instruction generator is equal to the described flag threshold, if equal, then keep the described interrupt of the described instruction generator The processing program is in a closed state. If it is smaller than the value, the interrupt processing program of the instruction generator is opened, and the information of the opened interrupt processing program is stored in the system event log.

Further, the first increment strategy is to increment the interrupt parameter or the flag based on the first parameter, and when the interrupt parameter or the flag reaches the interrupt parameter threshold or the flag threshold, continue to increase from the initial value;

The second increment strategy is to increment the interrupt parameter or the flag based on the second parameter, and when the interrupt parameter or the flag reaches the interrupt parameter threshold or the flag threshold, the initial The value continues to increment.

The present invention also provides a system for avoiding BMC crash, which is applied to the server, and the system includes: a monitoring module, a weighting module, a strategy module and a system event module;

Described monitoring module is used to monitor the CPU utilization rate of described server, and the interrupt instruction of the instruction generator of described server, and set flag area and non-flag area;

The strategy module stores an increasing strategy, a decreasing strategy and an emergency strategy;

When the CPU utilization rate reaches the flag area, the weighting module executes the incremental strategy, executes interrupt parameter weighting and flag weighting according to the interrupt instruction, and executes the interrupt handler corresponding to the instruction generator. describe emergency strategies;

When the CPU utilization rate reaches the non-flag area, the weighting module executes the decrement strategy to the flag, and executes the emergency strategy to the interrupt handler corresponding to the instruction generator;

The system event module is used for storing the information of the interrupt handler.

The present invention also provides a computer-readable storage medium, where a computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, any one of the method steps for avoiding a BMC crash is implemented.

The beneficial effects of the present invention are:

1. The method for avoiding BMC crash according to the present invention can realize that when the server is manufactured in the factory, when the server command generator is abnormal, the BMC will not be unable to respond due to the crash of the BMC, causing the server tester to mistakenly think that it is the server BMC firmware. If there is a problem with the design, and the abnormal command generator can be located through the system event log, there is no need for the server R&D personnel to go to the factory to confirm the cause of the abnormal server command generator.

2. The system for avoiding BMC collapse according to the present invention can perform weighting of the interrupt instruction of the instruction generator according to the strategy of the strategy module, and after the weighting ends, execute the shutdown of the interrupt handler according to the interrupt threshold, when the interrupt handler is satisfied. When the condition is turned on, the interrupt handler will be turned on again, and the information of the turned off or turned on interrupt handler can be saved as a system event, which is convenient for maintenance personnel to conduct maintenance.

3. The medium for avoiding BMC collapse according to the present invention can realize the strategy of increasing or decreasing the interrupt parameter corresponding to the instruction generator, and when the interrupt parameter reaches the threshold value, a flag is generated, and the flag is used to determine whether the instruction generator is abnormal, And after the abnormality occurs, the interrupt handler of the abnormal instruction generator is closed to reduce the load pressure of the CPU and ensure the normal operation of the CPU.

Description of drawings

In order to illustrate the specific embodiments of the present invention or the technical solutions in the prior art more clearly, the following briefly introduces the accompanying drawings that need to be used in the description of the specific embodiments or the prior art. Obviously, the accompanying drawings in the following description The drawings are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative efforts.

1 is a schematic diagram of the method for avoiding BMC collapse according to Embodiment 1 of the present invention;

2 is a flowchart of the method for avoiding BMC collapse according to Embodiment 1 of the present invention;

3 is an interval division diagram of the method for avoiding BMC collapse according to Embodiment 1 of the present invention;

FIG. 4 is a schematic diagram of a system for avoiding BMC collapse according to Embodiment 2 of the present invention.

Detailed ways

The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

In the description of the present invention, it should be noted that the terms "installed", "connected" and "connected" should be understood in a broad sense, unless otherwise expressly specified and limited, for example, it may be a fixed connection or a detachable connection connection, or integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or it can be the internal connection of two components, which can be a wireless connection or a wired connection connect. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood in specific situations.

In addition, the terms "first" and "second" are only used for descriptive purposes, and should not be construed as indicating or implying relative importance or implying the number of indicated technical features. Thus, a feature delimited with "first", "second" may expressly or implicitly include at least one of that feature. In the description of the present invention, "plurality" means at least two, such as two, three, etc., unless otherwise expressly and specifically defined.

It should be noted that, in the description of the present invention, a BMC (Baseboard Management Controller) is a baseboard management controller, and a CPU (Central Processing Unit) is a central processing unit.

Example 1

An embodiment of the present invention provides a method for avoiding BMC crash, please refer to FIG. 1 to FIG. 3 , including the following steps;

In order to avoid the phenomenon that when the server instruction generator is abnormal, the instruction generator abnormally transmits the interrupt function to the BMC, causing the CPU load of the BMC to increase instantaneously, resulting in the collapse of the BMC, the present invention first creates a monitoring strategy to continuously monitor the CPU of the server. Utilization, set the CPU utilization threshold and interrupt parameters. According to the CPU utilization, the CPU utilization is divided into a flag area and a non-flag area. The flag area is the CPU utilization rate greater than or equal to 70%, and the non-flag area is the CPU utilization rate. Less than 70%, when the CPU utilization reaches 70%, less than 90%, confirm whether the interrupt of the command generator occurs on the server, because each command generator corresponds to an interrupt response program, and the interrupt counter is pressed for each command generator. The interrupts generated in the next step are counted, and the interrupt parameters are initialized at the beginning. The interrupt parameter is 0. When the interrupt occurs when the command generator is pressed, the interrupt parameter is incremented by 1. If no interrupt occurs, the monitoring strategy will continue to be used. Monitor CPU utilization;

When the monitoring strategy monitors that the CPU utilization is greater than or equal to 90%, the CPU can no longer respond to burst-intensive interrupts, so the interrupt parameters are further weighted to confirm whether the interrupt operation of the instruction generator occurs on the server. , the interrupt parameter is incremented by 2.

The above addition of 1 or 2 to the interrupt parameter is just a simple weighting relationship. The number is not what we want to discuss, nor is it the scope of protection. Adding 1 or 2 is just an ideal situation. According to the operating status of the server, or The application scenario of the server and the performance of the server CPU can appropriately select a better interrupt parameter for weighting.

When the interrupt parameter is greater than the upper limit of the interrupt parameter of 10, a flag is generated. When the number of flags is greater than 10, the instruction generator that generates the interrupt, such as the shutdown instruction generator and the interrupt handler corresponding to the restart instruction generator, will be closed. That is to say, it is determined that the instruction generator is abnormal, and the CPU refuses to receive the interrupt instruction of the instruction generator, protects the normal operation of the upper-layer business of the server, and records the information log of the instruction generator while turning off the instruction generator. In the SEL log system event log, in this way, the staff is reminded that the abnormality of the command generator has affected the normal operation of the CPU, and the staff is reminded to respond quickly.

The monitoring strategy monitors that the CPU utilization is lower than 70%. When the non-flag area is reached, the CPU utilization can already process some interrupt instructions. Therefore, the CPU opens the closed interrupt handler corresponding to the instruction generator according to the situation.

When the CPU utilization is lower than 70% and greater than or equal to 60%, it is equivalent to a buffer. The monitoring strategy does not process the interrupt handler, but continues to wait for the CPU utilization to continue to change and wait for the CPU utilization to increase or decrease.

The monitoring strategy monitors that the CPU utilization is less than or equal to 60%, and when it is greater than 50%, reduce the number of flags by one, and the number of flags is at least 0. Because the remaining utilization of the CPU is sufficient to deal with various interrupt instructions, you can try to perform interrupt operations. The monitoring strategy confirms the current interrupt handler status of the instruction generator. When the number of flags of the instruction generator is reduced, if the number of flags of the instruction generator is equal to 10, that is, the original number of flags of the instruction generator is 11. , at this time, there may be a serious problem with the command generator, and the maintenance personnel have not dealt with it in time, and the interrupt handler of the command generator will continue to be closed. The utilization rate of the CPU remains stable, and in order to exceed the non-flag area, the utilization rate of the CPU performs at least one reciprocating motion in the buffer area and the CPU utilization rate of 60% to 50%. The monitoring strategy monitors the number of flags of the instruction generator as When it is less than 10, open the interrupt handler that the command generator has been closed, and record the information that the command generator is turned on in the system time log, and notify the maintenance personnel that the command generator has been turned on, and the CPU can execute the command The processing of the interrupt instruction of the generator.

When the monitoring strategy monitors that the CPU utilization is less than or equal to 50%, the initialization strategy is triggered, and the initialization strategy clears the interrupt parameters and the number of flags, because the CPU at this time can process a large number of services and execute a large number of interrupt instructions.

Example 2

An embodiment of the present invention provides a system for avoiding BMC crash, which is applied to a server. Please refer to FIG. 4 . The system includes;

Monitoring module, weighting module, command generator, policy module and system event module.

The command generator is used to control the server with various commands, such as shutdown command, restart command, and standby command. Every time the command generator generates a command, it will interrupt the server BMC. instruction, the CPU of the BMC will crash due to instantaneously receiving a large number of instructions, resulting in the BMC not working properly.

The monitoring module monitors the utilization rate of the CPU, and divides the flag interval and the flag interval according to the utilization rate of the CPU. In this embodiment, according to the performance index of the CPU, when the utilization rate of the CPU is greater than or equal to 70%, it is set as the flag interval, which is less than 70%. % is set as a non-flag interval. When the CPU utilization reaches 70%, the monitoring module further divides the interval.

When the CPU utilization is greater than or equal to 70% and less than 90%, the monitoring module monitors whether the instruction generator generates an interrupt command. If an interrupt command is generated, the weighting of the interrupt parameters and flags is carried out through the cooperation of the weighting module and the strategy module. When the shutdown command generator or the restart command generator is monitored to generate an interrupt command, each command generator is individually weighted, and each time an interrupt command is monitored, the weighting module commands the first increment strategy of the strategy module, that is, each time an interrupt command is generated. Increment by 1 each time, the interrupt instruction of the instruction generator will be interrupted and incremented. When the interrupt parameter is incremented to 10, the first increment strategy of the strategy module is executed for the flag, and the flag is gradually incremented. When the number of flags is greater than 10, the The interrupt handler corresponding to the instruction generator stops, and the BMC no longer accepts the interrupt instruction of the interrupt generator. Because of frequent occurrence of interrupt instructions, it means that the interrupt generator is abnormal, and records the stopped interrupt handler in the system event module. In this way, the site can be protected so that maintenance personnel can repair the abnormal command generator.

Similarly, when the CPU utilization is greater than or equal to 90%, it means that the CPU has very little usable space, and a faster weighting strategy is required. Every time the monitoring module monitors the shutdown command generator or restarts the command generator, when an interrupt command occurs, the command will be updated accordingly. The interrupt parameter of the interrupt instruction generated by the generator is subjected to the second increment strategy, that is, adding 2 each time. When the interrupt parameter is incremented to 10, the first increment strategy of the strategy module is executed for the flag, and when the number of flags is greater than 10, the The interrupt handler corresponding to the instruction generator stops, and the BMC no longer accepts the interrupt instruction of the interrupt generator. Because of frequent occurrence of interrupt instructions, it means that the interrupt generator is abnormal, and records the stopped interrupt handler in the system event module. In this way, the site can be protected so that maintenance personnel can repair the abnormal command generator.

When the CPU utilization is reduced to the non-flag area, the CPU utilization is subdivided. When it is less than 70% and greater than or equal to 60%, the CPU utilization is divided into buffers. At this time, it is not known whether the CPU utilization is stable. Therefore, when the CPU utilization is the buffer, the interrupt instruction is not counted, and the interrupt handler of the instruction generator will not be opened or closed.

When the CPU utilization rate is reduced to less than 60% and greater than 50%, as long as the CPU utilization rate enters once, a flag decrement strategy is performed for all interrupt processors, that is, the flag of the instruction generator is decreased by 1, because when the flag is greater than 10, that is, 11, the interrupt handler will be closed, so when the CPU utilization is reduced to less than 60% and greater than 50%, and when the flag corresponding to the instruction generator is less than 10, the closed interrupt handler will be opened. , which means that the CPU utilization is very stable after two times when it is less than 60% and greater than 50%. Therefore, when the flag is equal to 10, the closed interrupt handler will not be opened. When the closed interrupt handler is opened, it will be processed. The situation is sent to the system event module for saving, and the maintenance personnel are notified that the command generator has undergone a restart and needs to be carefully confirmed.

When the CPU utilization is less than or equal to 50%, the interrupt parameters and flags are initialized. Usually initialization is to clear them. The initialization value can also be set according to the performance of the CPU.

The above-mentioned embodiments of the present invention disclose the serial numbers of the embodiments only for description, and do not represent the advantages and disadvantages of the embodiments.

Those of ordinary skill in the art can understand that all or part of the steps of implementing the above-mentioned embodiments can be completed by hardware, and the program completed by instructing relevant hardware through a program can also be stored in a computer-readable storage medium. The storage medium can be a read-only memory, a magnetic disk or an optical disk, and the like.

The above descriptions are only the embodiments of the present invention, and are not intended to limit the scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the description and drawings of the present invention, or directly or indirectly applied to other related technologies Fields are similarly included in the scope of patent protection of the present invention.

Claims (4)

1. A method for avoiding BMC crash is characterized by comprising the following steps:

executing a monitoring strategy, and monitoring the CPU utilization rate and an interrupt instruction of an instruction generator;

setting a flag area, a non-flag area, an interruption parameter threshold and a flag threshold, wherein the non-flag area comprises a first threshold interval, a second threshold interval and a buffer interval, and the flag area comprises a first parameter interval and a second parameter interval;

when the CPU utilization rate reaches the flag area, if the instruction generator generates the interrupt instruction, executing an increasing strategy on the interrupt parameter and the flag corresponding to the instruction generator; the increment policies include a first increment policy and a second increment policy;

when the CPU utilization rate reaches the first parameter interval, the monitoring strategy monitors whether the instruction generator generates the interrupt instruction, if so, the first incremental strategy is executed on the interrupt parameter according to the interrupt instruction, when the interrupt parameter reaches the interrupt parameter threshold, the first incremental strategy is executed on the flag, and when the flag is greater than the flag threshold, the emergency strategy is executed;

when the CPU utilization rate reaches the second parameter interval, the monitoring strategy monitors whether the instruction generator generates the interrupt instruction, if so, the second incremental strategy is executed on the interrupt parameter according to the interrupt instruction, when the interrupt parameter reaches the interrupt parameter threshold, the first incremental strategy is executed on the flag, and when the flag is greater than the flag threshold, the emergency strategy is executed;

the step of executing the emergency policy includes: closing an interrupt processing program corresponding to the instruction generator, stopping processing of the interrupt instruction of the instruction generator, storing information of closing the interrupt processing program in a system event log, and executing an alarm operation;

when the CPU utilization rate reaches the non-flag area, executing a decreasing strategy on the flag corresponding to the instruction generator; if the CPU utilization rate reaches a first threshold interval, initializing the interrupt parameters and the flag corresponding to the instruction generator; when the CPU utilization rate reaches the second threshold interval, executing the first decreasing strategy on the flag corresponding to the instruction generator, judging whether the interrupt processing program corresponding to the instruction generator is in a closed state, if the interrupt processing program of the instruction generator is in the closed state, judging whether the flag of the instruction generator is equal to the flag threshold, if so, keeping the interrupt processing program of the instruction generator in the closed state, if not, starting the interrupt processing program of the instruction generator, and storing the information of the started interrupt processing program in the system event log.

2. The method of claim 1, wherein the BMC crash prevention step comprises: the first increment strategy is to increment the interruption parameter or the flag by taking the first parameter as a base number, and when the interruption parameter or the flag reaches the interruption parameter threshold or the flag threshold, the increment is continued from an initial value;

the second increment strategy is to increment the interruption parameter or the flag by taking the second parameter as a base number, and when the interruption parameter or the flag reaches the interruption parameter threshold or the flag threshold, the increment is continued from an initial value.

3. A system for avoiding a BMC crash, applied to a server, using the method for avoiding a BMC crash of claim 1 or 2, wherein the system comprises: the system comprises a monitoring module, a weighting module, a strategy module and a system event module;

the monitoring module is used for monitoring the CPU utilization rate of the server and an interrupt instruction of an instruction generator of the server, and setting a flag area and a non-flag area;

the strategy module stores an increasing strategy, a decreasing strategy and an emergency strategy;

when the CPU utilization rate reaches the flag area, the weighting module executes the increment strategy, executes the interrupt parameter weighting and the flag weighting according to the interrupt instruction, and executes the emergency strategy on an interrupt processing program corresponding to the instruction generator;

when the CPU utilization rate reaches a non-flag area, the weighting module executes the decrement strategy on the flag and executes the emergency strategy on an interrupt processing program corresponding to the instruction generator;

the system event module is used for storing the information of the interrupt processing program.

4. A computer-readable storage medium, in which a computer program is stored, which, when being executed by a processor, carries out the method steps of avoiding a BMC crash of claim 1 or 2.

Images