System and method for process monitoring and multi-stage recovery

Abstract

The invention relates to a system and a method for process monitoring and multi-stage recovery. The system comprises a hardware part and a software part, wherein the hardware part comprises a communication module and a control module; a control terminal of the control module is electrically connected with a restart jumper on the main board of an outer computer; and the software part comprises a monitoring module, a processing module, a data transceiver module an event recording module and a system recalling and recording module. The method corresponds to the system. The system and the method can achieve the comprehensive monitoring of the processes, perform the multi-stage recovery operation, and leave the event record and the system recall record to provide valid clues for the factory technical personnel in tracing the causes of the problems.

Description

technical field [0001] The invention relates to a process monitoring and multi-level recovery system, and a process monitoring and multi-level recovery method corresponding to the system, and is especially suitable for industrial control computers. Background technique [0002] At present, computers, industrial computers, and embedded devices are usually equipped with operating systems to run target applications. When users need to run multiple applications for a long time without interruption, they often need to use a monitoring system. However, the existing monitoring system functions Limited, it is difficult to meet user needs. [0003] The most traditional monitoring system - motherboard watchdog is only suitable for single-tasking operating systems. If it is directly used in multi-tasking operating systems, it cannot accurately respond to restarting the computer when a certain process is abnormal. [0004] In order to still be able to use the motherboard watchdog under...

AI Technical Summary

Problems solved by technology

[0005] However, currently commonly used mainstream operating systems such as windows system, linux system, and unix system have all blocked the way to directly call the motherboard watchdog, and the above-mentioned improved technical solutions are difficult to apply to the current operating system

In addition, many servers and workstations do not adopt the industrial computer architecture, and there is no motherboard watchdog at all. In this case, the above-mentioned improved technical solution cannot be applied.

[0006] For the monitoring of application programs in the operating system, researchers have developed corresponding technical solutions. For example, the Chinese invention patent with patent number 200610104447.9 and authorized announcement number CN100389404C discloses a computer application program monitoring and exception handling method, but the The technical solution also has the following disadvantages: (1) Before each application module is started, it is necessary to establish an independent monitoring module to communicate with it, which will increase the operating overhead and increase the consumption of system resources, which is not conducive to maintaining the system Stablize

(2) Judging whether the application module is abnormal based only on the communication feedback of the application module. If the monitoring module cannot handle it in time, it will easily lead to the collapse of the entire system due to resource exhaustion, and the expected monitoring purpose cannot be achieved.

(3) If the entire system crashes unexpectedly, the computer cannot be restarted automatically. This is undoubtedly a huge disaster for unattended computers, and it will cause heavy losses to users in severe cases; at the same time, in this case It is also impossible to leave any effective clues that can be traced. You can only restart the computer to continue running, and leave hidden dangers that are likely to recur.

[0007] In addition, as far as the applicant knows, for a multi-tasking operating system (such as WINDOWS system), its system environment allows multi-process and multi-thread operation. Many reasons, such as deadlock and no response of the process in the project, crash but not automatically exit, error, abnormal exit, memory leak, process CPU usage and memory usage continue to increase, which eventually lead to slow operating system response and operation System failure and other phenomena, because the operator user does not understand the technical details of the project, he cannot deal with it on site, and the project manufacturer is usually not on the site, so it is beyond his reach, which often causes the on-site project function to be interrupted and unable to recover, and wait until the manufacturer's technical personnel arrives At the scene, it is impossible to know the process and cause of the failure, and it is impossible to specifically determine which process caused the problem, which aspect of the problem occurred, etc. As a result, even if the manufacturer's personnel rushed to the scene, they could not find the problem at the scene However, due to the complexity of the operation situation on site, the problems that arise are often not found when the products and systems are copied for a short time before leaving the factory.

For this technical problem, there is no perfect solution in the prior art

Images