Method and system for distributed automatic detection of abnormal operation of applications

Abstract

The present invention relates to a method and system for distributed automatic detection of abnormal operation of applications. The method includes: performing patrol inspection and matching on log files according to keywords in the logs, packaging the logs matched with the keywords into a task process, saving them to a database, and storing them in the database Each periodic task process in the system is loaded into the distributed task scheduling module, and the task process is executed periodically and sequentially, and the execution result is fed back. According to the feedback execution result, the abnormal information is obtained and the detection is completed; the present invention can automatically detect whether the application service is running Abnormal, avoiding the problem of a large number of repetitive logs due to the need to scan the full amount of logs for each detection, and by matching the incremental scans of the application logs for each detection, the detection process can be simplified while the detection results It is more accurate, and through multiple distributed task scheduling modules to balance the load, it greatly improves the disaster recovery capability of the system and ensures the safe operation of the system.

Description

technical field [0001] The invention relates to the field of mobile communication, in particular to a distributed automatic detection method and system for abnormal operation of applications. Background technique [0002] With the development of society and the advancement of science, my country has made great progress in the field of mobile communications. The scale of daily operation and maintenance management and control of operators has gradually expanded, and the number of production system services for daily operation and maintenance monitoring has increased sharply. , how to detect the abnormal operation of the application is particularly important. The existing method usually uses manual login to the remote host to check whether there is an abnormal error log in the operation log. This method is very passive. Only when there is a problem with the service application, the operation and maintenance personnel will Going to check the log content will increase the workload ...

AI Technical Summary

Problems solved by technology

[0003] At present, daily operation and maintenance personnel use the log analysis tool of the ELK platform to retrieve and analyze the log content, but the existing methods have the following disadvantages: First, when performing retrieval and analysis, the full amount of logs is returned every time, resulting in a large number of The repetitive logs cannot solve the problem of single point of failure, which increases the workload of operation and maintenance personnel; second, once there is a problem with Logstash, ELK will not work normally. Therefore, a new technical method is urgently needed to solve the single point of failure. Point failure problem, simplify the system detection process, reduce the workload of operation and maintenance personnel

Images