Centralized application system performance problem comprehensive analysis method

Abstract

The invention discloses a centralized application system performance problem comprehensive analysis method, comprising the following steps: S1, uniformly acquiring performance indexes, warning abnormity and LOG scattered each monitoring system; S2, parsing the unstructured LOG and performance indexes into structured information through a data processing module, and performing classification; S3, storing the classified data; S4, performing statistics and summarizing for the data through a data presentation module; S5, quickly screening out data of each hierarchy within a fault period through anend-to-end hierarchy presentation function of the data presentation module; S6, checking warning data or abnormal information of each hierarchy, finding out atomic warning or the abnormal information, removing derivative warning or abnormal information, and finding out the root of the problem. The method provided by the invention can solve a problem of currently scattering data, high professionalization demand of some data and inefficient data analysis through uniform data acquisition, data integration processing and uniform data presentation.

Description

technical field [0001] The invention relates to a method for analyzing system performance data, in particular to a method for comprehensively analyzing performance problems of a centralized application system. Background technique [0002] Performance problems of application systems generally involve multiple levels, such as hosts, databases, networks, middleware, storage, etc. in the system are all related. When a performance problem occurs in the application system, it will cause abnormalities in the front-end application of user interaction, affecting user perception. In general, during the design and operation of a large-scale application system, an independent monitoring system will be deployed and alarms will be configured for components at all levels. Maintained by different teams or vendors. Since the monitoring, alarm logs, and performance monitoring data of each component are scattered in different monitoring systems, and different components and monitoring are h...

AI Technical Summary

Problems solved by technology

[0008] 1. The original monitoring data of each component of the application system is scattered in their own independent monitoring systems. The analysis and location of the problem requires the cooperation of various operation and maintenance teams. It is impossible to directly and quickly locate and analyze the cause of the application performance problem. The whole process takes a long time and communication costs high

[0009] 2. The use of each monitoring system is complicated, and the technical threshold is high, requiring a lot of manual processing and analysis

At the same time, it is difficult for non-professionals to understand the data output from monitoring, which needs to be viewed by professionals, resulting in duplication of work

[0010] 3. Data between monitoring systems cannot be shared directly

When a performance problem occurs, each operation and maintenance team often spends a lot of time looking for the cause in its own maintenance system, but the real cause of the problem may exist in the alarms of other monitoring systems

Images