Micro-service fault root cause positioning method based on fault propagation graph

Abstract

The invention discloses a micro-service fault root cause positioning method based on a fault propagation graph (FRL)-MFPG (Microservice Fault Root Case On MFPG), which is characterized in that the FRL-MFPG can be used for positioning the root cause of a micro-service fault; the FRL-MFPG can be used for positioning the root cause of the micro-service fault; the method comprises the following steps: firstly, aiming at the problem that the dynamic displacement of a micro-service KPI (Key Performance Indicator) is not well considered in an existing correlation analysis method, designing a new anomaly scoring algorithm to measure the fault degree of a micro-service component, and constructing a fault propagation graph (MFPG); secondly, a traditional random walk algorithm is optimized, three search directions are designed, and the limitation of the search range is solved; according to the above contents, an FRL-MFPG algorithm is designed to measure the abnormal degree of each node in the MFPG through an abnormal scoring algorithm based on correlation analysis after a micro-service fault is found, then the globality and flexibility of a random walk algorithm are optimized, and micro-service fault root cause positioning is realized. Root cause indexes such as a micro-service component, a physical machine and a KPI of a fault source are listed, so that the globality and flexibility of a search range and the accuracy of fault positioning are improved.

Description

technical field [0001] The invention belongs to the field of computer technology, and more specifically relates to a method for locating the root cause of a microservice fault based on a fault propagation graph. Background technique [0002] Compared with traditional software faults, microservice faults are more complex and difficult to locate, and different from traditional software faults, most microservice faults are related to the asynchronous call interaction of service components, heterogeneous deployment environment and configuration. First of all, there are many monitoring key performance indicators (KPI: Key Performance Indicators) of microservices, and they change with business updates. It is difficult for operation and maintenance personnel to manually set the abnormal thresholds of various indicators; secondly, the relationship between microservice fault propagation is complicated , O&M personnel use link tracking tools to describe the fault propagation relations...

AI Technical Summary

Problems solved by technology

First of all, there are many monitoring key performance indicators (KPI: Key Performance Indicators) of microservices, and they change with business updates. It is difficult for operation and maintenance personnel to manually set the abnormal thresholds of various indicators; secondly, the relationship between microservice fault propagation is complicated , the operation and maintenance personnel use the link tracking tool to describe the fault propagation relationship is not complete

Some existing research methods have certain deficiencies in inferring the root cause of microservice failures based on the dependency call relationship between services and the similarity of service nodes. They do not fully consider the failure correlation of microservice historical failure data and cannot effectively predict the failure. Scope of influence; Finally, in the troubleshooting stage of microservices, operation and maintenance personnel need to rely on operation and maintenance experience and a lot of time to troubleshoot, locate, and analyze the root causes of failures one by one. The operation and maintenance costs are large and the efficiency is not high. unacceptable

Images