Font Size: a A A

Design And Implementation Of Call Chain Management System For Microservice Environment

Posted on:2024-05-16Degree:MasterType:Thesis
Country:ChinaCandidate:M M WeiFull Text:PDF
GTID:2568306944961389Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
In recent years,the emergence and wide application of business systems based on microservice architecture has brought great convenience to the construction of business systems,but also brought great challenges to the operation and maintenance of business systems.The call chain management system is mainly used to manage the call chain of microservices that realize various functions of the business system,and is an important support tool for the operation and maintenance of the business system based on the microservice architecture.At the same time,log-based non-intrusive call chain generation and call chain-based anomaly detection have become one of the research hotspots.This paper designs and implements a call chain management system for microservice environment.Based on the collection and storage of observable data,the system realizes the functions of call chain generation and anomaly detection based on call chain.The collection and storage scheme of heterogeneous data is designed to provide data support for call chain generation and anomaly detection.The system provides an exception detection function based on call chain to improve O&M efficiency.Meanwhile,This paper proposes a Time-aware Call Chain Generation method and Anomaly Detection method based on Call chain and Metrics fusion,TCCG-CMAD).Firstly,a log-based call chain generation method is proposed to realize non-intrusive call chain generation.Secondly,based on the fusion analysis of call chain and metric data,the system anomaly detection is realized,and the graph model is used to represent and learn complex microservice system in a unified manner.Finally,the detection method of multi-types of anomalies is designed.A series of experiments showed that this method improved recall rates and F1 scores compared to baseline methods such as DeepTraLog and DeepLog.This paper firstly introduces the research background of call chain management system for microservice environment.Secondly,combined with the investigation and analysis of the existing monitoring system and anomaly detection system,the call chain management system for microservice environment is proposed.Then,a time-aware call chain generation method and an anomaly detection method based on the fusion of call chain and metric are proposed.Then the design and implementation of the call chain management system for microservice environment are introduced in detail.Finally,a series of tests verify the effectiveness of the system.
Keywords/Search Tags:microservice, call chain, anomaly detection, graph neural network, data fusion
PDF Full Text Request
Related items