Distributed System Root Cause Anomaly Detection Using Dependency Graphs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Determining the root cause of application performance issues in distributed systems is challenging due to increased data usage and complexity, leading to prolonged resolution times and high costs with existing tools and resources.
Innovation Solution
A system that generates a graph based on infrastructure templates, metrics, and policies to identify dependencies between resources, uses machine learning to detect anomalies, and determines root cause anomalies through traversal and contribution scoring, providing recommendations for resolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If additional tools and resources are used to pinpoint the root cause, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The system segments the complex analysis task into distinct components: graph generation from infrastructure templates, anomaly detection on time series data, chain of dependency identification, and root cause determination. This segmentation allows each component to be handled by specialized algorithms rather than requiring a single complex tool.
Solution Approach 2:
The patent introduces an intermediary graph structure that represents dependencies between resources. This graph acts as a mediator between the raw time series data and the root cause analysis, transforming the problem from directly analyzing complex data to traversing a structured dependency representation.
2Measurement precision
If additional tools and resources are used to pinpoint the root cause, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The system performs preliminary actions by generating the dependency graph from infrastructure templates before anomaly detection is needed. This pre-computed graph structure is then reused during anomaly analysis, avoiding the need to reconstruct dependencies during the actual root cause investigation.
Solution Approach 2:
The patent replaces manual or complex mechanical analysis processes with automated computational algorithms. The system automatically traverses the dependency graph and identifies root causes using algorithmic approaches, substituting human analysis time with automated processing.
3Device complexity
If manual analysis methods are used, then device complexity is reduced, but productivity decreases
Solution Approach 1:
The system enables self-service by automatically performing the complete root cause analysis workflow. The anomaly analytics service autonomously generates graphs, detects anomalies, identifies chains of dependency, and determines root causes without requiring manual intervention at each step, thereby maintaining simplicity while improving productivity.
4Measurement precision
If comprehensive data analysis is performed, then measurement precision is improved, but use of energy increases
Solution Approach 1:
The system extracts only the relevant information needed for root cause analysis from the comprehensive time series data. By focusing on anomaly detection and dependency chain identification rather than analyzing all data points, the system reduces computing resource consumption while maintaining detection accuracy.
Data Source
AI summary
A system generates a recommendation that includes at least one action to address at least one root cause anomaly that causes other anomalies occurred within a distributed system. The at least one root cause anomaly is determined by at least using a graph that represents the distributed system and metrics that are associated with the distributed system.