Causality Service for Web Service Anomaly Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Managing and configuring web services applications is complex due to the need for logging, monitoring, and retaining account activity, often resulting in operational incidents and unusual behavior that are difficult to detect and address.
Innovation Solution
An insights and causality service using machine learning to analyze application data and API calls, identifying anomalies and providing actionable recommendations for remediation by determining causality between events and metrics, thus helping to detect operational incidents and risks proactively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning analysis is implemented to detect anomalies and determine causality, then detection precision and reliability are improved, but device complexity and computational resources increase
Solution Approach 1:
The system segments the complex anomaly detection task into multiple specialized components: an anomaly detection service that identifies unusual patterns, a causality determination service that analyzes causal relationships, and a recommendation service that provides remediation guidance. This segmentation allows each component to focus on a specific aspect of the problem, improving detection precision while managing overall system complexity through modular architecture.
Solution Approach 2:
The patent introduces an intermediary causality determination service that acts as a mediator between anomaly detection and remediation recommendation. This intermediary component analyzes the causal relationships between detected anomalies and potential root causes, bridging the gap between symptom detection and solution provision, thereby enhancing measurement precision without requiring the entire system to be fundamentally complex.
2Loss of information
If comprehensive logging and monitoring are implemented to track account activity, then information completeness is improved, but data storage requirements and processing overhead increase
Solution Approach 1:
The system extracts only the most relevant information from comprehensive logs and monitoring data using the anomaly detection service. Instead of storing and processing all account activity data, the service identifies and extracts unusual patterns and anomalies, thereby maintaining information completeness for detection purposes while significantly reducing the quantity of data that needs to be retained and processed.
Solution Approach 2:
The patent applies partial action by implementing monitoring and logging at selective levels. The system performs comprehensive analysis only when anomalies are detected, rather than continuously processing all data at full depth. This allows the system to maintain information completeness when needed while reducing average data processing overhead and storage requirements during normal operation.
Data Source
AI summary
Techniques for determining causality are described. An exemplary method includes receiving a request to determine a cause of an unhealthy system; receiving one or more of anomaly information and event information associated with the unhealthy system; evaluating the received one or more of the anomaly information and event information associated with the unhealthy system to determine there is a known causality between anomalies or events leading to the unhealthy system; and providing a causality indication for the known causality, the causality indication including an identification of a causality source and a causality target.