Change Diagnosis Scope Reduction via Pattern-Based Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for diagnosing computer system failures, particularly those with significant economic impacts, face challenges in efficiently identifying the source of failures amidst vast configuration changes, relying heavily on human experience or cumbersome rule-based tools that struggle to manage the complexity of configuration data.
Innovation Solution
A pattern-based method and apparatus that narrows down the diagnosis scope using domain-independent rules, feature extraction, normalization, and clustering to reduce irrelevant changes and identify problematic system changes, employing a data cleaning module, feature extraction and normalization module, and data clustering module to analyze change records.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If trial and error approach is used to identify root causes, then human experience and knowledge are utilized, but the process is highly dependent on administrator expertise and time-consuming
Solution Approach 1:
The system performs preliminary actions by automatically collecting and analyzing configuration change data before administrators need to diagnose issues. Change detection agents continuously monitor configuration items, capture changes, and prepare analysis-ready data structures in advance, so when a failure occurs, the diagnostic information is already prepared and available for immediate analysis.
Solution Approach 2:
The patent introduces an intermediary diagnostic system that acts as a mediator between the complex configuration data and the administrator. The system includes change detection agents, data collectors, and analysis components that translate raw configuration changes into meaningful diagnostic information, reducing the need for administrators to directly analyze overwhelming amounts of raw data.
2Extent of automation
If rule-based monitoring tools are used to assist diagnosis, then automation is provided, but it is very difficult to specify all the rules needed for complex configuration changes
Solution Approach 1:
The system changes parameters by automatically discovering configuration relationships and change patterns rather than requiring manual rule specification. The analysis component learns from historical configuration data and failure patterns, dynamically adjusting diagnostic parameters and thresholds based on the specific system being monitored, eliminating the need for administrators to manually specify complex rules for every configuration item.
Solution Approach 2:
The diagnostic system performs self-service by automatically collecting, analyzing, and interpreting configuration change data without requiring manual rule configuration. The change detection agents and analysis components work autonomously to identify problematic changes, generating diagnostic reports and recommendations without human intervention in the rule specification process.
3Loss of information
If all configuration change data is collected for analysis, then complete diagnostic information is available, but the volume of data is enormous and difficult to manage
Solution Approach 1:
The system extracts only the relevant diagnostic information from the enormous volume of configuration change data. The analysis component identifies and extracts changes that are actually related to the failure, filtering out unrelated configuration modifications. This extraction process isolates the critical few changes from the many irrelevant ones, providing complete diagnostic information about the failure cause without requiring analysis of all configuration data.
Solution Approach 2:
The patent segments the configuration data into meaningful groups and categories for targeted analysis. Instead of analyzing all configuration changes uniformly, the system divides changes by configuration item type, change frequency, and relevance to the failure, allowing administrators to focus on specific segments of data that are most likely to contain the root cause, thereby managing the volume of data effectively while maintaining diagnostic completeness.
Data Source
AI summary
The scope of the system changes to be considered for analysis for finding problematic changes is reduced in order to allow focusing on highly potential suspicious drifts caused by change sequences. The method and system includes a data cleaning module to remove irrelevant changes, a feature extraction and normalization module to extract the features of change objects, data annotation module to remove irrelevant changes based on patterns, and a clustering module to obtain groups for further analysis. Data cleaning is simplified using domain independent rules. Additional sources of change sequences are removed by application of pattern based techniques so as to narrow down problematic system changes to analyze for root cause analysis. Change error sequence and degree of temporal correlation to correlate system changes with errors, as well as change behavior patterns may be used for downsizing the diagnosis scope.