Storage System Performance Degradation Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In digital storage systems, identifying the cause of performance degradation is cumbersome due to the large amounts of performance-related data and complex relationships between components, requiring manual effort and expertise to sift through anomalies.
Innovation Solution
A method that records performance metric data and uses statistical criteria to locate outliers, calculating a problem contribution probability value based on configuration and performance metric dependency trees, temporal differences, and constants to rank potential causes, thereby guiding the investigation efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual analysis of performance data is performed, then identification accuracy may be maintained, but time consumption and operational complexity increase significantly
Solution Approach 1:
The patent introduces an intermediary system comprising automated algorithms, statistical analysis tools, and processing software that mediate between raw performance data and human analysts. This intermediary layer filters, processes, and prioritizes data automatically, maintaining identification accuracy while dramatically reducing the time analysts need to spend manually examining performance metrics
Solution Approach 2:
The patent replaces the mechanical manual analysis process with automated computational systems. Algorithms automatically detect anomalies, calculate performance deviations, and generate prioritized lists of potential issues, substituting human manual sifting of data with automated mechanical processing that maintains precision while reducing time consumption
2Reliability
If comprehensive performance monitoring is implemented, then system reliability improves, but data complexity and analysis difficulty increase
Solution Approach 1:
The patent segments the complex performance monitoring system into modular components: data collection modules, processing modules, analysis modules, and reporting modules. Each component handles specific aspects of performance data independently, making the overall complex system manageable through structured division while maintaining comprehensive monitoring capabilities
Solution Approach 2:
The patent applies different analysis methods and processing techniques to different types of performance data based on their specific characteristics. Critical performance indicators receive more intensive analysis while less critical data receives streamlined processing, optimizing the balance between comprehensive monitoring and manageable complexity
3Productivity
If automated analysis algorithms are used, then productivity increases, but system complexity increases
Solution Approach 1:
The patent implements self-service capabilities where the automated analysis system automatically configures itself, adapts to changing conditions, and requires minimal human intervention for maintenance. The system performs self-diagnosis, automatic parameter tuning, and adaptive learning, increasing productivity while keeping operational complexity low through autonomous functionality
Data Source
AI summary
A tool for improving identification of one or more storage system elements responsible for a performance degradation in a digital storage system. The tool records performance metric data for the one or more storage system elements in a database. The tool locates one or more outliers in the performance metric data for a focus time window using a statistical criterion. The tool calculates a problem contribution probability value for at least one of the one or more outliers. The tool determines a sequential list of outliers, wherein an order in the sequential list of outliers is determined using the problem contribution probability value.