Software Workload Performance Scenario Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies lack a systematic method for objectively identifying representative performance scenarios in software systems, leading to subjective decision-making and incomplete evaluation due to the absence of a knowledge base for storing and retrieving performance considerations, conversion of identification tasks into multi-criteria decision-making problems, and application of operational laws to performance metrics.
Innovation Solution
An automated system that uses a knowledge base and Multi-Criteria Decision Making (MCDM) techniques to select performance scenarios based on input performance criteria, involving an input module, selection module, and output module to generate ranked performance scenarios using operational laws and MCDM methods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing technology is used for performance scenario identification, then the process can be completed, but the results are prone to subjectivity and may be incorrect or incomplete due to lack of systematic method
Solution Approach 1:
The identification process is segmented into distinct modules: input module for collecting performance considerations, knowledge base for storing operational laws and metrics, selection module for applying MCDM techniques, and output module for generating ranked scenarios. This segmentation structures the complex process into manageable components that reduce subjectivity while maintaining systematic rigor.
Solution Approach 2:
The patent transforms the identification task from a qualitative expert judgment process into a quantitative multi-criteria decision-making problem by defining specific performance metrics and their relationships through operational laws. This parameter transformation enables objective evaluation and ranking of performance scenarios.
2Reliability
If expert knowledge is used to arrive at performance scenarios, then experience can be leveraged, but the decision-making process becomes subjective and may lead to incorrect evaluation
Solution Approach 1:
The patent replaces the manual expert judgment mechanism with an automated computer-based MCDM system. The selection module automatically applies operational laws and performance metrics to evaluate and rank scenarios, substituting human subjectivity with algorithmic objectivity while maintaining the expertise embedded in the operational laws.
Solution Approach 2:
The system incorporates feedback mechanisms where the knowledge base stores and retrieves operational laws and performance metrics that guide the selection process. The ranked output scenarios can be fed back for validation, creating a closed-loop system that improves reliability through iterative refinement.
3Adaptability or versatility
If no knowledge base is used, then the system remains simple, but there is no method to store and retrieve performance considerations consisting of multiple criterions
Solution Approach 1:
The knowledge base is designed as a universal repository that stores multiple types of information: operational laws, performance metrics, their relationships, and criterion definitions. This single multi-functional knowledge base serves all modules (input, selection, output) and can handle various performance considerations, making the system adaptable to different scenarios without requiring separate storage structures.
Data Source
AI summary
The present technique is an apparatus and method evaluating software performance. The method identifies performance scenarios using a knowledge base and selects the performance scenarios from a context module using operational laws. The system analyzes performance criticality of an application workload. Furthermore, the system comprises a context module that ranks the performance criterions depending on the input and the knowledge base stores and retrieves the performance criterions using the operational laws. The knowledge base comprises the criterions and their relevant ranks based of the application context. The present technique automates classification of performance criterions into benefit and cost categories with the usage of the operational laws.