Autocorrelation Test Script Generation for Load Testing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current rule-based correlation of test scripts for software applications requires significant computing resources and does not scale well, often resulting in high CPU usage and memory issues, and manual reduction of correlations is error-prone and not customizable.
Innovation Solution
Record and compare first and second executions of a test script to identify varying response values, automatically generating a new test script based on these correlations, thereby reducing manual intervention and improving scalability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If rule-based correlation of test scripts is implemented, then test script automation is achieved, but computing resource consumption increases significantly
Solution Approach 1:
The patent applies preliminary action by pre-processing and indexing correlation rules before the actual test execution. The system prepares correlation patterns, token lists, and matching criteria in advance, so that during test script correlation, the system can quickly match against pre-prepared data structures rather than performing exhaustive searches, thereby reducing real-time computing resource consumption.
Solution Approach 2:
The patent uses copying by creating simplified representations or models of correlation rules and test script structures. Instead of directly comparing entire test scripts against complex rule sets, the system creates intermediate representations (such as tokenized formats or abstracted correlation models) that enable faster comparison while preserving the essential correlation logic, thus reducing computational overhead.
2Reliability
If exhaustive rule searching is performed in all requests and responses, then complete correlation coverage is achieved, but test execution time increases to hours
Solution Approach 1:
The patent applies segmentation by dividing the test script and correlation rules into smaller, manageable segments or tokens. Instead of searching through entire test scripts as monolithic blocks, the system breaks them down into discrete units (such as individual correlation points, parameter tokens, or rule segments) that can be processed independently and in parallel, significantly reducing search time while maintaining complete coverage.
Solution Approach 2:
The patent implements partial action by performing correlation matching in stages or iterations rather than attempting to match everything at once. The system identifies and processes high-priority or most-likely correlation points first, using heuristics to focus computational effort on the most promising areas, thereby achieving sufficient correlation coverage without the exhaustive time cost of complete brute-force searching.
3Use of energy by moving object
If manual reduction of correlations is performed, then computing resource usage is reduced, but error rate increases
Solution Approach 1:
The patent applies self-service by enabling the test script correlation system to automatically adjust and optimize its own operation. The system monitors correlation results, identifies patterns in matching success rates, and autonomously tunes parameters such as rule priority, matching sensitivity, and resource allocation without requiring manual intervention, thereby maintaining high reliability while optimizing resource usage.
Solution Approach 2:
The patent implements feedback mechanisms where correlation results are continuously monitored and fed back into the system to improve future matching accuracy. The system learns from successful and unsuccessful correlation attempts, adjusts rule weights and priorities based on observed patterns, and refines its matching algorithms over time, ensuring high reliability without requiring manual error correction.
4Ease of manufacture
If fixed correlations are used, then implementation simplicity is maintained, but adaptability to customized rules decreases
Solution Approach 1:
The patent applies dynamics by making the correlation rule system adaptable and configurable rather than fixed. The system allows rules to be dynamically added, modified, or prioritized based on specific testing needs. Correlation rules can be weighted, conditional logic can be applied, and the system can adapt to different test scenarios by loading appropriate rule sets, thereby maintaining ease of implementation through a standardized framework while providing flexibility for customization.
Data Source
AI summary
A first execution of a test script is recorded. The recording of the first execution of the test script is of a first interaction between a communication device and an Application Under Test (AUT). First request and response data for the first execution of the test script is captured. A second execution of the test script is recorded. Second request and response data for the second execution of the test script is captured. The first execution of the test script and the second execution of the test script are isolated sessions. The first request and response data is compared to the second request and response data to find one or more varying response values. The one or more varying response values are searched to identify correlations. A second test script is automatically created based on the identified correlations.