Workload Mix Test Case Generation Using Queuing Curve Analysis

Solution Overview

Problem

It is challenging to generate effective test cases for information handling systems that accommodate varying and dynamic workloads across different user systems and industries, as existing methods fail to account for the unique characteristics and requirements of these systems.

Innovation Solution

A method and system for generating test cases that involve receiving workload data, dividing it into bins based on characteristics, clustering the data, and determining workload mixes using queuing theory and K-means clustering, to create tailored test cases for target information handling systems.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If workload data is divided into multiple bins and clustered using queuing theory and K-means clustering, then test case accuracy and adaptability are improved, but system complexity and processing time increase

Engineering Contradiction:

Improvetest case accuracyVSAvoidsystem complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments workload data into multiple workload data bins based on different workload characteristics (e.g., I/O intensity, compute intensity, memory intensity). This segmentation allows the system to analyze and generate test cases for specific workload types independently, improving accuracy without requiring the entire dataset to be processed simultaneously, thus managing complexity through modular processing.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary clustering of workload data into representative patterns before actual test case generation. By pre-processing the workload data to identify common patterns and characteristics, the system reduces the complexity of subsequent test case generation while maintaining high accuracy. The clustering results are stored and reused for generating test cases across multiple targets.

Inventive Principle:

Principle #10Preliminary action

2Adaptability or versatility

If workload data is collected from multiple information handling systems across different industries, then test case versatility and coverage are improved, but data processing complexity and time requirements increase

Engineering Contradiction:

Improvetest case versatilityVSAvoiddata processing time

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent performs preliminary clustering of workload data into representative patterns before actual test case generation. By pre-processing the workload data to identify common patterns and characteristics, the system reduces the complexity of subsequent test case generation while maintaining high accuracy. The clustering results are stored and reused for generating test cases across multiple targets.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent creates representative workload patterns through clustering that can be copied and applied to generate test cases for multiple target systems. Instead of processing raw workload data from each individual system separately, the system creates abstracted workload models that capture essential characteristics and can be reused across different targets, significantly reducing processing time while maintaining versatility.

Inventive Principle:

Principle #26Copying

3Manufacturing precision

If clustering algorithms like K-means are applied to workload data bins, then workload mix accuracy is improved, but computational requirements and processing complexity increase

Engineering Contradiction:

Improveworkload mix accuracyVSAvoidcomputational energy

Core Design Contradiction:

Manufacturing precisionVSUse of energy by moving object

Solution Approach 1:

The patent segments workload data into multiple workload data bins based on different workload characteristics (e.g., I/O intensity, compute intensity, memory intensity). This segmentation allows the system to analyze and generate test cases for specific workload types independently, improving accuracy without requiring the entire dataset to be processed simultaneously, thus managing complexity through modular processing.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies clustering algorithms to selected workload data bins rather than processing all data with full complexity. By identifying and focusing on the most representative or critical workload bins for clustering, the system achieves high workload mix accuracy while reducing overall computational energy requirements compared to applying clustering to the entire dataset.

Inventive Principle:

Principle #16Partial or excessive action

Data Source

PatentUS12481575B2System and method for automated test case generation based on queuing curve analysis

Publication Date: 2025.11.25 DELL PROD LP

AI summary

A system for generating test cases with workload mixes for a set of target information handling systems includes gathering workload data from a plurality of information handling systems, dividing the workload data into a plurality of workload data bins, identifying workload data characteristics for each workload data bin and identifying workload data sets that may be applicable to a set of target information handling systems. A workload mix may be determined based on workload characteristics of the set of target information handling systems. Real customer workload data including real-time or near real-time workload data may be used to check a test case for accuracy before deploying a test case to a target information handling system.