Workload Identification via Usage Pattern Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Computing environments typically cannot be optimized for multiple usage scenarios simultaneously, requiring manual user intervention to balance performance and efficiency, similar to automobiles with 'Sport' and 'Eco' modes.
Innovation Solution
A system that uses user-behavior data to identify and optimize user-centric workloads by detecting usage patterns across multiple software applications, generating a workload specification, and associating it with optimization triggers and profiles to automatically adjust system settings for improved performance and user experience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If computing environment is optimized for one usage scenario, then performance for that scenario is improved, but performance for other scenarios deteriorates
Solution Approach 1:
The system dynamically adjusts computing environment optimizations based on detected usage patterns. Instead of static optimization for a single scenario, the system monitors user behavior, identifies patterns of application usage, and automatically applies appropriate optimization profiles when specific usage scenarios are detected, allowing the environment to adapt between different usage scenarios over time
Solution Approach 2:
The system performs self-service by automatically detecting usage patterns and applying optimizations without manual user intervention. The workload identification system autonomously monitors application usage, generates usage patterns, identifies user-centric workloads, and triggers appropriate optimizations, eliminating the need for users to manually choose between different optimization modes
2Reliability
If manual user intervention is used to balance performance and efficiency, then optimization accuracy is improved, but user time and operational complexity increase
Solution Approach 1:
The system eliminates manual user intervention by implementing self-service automation. The workload identification system automatically collects user-behavior data, determines activity windows, generates usage patterns, identifies user-centric workloads, and applies optimizations based on detected patterns, completely removing the need for users to manually balance performance and efficiency settings
Solution Approach 2:
The system uses feedback from monitored user behavior to automatically adjust optimizations. By continuously collecting user-behavior data and analyzing usage patterns, the system receives feedback about actual usage scenarios and automatically applies appropriate optimization profiles, achieving accurate optimization without requiring user time or manual intervention
Data Source
AI summary
In an embodiment, a method includes receiving user-behavior data for a plurality of software applications. The method also includes determining activity windows for the plurality of software applications. The method also includes generating a time map of the activity windows. The method also includes detecting a usage pattern, where the usage pattern indicates two or more applications of the plurality of software applications that are used in combination. The method also includes identifying a user-centric workload from the usage pattern. The method also includes generating a user-centric workload specification for the user-centric workload, where the user-centric workload identifies the two or more applications indicated by the usage pattern. The method also includes associating the user-centric workload specification with an optimization trigger and an optimization profile. The method also includes optimizing the user-centric workload in accordance with the optimization trigger and the optimization profile.