Automated Performance Measurement via Resource Usage State Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual starting and stopping of performance measurement tools becomes cumbersome and inaccurate when running at scale in virtualization environments, where multiple virtual machines operate asynchronously, making it difficult for users to monitor and manage the tools effectively.
Innovation Solution
A tool manager automatically determines the start and stop times of performance measurement tools by analyzing usage changes in computing resources, sending commands to initiate or terminate measurements based on state changes, thereby eliminating the need for manual intervention.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple performance measurement tools are run concurrently in a virtualization environment, then measurement coverage and scale are improved, but manual operation complexity and difficulty increase significantly
Solution Approach 1:
The performance measurement tool automatically detects application execution states and controls its own operation without user intervention. The tool monitors computing resource usage patterns, identifies when applications start and stop, and automatically starts/stops measurement operations based on these detected states, making the system self-managing even at large scale
Solution Approach 2:
The patent introduces an intermediary mechanism that monitors computing resource usage as a signal to coordinate between application execution and measurement tool operation. This intermediary resource usage monitoring serves as a trigger mechanism that automatically synchronizes measurement tool start/stop events with application execution without requiring direct user coordination
2Adaptability or versatility
If virtual machines operate asynchronously, then system flexibility and independence are improved, but synchronization of measurement operations becomes difficult
Solution Approach 1:
The patent applies local quality by implementing independent resource usage monitoring and state detection for each virtual machine individually. Each VM's measurement tool operates autonomously based on that specific VM's application execution state, allowing asynchronous operation while maintaining precise local measurement synchronization without requiring global synchronization across all VMs
3Measurement precision
If user manually monitors each virtual machine, then measurement accuracy is maintained, but user time and operational burden increase
Solution Approach 1:
The performance measurement tool performs self-service by automatically monitoring computing resource usage, detecting application execution states, and controlling its own operation. This eliminates the need for user monitoring while maintaining measurement precision through automated resource usage pattern analysis and state-based trigger mechanisms
Data Source
AI summary
A method measures performance for computing entities being run concurrently. A tool manager samples usage of a computing resource associated with a computing entity. A first state is determined based on the usage. Then, the tool manager determines that an application running in the computing entity started execution of a test that is being run in the plurality of computing entities when the usage changes from the first state. Upon determining the change, the tool manager sends a start command to a performance measurement tool that is configured to measure the performance of the computing entity. The tool manager determines the computing entity is in a second state based on the usage and determines the application ended execution of the test when the usage changes from the second state. Upon determining the change, the tool manager sends a stop command to the performance measurement tool.