Automated Virtual Machine Image Rollback via Problem Signature Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems lack an automated method to determine the appropriate prior virtual machine (VM) image to revert to when a VM is behaving poorly, relying on human-driven analysis and prone to error, as they do not evaluate the underlying causes of the issue.
Innovation Solution
An automated approach that analyzes prior VM images by creating a problem signature based on resource metrics, comparing it with historic signatures to select the 'best' prior VM image that does not exhibit the current problem, using a historical pattern matching system to identify and apply the most suitable VM image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If automated analysis of prior VM images is performed using problem signatures and resource metrics, then the accuracy and reliability of VM image selection is improved, but the complexity of the system increases
Solution Approach 1:
The system performs preliminary actions by gathering resource metrics and generating problem signatures for prior VM images before a rollback is needed. This advance preparation creates a historical database that enables automated, accurate selection without requiring complex real-time analysis during the rollback process.
Solution Approach 2:
The patent introduces problem signatures as an intermediary that mediates between raw resource metrics and VM image selection decisions. These signatures simplify the comparison process by transforming complex metric data into comparable problem characteristics, reducing system complexity while maintaining reliability.
2Productivity
If manual analysis and selection of prior VM images is performed, then the system complexity is reduced, but the time required for rollback and productivity is reduced
Solution Approach 1:
The system performs self-service by automatically analyzing resource metrics, generating problem signatures, and selecting the appropriate prior VM image without human intervention. This automation eliminates manual analysis time and accelerates the rollback process while maintaining systematic decision-making.
Solution Approach 2:
The patent replaces the mechanical process of manual analysis with an automated computational system that uses resource metrics and problem signatures. This substitution eliminates human-driven analysis time and enables faster, more consistent VM image selection.
3Ease of operation
If rollback to the most recent prior VM image is performed without evaluation, then the ease of operation is improved, but the reliability of resolving the underlying issue is reduced
Solution Approach 1:
The system uses feedback from historical resource metrics and problem signatures to inform the rollback decision. By comparing the current problem signature with historical data, the system identifies prior VM images that did not exhibit the same problems, ensuring that the rollback addresses the underlying issue rather than simply returning to the most recent version.
4Measurement precision
If comprehensive analysis of resource metrics is performed on all prior VM images, then the measurement precision of problem identification is improved, but the time and computational resources required increase
Solution Approach 1:
The patent extracts only the essential characteristics of VM performance into problem signatures from comprehensive resource metrics. This extraction process maintains measurement precision by focusing on the key indicators of problems while eliminating unnecessary detailed analysis, thereby reducing the time and computational resources required.
Data Source
AI summary
An approach is provided to apply a virtual machine (VM) image to a computer system. In the approach, implemented by an information handling system, a detection is made that a current VM image executing on the computer system is experiencing a problem. In response, prior VM images are analyzed, with each of the prior VM images being an image that was previously executed on the computer system. Based on the analysis, one of the prior VM images is selected and the selected image is used to replace the current VM image on the computer system.