Virtual Machine Image Analysis Using Bayesian Peer-Pressure

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In virtual machine management, it is challenging to distinguish between relevant and insignificant software settings, and to identify operational data correlated with the virtual machine's state from among numerous changing state parameters, making it difficult to determine the cause of desirable or undesirable states.

Innovation Solution

The use of Bayesian-based peer-pressure techniques to analyze virtual machine images by extracting a subset of settings, identifying similarities and differences, and statistically ranking and categorizing parameters to identify meaningful settings, thereby distinguishing between relevant and irrelevant settings.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual monitoring and analysis of virtual machine settings is performed, then understanding of individual settings is possible, but the complexity and time required increase significantly due to the large number of variables

Engineering Contradiction:
Improveidentification accuracy of relevant settingsVSAvoidcomplexity of settings analysis
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent introduces a statistical analysis system as an intermediary between the virtual machine settings and the user. This system automatically collects, analyzes, and presents settings information, eliminating the need for users to manually examine numerous variables. The intermediary processes the complex data and delivers only the relevant findings, resolving the contradiction between identification accuracy and analysis complexity.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces manual mechanical analysis of settings with automated statistical computation. Instead of users manually reviewing and comparing settings across multiple virtual machines, the system uses statistical algorithms to automatically identify patterns, correlations, and significant settings. This substitution of manual mechanical processes with automated computational methods reduces complexity while maintaining or improving identification accuracy.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Loss of information

If all configuration changes are tracked manually, then complete record of changes is obtained, but it becomes difficult to retrace steps and identify causes of state changes

Engineering Contradiction:
Improvecompleteness of configuration change recordVSAvoidtime to retrace configuration steps
Core Design Contradiction:
Loss of informationVSLoss of time

Solution Approach 1:

The patent implements feedback mechanisms where the statistical analysis system continuously monitors virtual machine settings and provides feedback about changes and their potential causes. The system analyzes settings over time and feeds back information about which changes are most likely to have caused observed state changes, based on statistical correlations. This feedback loop enables rapid identification of cause-effect relationships without requiring manual retracing of all configuration steps.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent performs preliminary statistical analysis of configuration changes and their relationships in advance. By pre-analyzing the data and establishing correlations between settings and virtual machine states, the system prepares information that enables rapid troubleshooting. When a state change occurs, the pre-computed statistical relationships allow immediate identification of likely causes without requiring time-consuming manual investigation of all previous changes.

Inventive Principle:
Principle #10Preliminary action

3Quantity of substance

If operational data is collected from virtual machines, then performance metrics are available, but it is difficult to determine which data is correlated with desirable or undesirable states

Engineering Contradiction:
Improveamount of operational data collectedVSAvoiddifficulty of identifying correlated operational data
Core Design Contradiction:
Quantity of substanceVSDifficulty of detecting and measuring

Solution Approach 1:

The patent replaces manual analysis of operational data with automated statistical computation. The system uses statistical algorithms to automatically analyze large volumes of operational data and identify correlations with virtual machine states. This computational approach can process and analyze data at scales and speeds that manual analysis cannot achieve, while systematically identifying meaningful correlations among the collected data.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent transforms operational data from its raw form into statistically analyzed parameters that reveal correlations with virtual machine states. By applying statistical transformations and analyses to the collected data, the system converts large quantities of raw operational metrics into meaningful indicators that clearly show their relationship to desirable or undesirable virtual machine states, making the correlations detectable and measurable.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP2726977B1Virtual machine image analysis
Publication Date: 2019.12.11 MICROSOFT TECHNOLOGY LICENSING LLC
  • EP2726977B1 patent drawingFigure 1
  • EP2726977B1 patent drawingFigure 2
  • EP2726977B1 patent drawingFigure 3

AI summary

Techniques for analyzing virtual machine images are described. In one embodiment, a subset of settings is extracted from one or more virtual machine images, the virtual machine images store therein values of the settings. The settings are used by software executing in virtual machines of the virtual machine images, respectively. A target one of the virtual machine images is selected and target values of the settings are obtained from the target virtual machine image. Sample values of the settings are obtained from a plurality of virtual machine images. The subset formed by identifying similarities and differences of the values between the virtual machine images.