Virtual Slot Scoring for Distributed Application Instance Allocation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current distributed systems require manual programming to determine virtual machines for application instance execution, leading to resource degradation and high acquisition costs due to the use of type 1 hypervisors, which also fail to meet real-time constraints effectively.

Innovation Solution

A method for automatically and autonomously selecting a virtual location within a virtual machine to run an application instance by calculating a score based on various constraints such as hardware resources, network reachability, application versions, and user interface availability, without requiring complex equipment or software, ensuring efficient resource allocation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If manual programming is used to determine virtual machines for application execution, then system configuration is straightforward, but resource allocation efficiency deteriorates and real-time constraints cannot be met

Engineering Contradiction:
Improvesystem configurationVSAvoidresource allocation efficiency
Core Design Contradiction:
Ease of operationVSProductivity

Solution Approach 1:

The system performs automatic selection of virtual machines for application execution based on predefined constraints and scoring criteria. The resource manager autonomously evaluates available virtual machines against application requirements and selects the optimal host without manual intervention, enabling self-service resource allocation that improves efficiency while maintaining operational simplicity

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The invention transforms manual configuration into automated parameter-based decision making. By converting selection criteria into quantifiable parameters (score calculation based on constraints) and dynamically adjusting virtual machine allocation based on real-time system state, the system achieves efficient resource distribution without compromising configuration simplicity

Inventive Principle:
Principle #35Parameter changes

2Adaptability or versatility

If type 1 hypervisor is implemented to manage virtual machines, then virtual machine execution is enabled, but system performance degrades and acquisition cost increases

Engineering Contradiction:
Improvevirtual machine execution capabilityVSAvoidsystem performance
Core Design Contradiction:
Adaptability or versatilityVSLoss of energy

Solution Approach 1:

The invention extracts the essential function of virtual machine management from complex type 1 hypervisors. By implementing a simplified resource manager that operates at the operating system level rather than requiring dedicated hardware hypervisors, the system maintains virtual machine execution capability while eliminating the performance overhead and cost associated with type 1 hypervisor infrastructure

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The system replaces expensive, complex type 1 hypervisor hardware with a software-based resource management approach. The virtual machine selection and management functions are implemented as lightweight software components that can be deployed without dedicated hardware infrastructure, reducing acquisition costs and performance degradation

Inventive Principle:
Principle #27Cheap short-living objects (Disposable)

3Adaptability or versatility

If type 1 hypervisor is used for virtual machine management, then application execution on virtual machines is achieved, but acquisition cost becomes very high

Engineering Contradiction:
Improvevirtual machine execution capabilityVSAvoidacquisition cost
Core Design Contradiction:
Adaptability or versatilityVSEase of manufacture

Solution Approach 1:

The invention replaces expensive type 1 hypervisor hardware with affordable software-based resource management. The system implements virtual machine selection and coordination functions using standard operating system tools and network protocols, eliminating the need for costly dedicated hypervisor infrastructure while maintaining full virtual machine execution capability

Inventive Principle:
Principle #27Cheap short-living objects (Disposable)

Solution Approach 2:

The system creates a software-based copy of hypervisor functionality that performs the same virtual machine management tasks without requiring expensive hardware. The resource manager replicates hypervisor coordination functions using operating system-level processes and network communication, achieving equivalent capability at lower cost

Inventive Principle:
Principle #26Copying

Data Source

PatentEP2667302B1Method for managing the starting of instances of applications on virtual machines of a distributed network
Publication Date: 2019.08.07 SCHNEIDER ELECTRIC IND SAS
  • EP2667302B1 patent drawingFigure 1
  • EP2667302B1 patent drawingFigure 2

AI summary

The method involves determining a set of free application slots not occupied by an application instance in use, and computing a score for each free virtual slot depending on an application constraint. The virtual slot having the best score is selected to start up an application instance, where the application constraint associated with the free virtual slot consists of determining whether the requested application is already open in the selected free virtual slot, and awarding a score associated with the application constraint as a function of the determination.