Dynamic Virtual Machine Migration for Network-Aware Resource Allocation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Data centers often underutilize available CPU, storage, and memory resources, leading to unnecessary server deployments and increased costs, as existing resource allocation methods do not effectively consider network traffic patterns, latency, and bandwidth requirements in virtual machine placement.

Innovation Solution

A method and apparatus for dynamically distributing virtual machines across physical devices by monitoring network traffic, determining if parameters exceed predefined thresholds, and moving virtual machines to optimize placement based on network information, including application traffic patterns, network topology, and latency, using distributed or centralized resource scheduling.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of manufacture

If virtual machines are statically allocated to physical servers, then deployment is simplified, but resource utilization is low and costs increase

Engineering Contradiction:
Improvedeployment simplicityVSAvoidresource utilization
Core Design Contradiction:
Ease of manufactureVSLoss of energy

Solution Approach 1:

The patent implements dynamic virtual machine migration capabilities that allow VMs to be automatically moved between physical servers based on real-time resource utilization metrics. This dynamic approach resolves the contradiction by enabling both simplified initial deployment and continuous optimization of resource allocation, preventing both over-provisioning and under-utilization.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system continuously monitors resource utilization metrics and uses this feedback to automatically trigger migration decisions. When thresholds are exceeded or under-utilization is detected, the system responds by migrating VMs to optimize resource distribution, thus resolving the contradiction between deployment simplicity and resource efficiency.

Inventive Principle:
Principle #23Feedback

2Reliability

If more servers are deployed to ensure capacity, then service availability is improved, but costs and system complexity increase

Engineering Contradiction:
Improveservice availabilityVSAvoidsystem complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent creates a universal resource pool where physical servers can dynamically serve multiple virtual machines based on demand. This multi-functionality allows the system to maintain service availability across fewer physical servers by efficiently consolidating workloads, thereby reducing overall system complexity while preserving reliability.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system merges multiple virtual machines onto shared physical infrastructure through virtualization, allowing resources to be consolidated and dynamically allocated. This combining approach reduces the total number of physical servers needed while maintaining service availability through live migration capabilities.

Inventive Principle:
Principle #5Merging (Combining)

3Loss of energy

If virtual machines are frequently migrated to optimize resource allocation, then resource utilization improves, but network traffic and migration overhead increase

Engineering Contradiction:
Improveresource utilizationVSAvoidmigration overhead
Core Design Contradiction:
Loss of energyVSLoss of time

Solution Approach 1:

The patent implements threshold-based migration triggers that initiate migration only when resource utilization exceeds or falls below predetermined levels. This partial action approach optimizes resource utilization without triggering unnecessary migrations, thereby reducing migration overhead and network traffic while maintaining efficient resource allocation.

Inventive Principle:
Principle #16Partial or excessive action

4Device complexity

If resource allocation does not consider network traffic patterns, then allocation is simpler, but application performance deteriorates

Engineering Contradiction:
Improveallocation complexityVSAvoidapplication performance
Core Design Contradiction:
Device complexityVSProductivity

Solution Approach 1:

The patent implements network-aware resource allocation that considers local network conditions, traffic patterns, and bandwidth requirements when placing virtual machines on physical servers. This local quality approach ensures that VMs are allocated to servers based on their specific network requirements, improving application performance without requiring complex global optimization.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP2335162B1Dynamic distribution of virtual machines in a communication network
Publication Date: 2013.11.20 CISCO TECHNOLOGY INC
  • EP2335162B1 patent drawingFigure 1
  • EP2335162B1 patent drawingFigure 2
  • EP2335162B1 patent drawingFigure 3

AI summary

In one embodiment, a method generally includes monitoring network traffic associated with a first network device comprising at least one of a plurality of virtual machines and determining if a parameter exceeds a predefined threshold at the first network device. If the parameter exceeds the predefined threshold, one of the virtual machines is selected to move to a second network device, the second network device selected based on network information, and the virtual machine is moved. An apparatus for resource allocation is also disclosed.