VM Affinity Management via UUID Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In modern data centers with multiple blade servers, managing the affinity requirements of virtual machines (VMs) to ensure high availability and separate physical hardware installation poses a challenge, particularly for multi-tiered applications that need VMs on different hardware for high availability.

Innovation Solution

A data center administration server communicates with VMs to identify and manage affinity requirements by moving VMs with specific UUIDs to separate computers, utilizing a hypervisor and affinity managers to maintain high availability and separate installations.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If multiple virtual machines are consolidated on a single blade server to improve resource utilization, then productivity and resource efficiency are improved, but reliability deteriorates because high availability requirements cannot be satisfied

Engineering Contradiction:
Improveresource utilizationVSAvoidhigh availability
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The system segments virtual machines into different affinity groups based on their high availability requirements. VMs that require separation are placed in different affinity groups, which then are assigned to different physical hosts. This segmentation allows the system to maintain both high resource utilization through consolidation and high availability through strategic distribution of critical VMs across multiple physical servers.

Inventive Principle:
Principle #1Segmentation

2Reliability

If virtual machines with affinity requirements are manually distributed across separate physical hosts to ensure high availability, then reliability is improved, but device complexity and management difficulty increase

Engineering Contradiction:
Improvehigh availabilityVSAvoidmanagement complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent introduces an affinity manager as an intermediary component that automatically handles the complex task of distributing virtual machines across physical hosts according to affinity requirements. The affinity manager communicates with the virtual machine monitor and host system, translating high-level affinity requirements into concrete placement decisions. This intermediary layer shields administrators from the complexity of manual distribution while ensuring high availability requirements are met.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Ease of operation

If an automated system is implemented to manage virtual machine affinity and distribution, then ease of operation is improved, but device complexity increases due to additional management components

Engineering Contradiction:
Improveautomation levelVSAvoidsystem architecture
Core Design Contradiction:
Ease of operationVSDevice complexity

Solution Approach 1:

The affinity manager is designed as a multi-functional component that performs multiple tasks: it monitors virtual machine placement, evaluates affinity requirements, makes distribution decisions, and coordinates with both the virtual machine monitor and physical hosts. By consolidating these diverse functions into a single universal management component, the system achieves high automation levels without proportionally increasing overall system complexity.

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

Data Source

PatentUS9003007B2Administration of virtual machine affinity in a data center
Publication Date: 2015.04.07 INTERNATIONAL BUSINESS MACHINE CORPORATION
  • US9003007B2 patent drawing
  • US9003007B2 patent drawing
  • US9003007B2 patent drawing

AI summary

Administration of virtual machine affinity in a data center, where the data center includes a plurality of virtual machines (‘VMs’), each VM being a module of automated computing machinery installed upon a computer in the data center and characterized by a Universally Unique Identifier (‘UUID’), at least two of the VMs having an affinity requirement to be installed on separate computers, the data center further including a data center administration server operably coupled to the VMs, including communicating, by at least one of the VMs having an affinity requirement to the data center administration server, the UUIDs of the VMs having an affinity requirement; and moving by the data center administration server the VMs having an affinity requirement to separate computers in the data center.