Policy-Based Virtual Resource Allocation for Host Pools

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In virtual system environments, efficient and dynamic allocation of virtual resource adapters, such as I/O adapters, is challenging due to over-commitment of physical resources and frequent activation/deactivation of virtual machines, leading to suboptimal resource utilization and potential failures in meeting availability and capacity requirements.

Innovation Solution

A policy-based method for virtual resource allocation, where a resource manager identifies eligible host resource pools and allocates resources based on predefined policies, ensuring maximum availability, capacity, and utilization, while adjusting allocations to meet changing virtual machine needs.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If virtual machines are activated and deactivated with high frequency, then system adaptability is improved, but resource allocation complexity increases

Engineering Contradiction:
Improvesystem adaptabilityVSAvoidresource allocation complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent implements dynamic resource allocation where the resource manager continuously monitors virtual machine states and automatically reassigns adapter resources based on current demands. This dynamic approach allows the system to adapt to frequent VM activations and deactivations without manual intervention, resolving the contradiction between system adaptability and allocation complexity.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The resource manager autonomously performs resource identification, eligibility determination, and allocation without requiring manual administrative actions. The system self-adjusts to VM state changes by automatically detecting available resources and reassigning them according to defined policies, thereby maintaining adaptability while reducing operational complexity.

Inventive Principle:
Principle #25Self-service

2Productivity

If physical resources are over-committed, then resource utilization efficiency is improved, but system reliability deteriorates

Engineering Contradiction:
Improveresource utilization efficiencyVSAvoidsystem reliability
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The system performs preliminary identification and eligibility assessment of host resources before allocation decisions are made. By pre-evaluating resource availability and suitability based on VM requirements, the system ensures that over-committed resources can be reliably reassigned when needed, maintaining both high utilization and system reliability.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The resource manager implements continuous monitoring of resource allocation states and VM performance. This feedback mechanism allows the system to detect when over-committed resources need reassignment and automatically adjusts allocations to maintain reliability while preserving high utilization efficiency through intelligent resource sharing.

Inventive Principle:
Principle #23Feedback

3Adaptability or versatility

If adapter resources are frequently assigned and released, then virtual machine flexibility is improved, but resource allocation time increases

Engineering Contradiction:
Improvevirtual machine flexibilityVSAvoidresource allocation time
Core Design Contradiction:
Adaptability or versatilityVSLoss of time

Solution Approach 1:

The system pre-identifies eligible host adapter resources and maintains readiness information before allocation is actually needed. This preliminary preparation allows rapid reassignment when VMs are activated or deactivated, reducing allocation time while maintaining the flexibility to serve multiple VMs dynamically.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The resource allocation system operates dynamically, automatically detecting VM state changes and triggering reassignment only when necessary. This event-driven approach minimizes unnecessary allocation operations, reducing the time loss associated with frequent assignments and releases while preserving VM flexibility through on-demand resource availability.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentUS9946567B2Policy based virtual resource allocation and allocation adjustment
Publication Date: 2018.04.17 INTERNATIONAL BUSINESS MACHINE CORPORATION
  • US9946567B2 patent drawing
  • US9946567B2 patent drawing
  • US9946567B2 patent drawing

AI summary

A computer-implemented method, carried out by one or more processors, for policy based virtual resource allocation. In an embodiment, the method includes identifying a number of host resources specified by host resource requirements for a first resource consumer. The method determines if the host resource requirements include a list of host resource pools for the first resource consumer. Responsive to determining that the host resource requirements include the list of host resource pools for the first resource consumer, a first set of eligible host resource pools is identified. An allocation policy may be identified, where the allocation policy includes one or more parameters for allocating host resources. Host resources from the first set of eligible host resource pools are allocated based on the allocation policy.