Virtual Machine Monitor Tree Topology for Distributed VM Management

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for managing virtual machines (VMs) in distributed computing environments are inefficient and time-consuming, particularly as the number and size of VMs increase, leading to suboptimal resource utilization and management challenges.

Innovation Solution

The implementation of a logical tree topology within a distributed computing environment, where a virtual machine monitor (VMM) manager assigns VMMs to a tree structure, enabling collective operations such as broadcast, scatter, gather, allgather, and reduce operations to efficiently manage VMs, including pausing, transferring, and duplicating them across hosts.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional VM management techniques are used, then VM management operations can be performed, but the operations are time-consuming and inefficient, especially as the number and size of VMs increase

Engineering Contradiction:
ImproveVM management efficiencyVSAvoidTime for management operations
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The system segments the distributed computing environment into a hierarchical tree topology with root VMMs and leaf VMMs. This segmentation allows management operations to be performed in a structured manner, where root VMMs coordinate operations and leaf VMMs execute them locally, significantly improving management efficiency and reducing operation time compared to traditional flat management approaches.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements preliminary action by having the VMM manager pre-assign VMMs to specific roles (root or leaf) and pre-establish the tree topology structure. This preliminary organization enables faster execution of management operations like migration and checkpointing, as the hierarchical relationships and communication paths are already established before operations begin.

Inventive Principle:
Principle #10Preliminary action

2Productivity

If the number of VMs increases to utilize more resources, then resource utilization improves, but the complexity of managing these VMs increases

Engineering Contradiction:
ImproveResource utilizationVSAvoidManagement complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The tree topology segments the large number of VMMs into hierarchical groups managed by root VMMs. Each root VMM manages a subset of leaf VMMs, breaking down the complexity of managing numerous VMs into smaller, more manageable units. This segmentation allows the system to scale to more VMs while maintaining manageable complexity through the hierarchical structure.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Root VMMs act as intermediaries between the VMM manager and leaf VMMs. They receive management operations from the VMM manager and coordinate their execution across multiple leaf VMMs, reducing the direct management burden on the central VMM manager and simplifying the overall system complexity while enabling support for a larger number of VMs.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Productivity

If collective operations are implemented across all VMMs, then management efficiency improves, but the coordination overhead among VMMs increases

Engineering Contradiction:
ImproveManagement efficiencyVSAvoidCoordination overhead
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The tree topology segments coordination responsibilities by assigning root VMMs to manage specific subsets of leaf VMMs. This segmentation reduces coordination overhead compared to all-to-all communication, as each root VMM only needs to coordinate with its assigned leaf VMMs rather than all VMMs in the system, while still enabling efficient collective operations within each subtree.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentUS10255097B2Administering virtual machines in a distributed computing environment
Publication Date: 2019.04.09 INTERNATIONAL BUSINESS MACHINE CORPORATION
  • US10255097B2 patent drawing
  • US10255097B2 patent drawing
  • US10255097B2 patent drawing

AI summary

In a distributed computing environment that includes hosts that execute a VMM, where each VMM supports execution of one or more VMs, administering VMs may include: assigning, by a VMM manager, the VMMs of the distributed computing environment to a logical tree topology, including assigning one of the VMMs as a root VMM of the tree topology; and executing, amongst the VMMs of the tree topology, a broadcast operation, including: pausing, by the root VMM, execution of one or more VMs supported by the root VMM; sending, by the root VMM, to other VMMs in the tree topology, a message indicating a pending transfer of the paused VMs; and transferring the paused VMs from the root VMM to the other VMMs.