Overlay Network Partitioning for VM Reachability Updates

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Large computer networks with numerous physical and virtualized servers face scalability issues, leading to increased operational complexity and potential service interruptions due to time-consuming address updates of virtual machines during relocation, which can result in unreliable communications and system failures.

Innovation Solution

The solution involves partitioning the overlay network into manageable overlay partitions, each independently managing and updating reachability information, allowing for rapid address updates and reducing the risk of communication interruptions by using Local Reachability Access Gateways (LRAGs) to populate and update virtual routing tables in a partition-by-partition basis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If the number of virtual machines is increased to improve network capacity, then network scalability is improved, but address updating time increases and communication reliability deteriorates

Engineering Contradiction:

Improvenetwork capacityVSAvoidcommunication reliability

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The overlay network is divided into multiple overlay partitions, each managed independently by a LRAG. This segmentation allows address updates to occur in smaller, manageable groups rather than across the entire network, reducing update time and maintaining communication reliability even as the total number of virtual machines increases.

Inventive Principle:

Principle #1Segmentation

2Productivity

If the number of virtual machines is increased to improve network capacity, then network scalability is improved, but operational complexity increases

Engineering Contradiction:

Improvenetwork capacityVSAvoidoperational complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

By dividing the overlay network into multiple overlay partitions managed by separate LRAGs, the operational complexity is distributed across multiple independent management entities. Each LRAG handles only its partition, making the system manageable even as the total number of virtual machines grows to millions or tens of millions.

Inventive Principle:

Principle #1Segmentation

3Reliability

If address updates are performed across the entire overlay network to ensure reachability, then communication reliability is improved, but update time increases causing service interruptions

Engineering Contradiction:

Improvecommunication reliabilityVSAvoidupdate time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

Address updates are performed independently within each overlay partition rather than across the entire network. This segmentation reduces update time significantly while maintaining communication reliability within each partition, preventing service interruptions even as the network scales to millions of virtual machines.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentEP3210111B1Method for reachability management in computer networks

Publication Date: 2019.11.20 MICROSOFT TECHNOLOGY LICENSING LLC

Figure 1
Figure 2
Figure 3

AI summary

Various techniques for partitioning an overlay network is disclosed herein. In certain embodiments, an overlay network can be partitioned into overlay partitions with manageable sizes. Each overlay partition can independently manage and update reachability information only for end points that belong to a virtual network with at least one end point in the overlay partition. Thus, each overlay partition can operate independently from others to achieve fast reachability updating for relocated virtual machines or other end points.