Transparent Virtual Network Function Migration

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Scaling virtual networks to form large networks is problematic due to the decoupling of control and data planes in software-defined networking, leading to inefficiencies and bottlenecks, particularly with stateful middlebox functions.

Innovation Solution

A method and system for transparent migration of virtual network functions, where data processing hardware identifies and migrates network flows from middleboxes to end hosts based on flow statistics and offload rules, using techniques like consistent hashing and connection tracking to maintain network performance and scalability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If network flow is processed through middlebox functions in software-defined networks, then network control and management is simplified, but network latency increases and bottlenecks occur

Engineering Contradiction:
Improvenetwork controlVSAvoidnetwork latency
Core Design Contradiction:
Ease of operationVSLoss of time

Solution Approach 1:

The patent extracts middlebox functions from centralized network devices and migrates them to distributed end hosts. Flow offloading identifies eligible flows and redirects them directly to end hosts, bypassing the middlebox processing path. This extraction eliminates the latency introduced by centralized middlebox processing while maintaining the simplified network control provided by software-defined networking.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent segments network flows into two categories: flows that require middlebox processing and flows that can be offloaded to end hosts. By segmenting the traffic flow based on flow offloading criteria, the system allows eligible flows to bypass middlebox processing entirely, reducing latency for those flows while maintaining centralized control for flows that require middlebox functions.

Inventive Principle:
Principle #1Segmentation

2Quantity of substance

If virtual networks are scaled to form large networks, then network capacity increases, but management complexity and bottlenecks increase

Engineering Contradiction:
Improvenetwork capacityVSAvoidmanagement complexity
Core Design Contradiction:
Quantity of substanceVSDevice complexity

Solution Approach 1:

The patent segments network management functions by distributing flow offloading decisions to end hosts based on flow offloading criteria. This segmentation allows the network to scale by adding more end hosts without proportionally increasing centralized management complexity, as each end host independently makes offloading decisions based on received criteria.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements dynamic flow offloading where the network adapts to changing conditions by adjusting flow offloading criteria and eligibility determinations. This dynamic approach allows the network to optimize performance and manage complexity adaptively as the network scales, rather than requiring static, overly complex management structures.

Inventive Principle:
Principle #15Dynamics

3Loss of time

If network flow is migrated from middlebox to end host, then network latency is reduced, but connection stability may be affected

Engineering Contradiction:
Improvenetwork latencyVSAvoidconnection stability
Core Design Contradiction:
Loss of timeVSReliability

Solution Approach 1:

The patent implements feedback mechanisms where end hosts report flow offloading status and performance metrics back to the network controller. This feedback allows the system to monitor connection stability and adjust flow offloading criteria accordingly, ensuring that migrations to end hosts maintain reliability while achieving latency reduction goals.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent prepares for potential connection stability issues by implementing fallback mechanisms and monitoring systems before problems occur. The flow offloading criteria include conditions that prevent offloading of flows that would be unstable when migrated, and the system can revert flows to middlebox processing if stability issues arise, providing a cushion against reliability degradation.

Inventive Principle:
Principle #11Beforehand cushioning (Prior cushioning)

Data Source

PatentUS12010024B2Transparent migration of virtual network functions
Publication Date: 2024.06.11 GOOGLE LLC
  • US12010024B2 patent drawing
  • US12010024B2 patent drawing
  • US12010024B2 patent drawing

AI summary

A method includes identifying a middlebox receiving network flow and communicating with one or more backend virtual machines. The method also includes receiving flow statistics corresponding to the network flow of the middlebox and determining whether the flow statistics satisfy an offload rule. The offload rule indicates when to migrate the network flow from the middlebox to an end host. When the flow statistics satisfy the offload rule, the method also includes migrating the network flow from the middlebox to the end host.