NFV Service Chaining via BGP Load Balancing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Data center networks face challenges in efficiently routing traffic between distributed Network Function Virtualized (NFV) applications due to impedance mismatches and oversubscribed links, leading to potential bottlenecks and inefficient use of resources, with existing solutions lacking a universal mechanism for dynamic load balancing and efficient connectivity between NFV-based components.

Innovation Solution

A system and method for chaining distributed NFV applications using Border Gateway Protocol (BGP) exchanges to establish connectivity, signal load distribution, and share application details, with data plane load balancers embedded in egress points to manage traffic and optimize routing between NFV applications, leveraging BGP/NSH interconnection and other routing techniques to create efficient service chains.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If distributed NFV applications are chained in a data center network, then service functionality and resource utilization are improved, but link overloads and bottlenecks occur due to impedance mismatches and oversubscribed links

Engineering Contradiction:
Improveresource utilizationVSAvoidlink overload risk
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent implements dynamic load balancing by enabling NFV applications to signal their load distribution requirements to previous applications in the service chain through BGP exchanges. This allows the system to adapt traffic routing in real-time based on current application states, preventing link overloads while maximizing resource utilization across the distributed NFV infrastructure.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent introduces BGP (Border Gateway Protocol) as an intermediary mechanism between distributed NFV applications. BGP exchanges enable applications to share connectivity information, ingress/egress point details, and load distribution requirements, facilitating coordinated traffic management without requiring direct peer-to-peer communication between all application pairs.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Adaptability or versatility

If distributed NFV applications are chained to provide diverse services, then service availability and functionality are improved, but network complexity increases due to impedance mismatches and routing challenges

Engineering Contradiction:
Improveservice diversityVSAvoidnetwork complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent applies a universal BGP-based signaling mechanism that works across all NFV applications regardless of their specific function or type. This single standardized approach handles diverse service chains, load balancing requirements, and connectivity scenarios, reducing network complexity by eliminating the need for application-specific routing protocols or custom signaling mechanisms.

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

Solution Approach 2:

The patent enables NFV applications to dynamically signal changes in their operational parameters (such as load capacity, ingress/egress points, and capabilities) through BGP exchanges. This allows the network to automatically adapt to changing service requirements and application states without manual reconfiguration, managing complexity through automated parameter propagation.

Inventive Principle:
Principle #35Parameter changes

3Productivity

If BGP exchanges are used for connectivity signaling between NFV applications, then routing efficiency and load balancing are improved, but information exchange overhead increases

Engineering Contradiction:
Improverouting efficiencyVSAvoidinformation exchange overhead
Core Design Contradiction:
ProductivityVSQuantity of substance

Solution Approach 1:

The patent implements preliminary action by having NFV applications pre-signal their connectivity requirements, capabilities, and load distribution preferences through BGP exchanges before traffic flows are established. This allows the network to proactively set up optimal routing paths and load balancing configurations, reducing the need for continuous real-time signaling and minimizing information exchange overhead during operation.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3222062B1System and method to chain distributed applications in a network environment
Publication Date: 2020.01.08 CISCO TECHNOLOGY INC
  • EP3222062B1 patent drawingFigure 1A
  • EP3222062B1 patent drawingFigure 1B~1C
  • EP3222062B1 patent drawingFigure 2~3

AI summary

A method is provided in one example embodiment and may include communicating information between a plurality of network function virtualized (NFV) based applications; and creating at least one service chain using at least two of the plurality of NFV-based applications based on the information communicated between the plurality NFV based applications. In some instances, the information can be communicated using border gateway protocol (BGP) exchanges between the NFV-based applications. In some instances, the information can include at least one of: next-hop address information for one or more ingress points of a particular NFV-based application; one or more capabilities by which a particular NFV-based application can receive data on one or more ingress points; and a method by which one or more egress points of a previous NFV-based application in a particular service chain is to perform load balancing for a subsequent NFV-based application in the particular service chain.