Distributed Virtual Routing and Switching for Data Center Traffic

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current data center networks face challenges in dynamically assigning resources to tenants while maintaining performance isolation and scalability, particularly as cloud providers expand to serve more customers, necessitating a move beyond basic VLANs and shared IP routing.

Innovation Solution

The implementation of distributed Virtual Routing and Switching (dVRS) systems, where virtual nodes are logically represented as IP infrastructure based on physical network nodes, with virtual switches linked via VLANs, allowing for control plane processing and direct traffic routing without involving a controller, enabling scalable multi-tenancy and optimized traffic steering.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If distributed Virtual Routing and Switching (dVRS) is implemented with virtual switches routing traffic directly without controller involvement, then forwarding speed and scalability are improved, but system complexity increases

Engineering Contradiction:
Improveforwarding speedVSAvoidsystem complexity
Core Design Contradiction:
SpeedVSDevice complexity

Solution Approach 1:

The system separates control plane functions (performed by controller 110) from data plane functions (performed by virtual switches 220). This segmentation allows fast local forwarding decisions at virtual switches while the controller handles only control plane processing, achieving high-speed forwarding without proportionally increasing overall system complexity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a multi-dimensional architecture where virtual switches operate in the data plane dimension with direct routing capabilities, while the controller operates in the control plane dimension. This dimensional separation enables parallel processing of control and forwarding functions, improving speed without linearly increasing complexity.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Adaptability or versatility

If basic VLANs and shared IP routing are used for connectivity, then implementation simplicity is maintained, but scalability is limited as cloud providers expand to serve more customers

Engineering Contradiction:
ImprovescalabilityVSAvoidnetwork architecture complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system implements dynamic resource allocation where virtual switches can be dynamically created, configured, and removed based on customer demands. The control plane dynamically programs virtual switches with routing information, enabling the network to adapt and scale flexibly as cloud providers expand to serve more customers without requiring static pre-configuration.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The virtual switch architecture provides universal functionality that can serve multiple purposes: L2 switching, L3 routing, VLAN segmentation, and dynamic resource allocation. This multi-functional design allows a single infrastructure to support diverse customer requirements and scale across different service types without requiring separate specialized systems for each function.

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

3Productivity

If a centralized controller is involved in all traffic forwarding decisions, then centralized control and management are achieved, but forwarding efficiency and speed are reduced

Engineering Contradiction:
Improveforwarding efficiencyVSAvoidcentralized control capability
Core Design Contradiction:
ProductivityVSEase of operation

Solution Approach 1:

The patent extracts the forwarding decision-making function from the centralized controller and places it directly in the virtual switches. The controller retains only control plane functions (programming virtual switches with routing information), while data plane forwarding is performed locally at virtual switches. This extraction eliminates the performance bottleneck of centralized forwarding while maintaining centralized control over policy and configuration.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentEP2891282B1System and method providing distributed virtual routing and switching (DVRS)
Publication Date: 2019.10.09 ALCATEL LUCENT SA
  • EP2891282B1 patent drawingFigure 1
  • EP2891282B1 patent drawingFigure 2
  • EP2891282B1 patent drawingFigure 3

AI summary

Systems, methods, architectures and/or apparatus for providing a distributed Virtual Routing and Switching (dVRS) solution at a data center (DC) by logically representing the networking resources of the data center as a plurality virtual nodes, each virtual node comprising an IP infrastructure based on a physical network node and a plurality of virtual switches, each virtual switch including a VRS instance linked to one or more VRS instances at other virtual switches via a VLAN to provide thereby a virtual service, wherein the physical network node performs control plane processing and provides to each virtual switch the specific routing information associated with the local services of that virtual switch. Thus, from a data plane perspective the virtual switches route and/or switch traffic directly to tunnels pointing to remote VRS instances without involving the Controller in the forwarding of the packets.