Distributed SD-WAN Gateways for Multi-Region Edge-to-Edge Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing routing solutions for large-scale SD-WAN deployments face limitations in interconnectivity and scalability due to constraints on the number of edges that can connect to a given transit point or gateway router, leading to issues such as loss of end-to-end visibility and limited hop support, which hinder seamless connectivity across multiple geographical regions.
Innovation Solution
Implementing a distributed, disjoint gateway router model with multi-hop routing capabilities, utilizing peer-conn notifications and route reflectors to establish dynamic edge-to-edge connections across multiple regions, enabling customizable VPNs and seamless switching between redundant transit points.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a common controller model is used for route exchanges, then routing control is simplified, but the maximum supported overlay hops are limited to two, constraining hierarchical deployments
Solution Approach 1:
The network is segmented into multiple regions, each with its own route reflector and hub router. This segmentation allows hierarchical deployments with more than two overlay hops by distributing routing control across regional boundaries while maintaining simplified control within each region.
Solution Approach 2:
Hub routers are introduced as intermediaries between edge routers and route reflectors across regional boundaries. These hub routers facilitate route exchanges and peer-conn notifications between regions, enabling multi-hop connectivity while maintaining the simplified common controller model within each region.
2Reliability
If the number of edges connecting to a controller is limited, then controller performance is maintained, but manual assignment of controllers to edges is required
Solution Approach 1:
Edge routers automatically discover and connect to appropriate hub routers and route reflectors without manual assignment. The system self-organizes by having edge routers send peer-conn notifications that are propagated through hub routers to the appropriate route reflectors, eliminating the need for manual controller-to-edge assignment while maintaining controller performance limits.
3Adaptability or versatility
If underlays are used for routing, then deployment flexibility is improved, but end-to-end visibility is lost when overlay to underlay handoffs occur
Solution Approach 1:
Hub routers serve as intermediaries that maintain overlay routing information during handoffs to underlay networks. They propagate peer-conn notifications and routing information through the overlay, preserving end-to-end visibility even when packets traverse underlay infrastructure, thus combining the flexibility of underlays with the visibility of overlays.
Data Source
AI summary
Some embodiments of the invention provide a method for providing dynamic edge-to-edge support across multi-hops in an SD-WAN connecting multiple regions. The method is performed at a first route reflector for a first of the multiple regions. The method receives, from a first edge router at a first site of the first region, a first request for endpoint information associated with a second edge router at a second site of a second region. After determining that the first route reflector does not have a direct connection to the second edge router, the method identifies a next-hop hub router for reaching the second edge router. The method sends a second request to the identified next-hop hub router to request the identified next-hop hub router to forward endpoint information for the second edge router to the first edge router for use in establishing a dynamic edge-to-edge connection with the second edge router.