Ethernet Ring Hub-and-Spoke Scalability via Localized MAC Learning

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing Ethernet ring networks face scalability issues when implementing hub-and-spoke connectivity due to the need for MAC learning on routes extending through multiple spokes, which increases the load on the hub node.

Innovation Solution

Each spoke node uses a unique Ring VLAN Identifier (RVID) to identify its traffic to the hub, and the hub node dynamically creates bindings between Customer VLAN ID (C-VID) and RVID to localize MAC learning, allowing only frames tagged with its RVID to be processed and forwarded, thereby reducing the hub's learning burden and increasing scalability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If MAC learning is performed on routes extending through multiple spokes in a hub-and-spoke Ethernet ring network, then connectivity between spokes is enabled, but the hub node's forwarding database size and processing load increase, reducing scalability

Engineering Contradiction:
Improvehub-and-spoke connectivityVSAvoidhub node forwarding database size
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent segments the Ethernet ring network into distinct hub-and-spoke domains by introducing ring blocking at spoke nodes. Each spoke node blocks certain ring directions for specific VLANs, creating segmented traffic flows that prevent MAC learning propagation across the entire ring. This segmentation isolates MAC learning to local hub-spoke segments rather than extending across all spokes, reducing the hub's forwarding database requirements while maintaining necessary connectivity.

Inventive Principle:
Principle #1Segmentation

2Adaptability or versatility

If MAC learning is performed on routes extending through multiple spokes, then spoke-to-spoke communication is enabled, but the processing load and time on the hub node increase

Engineering Contradiction:
Improvespoke-to-spoke communicationVSAvoidhub node processing time
Core Design Contradiction:
Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent implements preliminary action by pre-configuring ring blocking rules at spoke nodes before traffic flows occur. Each spoke node is pre-configured with blocking rules that prevent certain VLAN traffic from traversing specific ring segments. This preliminary configuration ensures that MAC learning is automatically limited to appropriate segments before traffic begins flowing, eliminating the need for the hub to process and learn MAC addresses from all spoke-to-spoke traffic paths, thereby reducing processing time and load.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP2417735B1Enabling an ethernet ring network to scalably support a hub-and-spoke connectivity model
Publication Date: 2017.06.28 CIENA LUXEMBOURG R L
  • EP2417735B1 patent drawingFigure 1~2
  • EP2417735B1 patent drawingFigure 3~5
  • EP2417735B1 patent drawingFigure 6~7

AI summary

A unique RVID is used for each spoke node to identify traffic flowing from that spoke node to the hub and from the hub to the spoke. Spoke nodes perform MAC learning on any frame containing their assigned unique RVID and only bridge traffic received on the ring to a client port if the traffic contains their assigned RVID. Thus, MAC learning at the spoke is localized to client routes, or to routes of interest that pass through the hub. The hub node learns C- VID/RVID - ringport bindings for traffic on the ring. When a frame is received on the ring, the hub will use the C-VID and RVID to determine the I-SID and forward the traffic onto the external network. When a frame is received from the external network, the hub node will use the I-SID & C-VID to determine the RVID of the spoke node, and then use the C-VID & RVID to determine, from its forwarding database, which ringport should be used to output the frame.