MLAG Peer Local Routing via Shared MAC Address

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In multichassis link aggregation group (MLAG) domains, the existing methods for network data unit (NDU) routing and bridging often require excessive use of peer links, leading to bandwidth inefficiency, as each MLAG peer typically transmits NDUs to another peer for processing, even when local processing is possible.

Innovation Solution

Implementing a shared media access control (MAC) address across all MLAG peers, allowing them to perform local routing and bridging instead of relying on peer links, by publishing route reachability messages with the shared MAC address, thereby minimizing peer link usage and optimizing bandwidth for external network device transmissions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If each MLAG peer transmits NDUs to another peer for processing, then routing and bridging functionality is maintained, but peer link bandwidth is consumed excessively

Engineering Contradiction:
Improverouting and bridging functionalityVSAvoidpeer link bandwidth
Core Design Contradiction:
ReliabilityVSLoss of energy

Solution Approach 1:

Each MLAG peer performs routing and bridging processing locally on received NDUs using its own routing table and bridging database, eliminating the need to forward NDUs to the peer for processing. This self-service approach resolves the contradiction by maintaining full routing/bridging functionality while eliminating unnecessary peer link traffic.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The invention extracts the routing and bridging processing functionality from the peer link dependency and implements it locally at each MLAG peer. By taking out the processing function from the peer-to-peer transmission model, the system eliminates the bandwidth consumption problem while preserving the essential routing and bridging capabilities.

Inventive Principle:
Principle #2Taking out (Extraction)

2Device complexity

If MLAG peers use peer links for NDU transmission, then centralized processing is simplified, but bandwidth efficiency deteriorates

Engineering Contradiction:
Improveprocessing architectureVSAvoidbandwidth efficiency
Core Design Contradiction:
Device complexityVSProductivity

Solution Approach 1:

Each MLAG peer independently processes NDUs locally using its own routing table and bridging database, eliminating the need for centralized peer processing. This distributed self-service model improves bandwidth efficiency by eliminating unnecessary peer link transmissions while maintaining manageable complexity through standardized local processing logic.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The invention segments the processing functionality by implementing independent routing and bridging processing at each MLAG peer rather than centralized processing. This segmentation allows each peer to operate autonomously with its own processing tables, improving bandwidth efficiency while keeping the architecture manageable through modular, standardized processing segments.

Inventive Principle:
Principle #1Segmentation

3Adaptability or versatility

If peer links are used for all NDU processing, then processing capability is distributed, but available bandwidth for external communications is reduced

Engineering Contradiction:
Improveprocessing capability distributionVSAvoidavailable bandwidth
Core Design Contradiction:
Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

Each MLAG peer performs routing and bridging processing locally on received NDUs, eliminating the need to forward NDUs to the peer for processing. This self-service approach resolves the contradiction by maintaining full routing/bridging functionality while eliminating unnecessary peer link traffic.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The invention extracts the routing and bridging processing functionality from the peer link dependency and implements it locally at each MLAG peer. By taking out the processing function from the peer-to-peer transmission model, the system eliminates the bandwidth consumption problem while preserving the essential routing and bridging capabilities.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentUS11296979B2Method and system for symmetric integrated routing and bridging
Publication Date: 2022.04.05 ARISTA NETWORKS INC
  • US11296979B2 patent drawing
  • US11296979B2 patent drawing
  • US11296979B2 patent drawing

AI summary

In general, embodiments of the invention relate to managing the processing of network data units (NDUs) received by a network device. More specifically, embodiments of the invention relate to minimize the use of a peer link between two multichassis link aggregation group (MLAG) peers to transmit NDUs that are to be routed or bridged by the MLAG peers. The aforementioned minimization of the use of the peer link may be achieved, e.g., using a shared media access control (MAC) address.