Spine-Leaf Data-Plane Bypass for Link Failure Recovery

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In spine-leaf architecture networks, existing leaf nodes rely on the control plane for route convergence, leading to slow convergence speeds and flow interruptions due to packet loss when a link failure occurs.

Innovation Solution

A communication method and apparatus that enables fast bypass protection switching at the data plane by using a fast recovery leaf node group, allowing sub-millisecond recovery for CPU forwarding and microsecond-scale recovery for hardware forwarding chips, by determining a third interface from a fast recovery leaf node group when a second interface fails.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If leaf nodes rely on the control plane for route convergence, then routing correctness is maintained, but convergence speed deteriorates and flow interruptions occur

Engineering Contradiction:
Improverouting correctnessVSAvoidconvergence speed
Core Design Contradiction:
ReliabilityVSSpeed

Solution Approach 1:

The patent segments the forwarding decision process into two independent parts: control plane routing computation and data plane fast bypass forwarding. The forwarding entry is segmented to include both traditional next-hop information and fast bypass tags, allowing the data plane to independently make forwarding decisions without control plane involvement during link failures.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent performs preliminary action by pre-configuring fast bypass tags in forwarding entries before link failures occur. When a link failure happens, the data plane can immediately use these pre-configured tags to switch to backup paths without waiting for control plane route convergence, achieving sub-millisecond recovery.

Inventive Principle:
Principle #10Preliminary action

2Productivity

If ECMP routing is used for load balancing, then traffic distribution is improved, but packet loss increases when a link fails

Engineering Contradiction:
Improvetraffic distributionVSAvoidpacket loss
Core Design Contradiction:
ProductivityVSLoss of energy

Solution Approach 1:

The patent implements feedback by monitoring link status and dynamically adjusting forwarding behavior. When a link failure is detected, the system feedback triggers immediate fast bypass switching at the data plane, redirecting traffic that would otherwise be lost to backup paths through pre-configured fast bypass tags.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent applies beforehand cushioning by pre-establishing backup forwarding paths with fast bypass tags in the forwarding table. These cushioning measures are prepared in advance so that when ECMP link failures occur, traffic can be immediately redirected to backup paths without packet loss, while maintaining the load balancing benefits of ECMP during normal operation.

Inventive Principle:
Principle #11Beforehand cushioning (Prior cushioning)

Data Source

PatentEP4637103A1Communication method and apparatus
Publication Date: 2025.10.22 NEW H3C TECH CO LTD
  • EP4637103A1 patent drawingFigure 1
  • EP4637103A1 patent drawingFigure 2
  • EP4637103A1 patent drawingFigure 3

AI summary

The present disclosure provides a communication method and apparatus. In an example of the present disclosure, a spine node may receive a first internet protocol (IP) packet sent by the first leaf through a first interface. The spine node may obtain a forwarding entry for forwarding the first IP packet; wherein, the forwarding entry comprises a fast bypass tag field and an egress interface field instructing a second interface for forwarding the first IP packet. The spine node may determine a third interface from a fast recovery leaf node group in response to determining that the second interface fails and the value of the fast bypass tag field is a first value, wherein the third interface is different from the second interface and is connected to a second leaf node. The spine node may send a second IP packet, which includes a destination media access control (MAC) address and the first IP packet, to the second leaf node through the third interface, so that the second leaf node may perform split-horizon forwarding for the second IP packet based on the destination MAC address.