Segment Routing Fast Reroute for Container Network Resilience

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current container network implementations are slow to detect unavailability of container pods or nodes and reprogram the network, leading to inefficiencies in rerouting traffic in case of overloading or failure.

Innovation Solution

Implementing Segment Routing (SR) with fast reroute capabilities, where source SR devices encode multiple SR routes or policies to immediately reroute traffic by encoding segment lists in packet headers, allowing for rapid reconfiguration in case of unavailability.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If traditional container network implementations are used, then network configuration is simple, but detection of unavailability and rerouting speed is slow

Engineering Contradiction:
Improvererouting speedVSAvoidnetwork configuration complexity
Core Design Contradiction:
SpeedVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by pre-configuring multiple SR routes (primary and backup routes) in the network before failures occur. Segment lists are pre-computed and stored at SR devices, enabling immediate rerouting when unavailability is detected without requiring real-time path computation or network reconfiguration

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent applies segmentation by dividing the network path into discrete segments using Segment Routing technology. Each segment is represented by a Segment Identifier (SID) in a segment list, allowing independent routing decisions at each segment boundary and enabling granular control over traffic flow paths for fast reroute capabilities

Inventive Principle:
Principle #1Segmentation

2Reliability

If traditional network reprogramming is used, then network stability is maintained, but network downtime during failures is extended

Engineering Contradiction:
Improvenetwork availabilityVSAvoidnetwork downtime
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent pre-establishes backup SR routes and segment lists before failures occur. When a failure is detected, the SR device immediately switches to the pre-computed backup route without requiring real-time path computation, thereby minimizing network downtime while maintaining reliability

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent implements feedback mechanisms where SR devices continuously monitor network conditions and detect unavailability of container pods or nodes. This feedback triggers automatic rerouting decisions based on pre-configured segment lists, enabling rapid response to failures while maintaining network availability

Inventive Principle:
Principle #23Feedback

3Adaptability or versatility

If real-time rerouting is implemented, then network resilience is improved, but detection and response time requirements increase system complexity

Engineering Contradiction:
Improvenetwork resilienceVSAvoiddetection and response system complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent makes SR devices multi-functional by combining route computation, segment list management, reachability detection, and automatic rerouting capabilities in a single device. This universal approach enables real-time rerouting and improves network resilience without requiring separate specialized systems for each function

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

Data Source

PatentEP3854038B1Segment routing with fast reroute for container networking
Publication Date: 2023.12.27 CISCO TECHNOLOGY INC
  • EP3854038B1 patent drawingFigure 1
  • EP3854038B1 patent drawingFigure 2
  • EP3854038B1 patent drawingFigure 3A~3B

AI summary

Systems and methods provide for segment routing (SR) with fast reroute in a container network. An SR ingress can receive a packet from a first container destined for a container service. The ingress can generate an SR packet including a segment list comprising a first segment to a first container service host, a second segment to a second service host, and a third segment to the service. The ingress can forward the SR packet to a first SR egress corresponding to the first host using the first segment. The first egress can determine whether the first service and/or host is reachable. If so, the first egress can forward the SR packet to the first host or the packet to the service. If not, the first egress can perform a fast reroute and forward the SR packet to a second SR egress corresponding to the second host using the second segment.