Segment Routing Policy Optimization for Traffic Engineering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional traffic engineering systems require manual deployment of RSVP-TE tunnels, leading to overprovisioning and increased costs, as they lack automated solutions to efficiently mitigate failure-induced Service Level Agreement (SLA) violations such as link congestion and traffic surges.
Innovation Solution
The implementation of Segment Routing (SR) policies with real-time traffic monitoring and optimization algorithms that limit the number of Segment Identifiers (SIDs) to automatically or semi-automatically resolve constraint violations, using a scalable collection of network topology and real-time notifications, allowing for tactical traffic engineering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If manual deployment of RSVP-TE tunnels is used for tactical traffic engineering, then network operators can achieve some level of traffic optimization, but the process requires excessive manual intervention and leads to overprovisioning of network capacity
Solution Approach 1:
The system enables automated traffic engineering where the network autonomously computes and deploys Segment Routing policies based on real-time traffic monitoring and constraint violation detection, eliminating the need for manual RSVP-TE tunnel deployment while achieving tactical optimization
Solution Approach 2:
The system pre-configures Segment Routing capabilities and policies in the network infrastructure, allowing automated real-time responses to traffic conditions without manual intervention during actual traffic engineering operations
2Reliability
If overprovisioning of network capacity is implemented, then Service Level Agreement violations are minimized, but network costs increase significantly
Solution Approach 1:
The system dynamically adjusts traffic engineering policies based on real-time network conditions and detected constraint violations, allowing the network to adapt capacity allocation to actual demand rather than relying on static overprovisioning, thus maintaining SLA compliance with optimized capacity utilization
Solution Approach 2:
The system continuously monitors network traffic and detects SLA constraint violations in real-time, using this feedback to automatically trigger and adjust traffic engineering actions, enabling precise capacity management that responds to actual network needs rather than predetermined overprovisioning
3Adaptability or versatility
If full mesh of RSVP-TE tunnels is deployed for global traffic optimization, then comprehensive traffic control is achieved, but the system complexity and deployment difficulty increase dramatically
Solution Approach 1:
The system segments the network into routing domains and uses Segment Routing to provide global traffic optimization through hierarchical policy deployment rather than requiring a full mesh of RSVP-TE tunnels, reducing complexity while maintaining adaptability
Solution Approach 2:
The system implements a universal Segment Routing framework that can provide both global and local traffic optimization capabilities through a single architecture, eliminating the need for separate RSVP-TE tunnel meshes while achieving comprehensive traffic control
4Ease of operation
If manual traffic engineering processes are used, then detailed control over network paths is possible, but response time to constraint violations increases
Solution Approach 1:
The system implements real-time monitoring and automated feedback loops that detect SLA constraint violations and trigger traffic engineering actions immediately, maintaining precise path control while eliminating manual response delays
Solution Approach 2:
The system autonomously detects constraint violations and executes traffic engineering decisions without manual intervention, maintaining detailed path control precision while achieving immediate automated response to network conditions
Data Source
AI summary
In one embodiment, a method includes monitoring traffic in a Segment Routing (SR) network through a collection of a Segment Routing Demand Matrix (SRDM) at a Traffic Engineering (TE) system operating at a network device, receiving topology information for the SR network at the TE system, modeling the SR network based on the topology information and the SRDM at the TE system, identifying a violation of a constraint in the SR network at the TE system, and running an optimization algorithm for SR optimization of constraints in the SR network at the TE system, wherein the optimization comprises limiting a number of Segment Identifiers (SIDs) used in a SR policy implemented to resolve the constraint violation. An apparatus is also disclosed herein.