Forwarding Tree Computation for Network Traffic Load Balancing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Shortest Path Bridging (SPB) networks fail to efficiently distribute traffic load across all network links, leading to hot links or nodes being overloaded, and require manual configuration for traffic engineering, which is not automatic and may not achieve balanced link utilization.
Innovation Solution
Implementing a node and method that perform multiple rounds of forwarding tree computations using standard Equal Cost Tree (ECT) algorithms for convergence-sensitive traffic and opaque ECT algorithms for non-sensitive traffic, taking into account the physical topology's equal cost characteristics, with Organizationally Unique Identifiers (OUI) to compute forwarding trees that balance link utilization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If standard Equal Cost Tree (ECT) algorithms are used for forwarding tree computation, then convergence-sensitive traffic is handled properly, but traffic load is not balanced across all network links leading to hot links being overloaded
Solution Approach 1:
The patent segments traffic into convergence-sensitive and non-sensitive categories, applying different ECT algorithm rounds appropriately. It also segments the ECT computation into multiple rounds (first round for convergence-sensitive traffic, second round for load balancing), allowing each segment to be optimized independently for its specific requirements.
Solution Approach 2:
The patent changes the parameters of the ECT algorithm by introducing multiple computation rounds with different objectives. The first round uses standard ECT parameters for convergence sensitivity, while the second round modifies parameters to prioritize load balancing across links, thereby resolving the contradiction between reliability and productivity.
2Ease of operation
If manual configuration is used for traffic engineering, then traffic distribution can be controlled, but the process is not automatic and requires significant manual intervention
Solution Approach 1:
The patent implements self-service by enabling the network to automatically perform multiple rounds of ECT computations and distribute traffic load without manual intervention. The system autonomously computes forwarding trees, monitors link utilization, and adjusts traffic distribution based on network conditions, eliminating the need for manual configuration while maintaining effective traffic engineering control.
3Loss of time
If single round of forwarding tree computation is performed, then computation time is reduced, but traffic load cannot be distributed evenly across all network links
Solution Approach 1:
The patent applies periodic action by performing ECT computations in multiple sequential rounds rather than a single computation. The first round establishes baseline forwarding trees for convergence-sensitive traffic, and subsequent rounds periodically adjust the trees to optimize load balancing. This periodic multi-round approach distributes computation over time while achieving better traffic distribution efficiency than a single round could provide.
Data Source
AI summary
A node and a method are described herein for computing forwarding trees to distribute traffic in a network. In addition, a network is described herein that has a plurality of nodes interconnected to one another by a plurality of network links, and each node is configured to perform multiple rounds of forwarding tree computations to distribute traffic load on one or more of the network links to the other nodes.