Leaf-Spine Switch Fabric Rate Limiting for Congestion Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In a leaf-spine topology switch fabric, oversubscription can lead to resource wastage and congestion, as traffic destined for one output port may consume more than its fair share of resources, resulting in discarded data and reduced system efficiency.
Innovation Solution
Implementing static and dynamic rate limiting methods, where data rate measurers and limiters on leaf and spine nodes measure and control data flows to prevent excessive egress, ensuring that traffic is not needlessly transmitted and consumed within the switch fabric, using rate limits that include port speeds and burst sizes to manage data flow effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If statistical multiplexing is employed to improve system utilization, then system utilization is improved, but internal fabric congestion increases and quality of service deteriorates
Solution Approach 1:
The patent implements preliminary rate limiting at ingress ports and spine nodes before traffic reaches the destination leaf node. By measuring data rates and applying rate limits in advance, the system prevents oversubscription and fabric congestion before they occur, ensuring quality of service is maintained while still allowing statistical multiplexing to improve overall utilization.
Solution Approach 2:
The patent employs feedback mechanisms where data rates are continuously measured at ingress ports and spine nodes, and rate limits are dynamically applied based on these measurements. This feedback loop ensures that traffic is controlled to match available capacity, preventing congestion while maximizing resource utilization through statistical multiplexing.
2Loss of energy
If rate limiting is applied to prevent resource wastage, then resource efficiency is improved, but device complexity increases
Solution Approach 1:
The patent segments the rate limiting function across multiple locations in the fabric - at ingress ports on leaf nodes and at spine nodes. This segmentation distributes the complexity burden and allows each component to perform simpler rate limiting operations based on local measurements, reducing overall system complexity while preventing resource wastage.
Solution Approach 2:
The patent implements self-service rate limiting where each node measures its own data rates and applies its own rate limits autonomously. This eliminates the need for centralized control and complex coordination, reducing device complexity while effectively preventing resource wastage through local rate limiting decisions.
Data Source
AI summary
Embodiments of the present invention provide methods and system to reduce needless data traffic in leaf-spine switch fabric. In embodiments, in a static solution, data rates of data flows having a common destination port may be measured and responsive to the data flows having a combined data rate that exceeding a rate limit of the common destination port, one or more of the data flows may be limited. Embodiments may also comprise a global scheduler to provide dynamic data rate controls of traffic flows from source ports to destination ports in which to reduce the handling of data traffic that would otherwise be discarded due to oversubscription.