Weighted ECMP Load Balancing Using Cumulative Capacity Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing load balancing algorithms for weighted equal cost multi-path (ECMP) groups become complex and resource-intensive when dealing with network link failures, leading to increased memory usage, power consumption, and hardware complexity.
Innovation Solution
A hardware-based weighted load balancing algorithm that eliminates the need for path replication, using cumulative data carrying capacities computed based on path weights to distribute traffic evenly across paths, with operations performed in hardware to minimize area and power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional load balancing algorithms are used to handle network link failures in multi-path groups, then traffic distribution can be maintained, but device complexity, memory usage, and power consumption increase significantly
Solution Approach 1:
The patent changes the parameter representation from individual path states to cumulative capacity thresholds. By computing cumulative data carrying capacities (thresholds) for each path based on path weights and available capacities, the system transforms the load balancing problem into a simpler threshold comparison task that is more resilient to link failures without increasing complexity
Solution Approach 2:
The patent performs preliminary computation of cumulative capacity thresholds for all paths in the multi-path group before actual load balancing decisions are needed. These pre-computed thresholds are stored and readily available for quick comparison with flow-specific values, enabling fast and simple path selection even when link failures occur
2Reliability
If path replication is used to handle network link failures, then traffic distribution is maintained, but memory usage and hardware size increase
Solution Approach 1:
The patent merges the functionality of multiple replicated paths into a single path selection mechanism based on cumulative capacity thresholds. Instead of maintaining separate replicated path entries for each possible failure scenario, the system uses a unified threshold-based approach that naturally adapts to any combination of working paths, significantly reducing memory requirements
3Adaptability or versatility
If complex algorithms are implemented to support numerous concurrent multi-path groups, then load balancing is maintained, but power consumption and hardware complexity increase
Solution Approach 1:
The patent segments the load balancing decision into two independent parts: (1) pre-computed cumulative capacity thresholds for each path, and (2) flow-specific value computation and comparison. This segmentation allows the system to handle numerous concurrent multi-path groups by processing each group independently with simple, fixed operations, minimizing power consumption per group while maintaining overall versatility
Data Source
AI summary
Techniques as described herein may be implemented to support selecting a transmission path in a multi-path network link. In an embodiment, respective cumulative data carrying capacities for selected network paths in a group of network paths defining a multi-path group used to forward network packets from a first network node to a second network node are computed. A cumulative capacity comparison value for a received network packet in a flow of network packets is computed based at least in part on a hash value used to distinguish the flow from other flows of network packets. A specific network path is selected from amongst the network paths of the multi-path group, over which to forward the received network packet from the first network node towards the second network node, based on comparing the cumulative capacity comparison value with at least a subset of the cumulative data carrying capacities.