Hash-Based Multi-Path Routing for Network Traffic Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing multi-path routing systems face inefficiencies in traffic distribution, leading to potential overload on alternative egress interfaces and non-uniform traffic distribution due to the lack of effective methods for routing packets to specific destination-address groups across multiple paths.
Innovation Solution
A network element with circuitry that assigns multiple egress interfaces for routing packets to a given destination-address group, utilizing hash tables to store state information and calculate hash results to efficiently forward packets based on destination addresses and flow identification, ensuring efficient distribution even under congestion.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple paths are used for routing packets to a given destination, then traffic distribution efficiency is improved, but the complexity of routing decision-making increases
Solution Approach 1:
The routing decision process is segmented into two independent stages: (1) destination-address group identification using routing tables, and (2) specific path selection using hash tables with hash results. This segmentation allows each stage to operate independently and simplifies the overall complexity while maintaining efficient multi-path traffic distribution.
Solution Approach 2:
Hash tables serve as an intermediary data structure between the routing table and the final egress interface selection. The hash tables store state information indexed by hash results, enabling fast lookup and path selection without complex real-time calculations, thus reducing decision-making complexity while improving traffic distribution efficiency.
2Reliability
If adaptive routing is implemented to re-route packets based on network state, then network reliability is improved, but the processing time and delay increase
Solution Approach 1:
State information for multiple possible paths is pre-computed and stored in hash tables before packets arrive. When a packet needs routing, the system performs a fast lookup in the pre-prepared hash tables rather than performing complex adaptive routing calculations in real-time, thus maintaining reliability while minimizing processing time.
Solution Approach 2:
The patent replaces complex real-time adaptive routing mechanisms with a simplified hash-based lookup system. Instead of dynamically evaluating network conditions and re-routing packets based on current state, the system uses pre-computed hash results to quickly determine egress interfaces, reducing processing time while maintaining reliable multi-path routing.
3Stability of the object's composition
If hash tables are used to store state information for multiple paths, then traffic distribution uniformity is improved, but the memory requirements increase
Solution Approach 1:
The hash tables store state information locally for each destination-address group and hash result combination, rather than maintaining global routing state. This localized storage approach ensures uniform traffic distribution for each destination group while minimizing overall memory requirements by only storing necessary path state information.
Solution Approach 2:
The system changes the parameter for storing routing state from comprehensive global state to localized hash result-based state. By indexing hash tables with hash results and storing only the necessary egress interface information, the system achieves uniform traffic distribution with reduced memory footprint compared to traditional routing table approaches.
Data Source
AI summary
A network element includes circuitry and one or more interfaces. The interfaces are configured to connect to a communication network. The circuitry is configured to assign multiple egress interfaces corresponding to respective different paths via the communication network for routing packets to a given destination-address group, to hold, for the given destination-address group, respective state information for each of multiple sets of hash results, to receive via an ingress interface a packet destined to the given destination-address group, to calculate a given hash result for the packet and identify a given set of hash results in which the given hash result falls, and to forward the packet via one of the multiple egress interfaces in accordance with the state information corresponding to the given destination-address group and the given set of hash results.


