Hash-Routed In-Network Packet Forwarding for Consistent Aggregation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing in-network computing technologies face challenges with static path configuration that leads to poor flexibility, high maintenance costs, and security vulnerabilities, especially when dealing with large quantities of computing nodes and network changes, and they fail to utilize high-capability forwarding nodes effectively.
Innovation Solution
Implement a hash routing algorithm using in-network computing identifiers to dynamically route packets, eliminating the need for static path configuration, ensuring path consistency, and enhancing security and efficiency by utilizing remote direct memory access over converged Ethernet (ROCE) for transmission.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If static path configuration is used to ensure path consistency for in-network computing packets, then packet forwarding reliability is improved, but network flexibility and adaptability deteriorate
Solution Approach 1:
The patent applies dynamics by transitioning from static path configuration to dynamic hash routing. The forwarding node uses hash routing algorithms that dynamically determine packet paths based on real-time network conditions and packet characteristics, while maintaining path consistency through deterministic hash functions. This resolves the contradiction by making the routing system adaptive without sacrificing reliability.
Solution Approach 2:
The patent changes the routing parameter from fixed static paths to dynamic hash-based paths. By using hash functions that take packet metadata (such as source/destination addresses and port numbers) as input, the system generates consistent routing decisions that adapt to different traffic patterns while ensuring packets of the same flow follow the same path, thus maintaining path consistency with improved flexibility.
2Reliability
If static path configuration is manually set for each forwarding path, then path consistency is ensured, but configuration effort and maintenance cost increase
Solution Approach 1:
The patent implements self-service by enabling forwarding nodes to automatically generate and update routing paths using hash routing algorithms. Instead of requiring manual configuration, the system autonomously determines packet paths based on hash functions applied to packet metadata. This eliminates manual configuration effort while maintaining path consistency, as the hash function deterministically maps packet characteristics to routing decisions.
Solution Approach 2:
The patent replaces the mechanical manual configuration process with an automated computational system. Hash routing algorithms automatically calculate optimal paths based on packet characteristics, substituting human operators with algorithmic decision-making. This reduces configuration effort and maintenance costs while ensuring consistent path selection through deterministic hashing.
3Reliability
If static path configuration is used, then path consistency is maintained, but network security against attacks deteriorates
Solution Approach 1:
The patent applies dynamics by replacing fixed static paths with dynamic hash-based routing paths. This makes the network less predictable to attackers, as packet paths change based on hash computations of packet metadata rather than following predetermined static routes. Path consistency is maintained through deterministic hashing, while security is improved by eliminating the predictability that static paths provide to malicious actors.
4Reliability
If static path configuration is implemented, then path consistency is achieved, but system complexity increases
Solution Approach 1:
The patent replaces complex manual path configuration and management systems with simple hash routing algorithms. Instead of requiring sophisticated mechanisms to track and manage static paths across multiple forwarding nodes, the system uses straightforward hash functions that automatically determine routing decisions. This reduces system complexity while maintaining path consistency through deterministic hashing behavior.
Data Source
AI summary
Embodiments of this application disclose an in-network computing packet forwarding method, a forwarding node, and a computer storage medium, and belong to the field of communication technologies. In the method, a forwarding node receives a plurality of in-network computing packets; determines an in-network computing identifier of each in-network computing packet; performs, based on the in-network computing identifier of each in-network computing packet, aggregation computing on in-network computing packets that belong to a same in-network computing message in the plurality of in-network computing packets, to obtain at least one aggregated packet; and forwards each aggregated packet by using a hash routing algorithm based on an in-network computing identifier corresponding to each aggregated packet.