Ingress Network Processing Unit Traffic Routing via Hashing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In networks with multiple service modules, routing traffic efficiently without a full flow routing table is challenging due to resource constraints and synchronization issues, especially when ensuring that traffic in both directions of a session is routed to the same service module.
Innovation Solution
The system employs an ingress network processing unit that classifies incoming data packets based on consistent destination information, orders them using a precedence table, and hashes the Internet protocol address and destination port to determine the appropriate service module for routing, eliminating the need for a complete flow routing table at each I/O module.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a complete flow routing table is used at every I/O module to route traffic to the correct service module, then traffic routing accuracy is improved, but device complexity and resource requirements increase significantly
Solution Approach 1:
The patent extracts only the necessary routing information (destination IP address and destination port) from the complete packet header, eliminating the need to store entire flow routing tables. By taking out only the essential fields needed for routing decisions, the system achieves accurate traffic routing with significantly reduced table size and complexity.
Solution Approach 2:
The routing decision process is segmented into two parts: (1) classification of incoming traffic based on destination information, and (2) hashing of specific fields (destination IP and port) to determine the service module. This segmentation allows the system to make routing decisions without requiring a complete flow table at every I/O module.
2Reliability
If a complete flow routing table is maintained at every I/O module, then traffic routing reliability is improved, but synchronization complexity and system delay increase
Solution Approach 1:
By extracting only the essential routing fields (destination IP and port) rather than maintaining complete flow tables, the patent eliminates the synchronization burden between multiple I/O modules. Each module can independently perform routing decisions based on local packet information without requiring table synchronization.
Solution Approach 2:
Each I/O module performs self-contained routing decisions using locally available packet information and the simplified routing algorithm. This self-service approach eliminates the need for inter-module synchronization while maintaining reliable routing through deterministic hashing of destination fields.
3Measurement precision
If a complete flow routing table is used, then traffic routing accuracy is improved, but processing speed and system throughput deteriorate due to table access delay
Solution Approach 1:
The patent applies partial action by processing only the necessary fields (destination IP and port) rather than examining entire packet flows. This selective processing reduces the computational burden and table access requirements, thereby increasing processing speed while maintaining accurate routing decisions through the hashing of essential fields.
Data Source
AI summary
Systems and methods are disclosed of routing traffic in a network that include classifying incoming data packets using the content of the incoming data packets that remains relatively consistent during a network session. These systems and methods also include ordering the incoming traffic according to at least one internet protocol address comprised within the incoming traffic, hashing the at least one internet protocol address and the destination port within the incoming traffic, and determining a destination service module for the incoming traffic. Through these systems and methods, data can be routed without the need for complete flow tables.