Unified Hash Index Table for Packet Processing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Network devices face inefficiencies in processing packets due to the mutual exclusivity of different forwarding operations, such as Layer 2 and Layer 3 forwarding, which requires separate memory spaces and configurations for each operation type, leading to increased memory usage and complexity.
Innovation Solution
The introduction of Hash index tables with operation types and result tables allows for efficient packet processing by integrating multiple operation types into a single memory space, using a Hash search mode that distinguishes between operation types and resolves conflicts through key values, thereby reducing memory requirements and improving forwarding performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If separate memory spaces are allocated for each forwarding operation type (Layer 2, Layer 3, unicast, multicast), then each operation can be processed independently and reliably, but memory usage increases and device complexity increases
Solution Approach 1:
The patent merges multiple separate forwarding operation types (Layer 2 forwarding, Layer 3 forwarding, unicast forwarding, multicast forwarding) into a single unified forwarding operation that shares common memory space and processing logic. This is achieved by introducing an operation type field in the forwarding entry that specifies which operation to perform, allowing the system to handle different forwarding types without requiring separate memory allocations for each type.
Solution Approach 2:
The patent creates a universal forwarding entry structure that can handle multiple operation types through the operation type field. This single forwarding entry template serves multiple functions by dynamically selecting different forwarding behaviors (Layer 2, Layer 3, unicast, multicast) based on the operation type value, eliminating the need for separate specialized structures for each operation type.
2Ease of operation
If separate memory spaces are allocated for each forwarding operation type, then configuration management is simplified for each operation, but overall device complexity increases due to multiple configurations
Solution Approach 1:
The patent implements a universal configuration structure where a single forwarding entry contains an operation type field that can represent multiple different forwarding operations. This allows the system to manage multiple operation types through a unified configuration interface rather than requiring separate configuration procedures for Layer 2, Layer 3, unicast, and multicast forwarding, thereby reducing overall device complexity while maintaining ease of operation.
3Measurement precision
If multiple Hash index tables are created for different operation types, then packet processing accuracy is improved by reducing conflicts, but memory usage and search complexity increase
Solution Approach 1:
The patent merges multiple operation-specific Hash index tables into a single unified Hash index table. The unified table stores forwarding entries that include an operation type field, allowing the system to perform Hash searches in one table rather than searching through multiple separate tables. This reduces memory usage while maintaining processing accuracy through the inclusion of operation type information in each forwarding entry.
4Productivity
If separate Hash index tables are used for different operation types, then packet processing speed is improved by avoiding conflicts, but device complexity and memory requirements increase
Solution Approach 1:
The patent combines multiple operation type-specific Hash index tables into a single unified Hash index table that handles all forwarding operations. This unification reduces system complexity by eliminating the need for multiple separate table structures and management mechanisms, while packet processing speed is maintained through efficient Hash search algorithms that operate on the unified table with operation type filtering.
Data Source
Figure 1
Figure 2
Figure 3~4
AI summary
After receiving the packet, an operation type of an operation to be performed for a packet is identified, a key value is extracted from the packet and Hash calculation is performed for the extracted key value; a calculation result of the Hash calculation is taken as a Hash entrance, and at least one Hash index table entry is found from a local Hash index table; match operation is performed between the identified operation type and the at least one Hash index table entry; the packet is processed according to a matched Hash index table entry if the matched Hash index table entry is found.