Forwarding Entry Update via Pre-calculated Write Operations
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing routing systems face inefficiencies in updating forwarding entries due to the limited processing performance of line card box processors, which restricts the rate at which servers can deliver instruction information, leading to low efficiency in updating forwarding entries.
Innovation Solution
A method where the packet receive end directly writes write operation data into memory based on the write operation address in the write operation packet, eliminating the need for data calculation and allowing only memory write operations, thereby bypassing processing performance limitations and improving the rate and efficiency of forwarding entry updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the processor calculates write operation data according to routing algorithms, then the forwarding entry can be updated, but the processing performance limitation restricts the delivery rate of instruction information
Solution Approach 1:
The patent extracts the data calculation function from the packet receive end and moves it to the packet transmit end. The packet transmit end pre-calculates the write operation data according to routing algorithms, and the packet receive end only needs to perform simple memory write operations, thereby eliminating the processing performance bottleneck at the receive end and significantly improving the forwarding entry update rate
Solution Approach 2:
The packet transmit end performs preliminary calculation of write operation data before transmission. By pre-processing the routing algorithm calculations and preparing the write operation data in advance, the system eliminates the need for real-time calculation at the packet receive end, enabling high-speed memory write operations without processing bottlenecks
2Productivity
If the server delivers instruction information at a high rate, then the forwarding entry update efficiency improves, but the packet receive end's processing performance becomes a bottleneck
Solution Approach 1:
The complex routing algorithm calculation function is extracted from the packet receive end and relocated to the packet transmit end. This extraction transforms the packet receive end into a simple execution device that only performs memory write operations, thereby removing the processing performance bottleneck and enabling high-rate instruction delivery without increasing receive end complexity
Solution Approach 2:
The patent replaces the mechanical processing system (processor calculating data) with a direct memory write system. Instead of the processor performing complex calculations and generating write data, the system directly writes pre-calculated data to memory, substituting a complex computational mechanism with a simpler storage operation mechanism
Data Source
Figure 1~2
Figure 3
Figure 4~6
AI summary
This application provides a forwarding entry update method and apparatus, and pertains to the communications field. In the method, write operation information in a write operation packet sent by a packet transmit end includes write operation data and a write operation address. A packet receive end may obtain the write operation information after receiving the write operation packet, and may directly write the write operation data into a memory based on the write operation address, to update a forwarding entry. The packet receive end needs to perform only a memory write operation, without a need to calculate the write operation data. Therefore, a rate at which the packet transmit end delivers a write operation packet is not restricted by processing performance of the packet receive end, thereby effectively improving a rate of delivering a forwarding entry and efficiency in updating the forwarding entry.