Network data packet aggregation method, system and device

Abstract

The invention provides a network data packet aggregation method which comprises that whether to-be-compressed data packets are enough in number is judged by a central processing unit (CPU), if the to-be-compressed data packets are enough in number, when the CPU detects that the number of the aggregated to-be-compressed data packets reaches a first preset number, the to-be-compressed data packets of the first preset number are aggregated to be data blocks to be loaded into a request queue, otherwise, the CPU detects data compression efficiency of an acceleration card, when the data compression efficiency accords with a preset condition, the to-be-compressed data packets aggregated during the period from a last aggregation end point to current time are aggregated into the data blocks to be loaded into the request queue, finally, the data blocks of the request queue are read in a compression engine by the acceleration card to be compressed and then the generated data packets are loaded into a response queue, and the CPU obtains compressed data packets in the response queue. The method can carry out self regulation according to compression efficiency of the acceleration card, improves efficiency of the acceleration card, and has timeliness and usability. The invention further discloses a network data packet aggregation system and a data packet aggregation processing device.

Description

technical field [0001] The present invention relates to the technical field of information processing, in particular to a network data packet aggregation method and system and a network data packet aggregation processing device Background technique [0002] At present, many literatures use the strategy of collecting a fixed number of data packets to gather batches of data packets, or use a pre-defined (Pre-defined) aggregation buffer size to complete an aggregation, or exceed a specified time threshold to complete an aggregation, or a combination of several strategy. In summary, these strategies either use a fixed delay strategy or a time-based maximum delay strategy to control the number of aggregated packets. [0003] In order to improve the efficiency of the coprocessor acceleration engine, in the process of network data packet compression and decompression, or encryption and decryption, it is usually used to optimize system overhead and improve the performance of the ac...

AI Technical Summary

Problems solved by technology

In the time-based aggregation delay control strategy, additional management of the delay time is required, the control is complicated, and the processing efficiency of the compression card and the buffering of the request compression queue buffer cannot be perceived. Only focusing on the time-based delay may cause CPU and Pipeling interruption between compression cards affects system throughput

Images