Method and device for obtaining data packet in network

Abstract

The invention is to provide a method and device for obtaining data packets in a network. Compared with the prior art, multiple processing cores of multi-core equipment are utilized to carry out data packet capture in a high-speed network, so that network data packages can be obtained losslessly through a quick path, the technical defects of conventional software implementation or ASIC-based implementation are solved, and the problem that a conventional common network transceiver interface cannot obtain the network data packages losslessly through quick path is solved; meanwhile, the function of carrying out distributed storage on the captured data packages is realized; and through introducing overall serial numbers, each data package is allocated with the serial number increasingly, and the data in distributed storage is stored to an independent target memorizer accurately and losslessly, so that under the condition of high-speed flow load, multi-core multiple cores and multidata memorizer are fully utilized to realize high-performance and high-reliability network acquisition, and meanwhile, further data processing can be carried out conveniently.

Description

technical field [0001] The invention relates to the communication field, in particular to a technique for acquiring data packets in a network. Background technique [0002] At present, there are three main problems in collecting network data packets in a high-speed network traffic environment: [0003] First of all, existing network acquisition software, such as wireshark and tcpdump, cannot support high-speed network traffic. This is because data packet loss will occur under high-speed network traffic, and further analysis of the network (including but not limited to service quality analysis, voice quality analysis, video quality analysis, etc.) largely depends on the collected data accuracy. The collected data packet loss rate will affect the accuracy of the network analyzer, so that the data packet loss problem under high-speed traffic becomes the primary problem restricting the reliability of the network analyzer. [0004] For the current method of sniffing and analyz...

AI Technical Summary

Problems solved by technology

The collected data packet loss rate will affect the accuracy of the network analyzer, so that the data packet loss problem under high-speed traffic becomes the primary problem restricting the reliability of the network analyzer

[0004] For the current method of sniffing and analyzing the network based on commonly used packet capture software (such as tcpdump and wireshark), there are mainly three problems as follows: a large amount of CPU cycles are wasted due to too many interrupts; A large number of system calls brought by the copy waste a lot of CPU cycles; the performance of the back-end storage has also become the bottleneck of network packet collection

In the prior art, some use multi-core processors to process data packets for data storage performance is very poor, while others require the introduction of storage devices based on specific hardware

This is due to the fact that under high-speed network traffic, it is easy to cause out-of-order problems when merging multiple collection data packets with the same timestamp

Images