A two-stage flow measurement method for software-defined networks

Abstract

The present invention proposes a software-defined network-oriented two-stage large flow measurement method. The method combines the sFlow sampling technology with the OpenFlow flow meter measurement technology to construct a two-stage large flow measurement model. The basic idea of ​​this method is: in the first stage, coarse-grained screening of suspected large flows is based on the sFlow sampling mechanism; in the second stage, the SDN controller generates corresponding measurement rules based on the collection information of suspected large flows, and loads the measurement flow entries to the OpenFlow switch to enable large flows. Flow measurement tasks, and then read the measured flow entries to obtain the accurate flow information of these suspected large flows, so as to find out the real large flows. The invention has the advantages of eliminating the influence of the sampling rate on the measurement accuracy, reducing the consumption of TCAM resources by the measurement, and realizing the measurement load balance between the switches, and the like.

Description

technical field [0001] The invention belongs to the field of network data communication, and in particular aims at measuring large flows in a network with high precision and low load, and establishes a two-stage measurement mechanism of comprehensive sFlow packet sampling measurement and OpenFlow switch flow table measurement. Background technique [0002] Currently, network measurement is an important means of understanding network health and understanding network behavior. By measuring and analyzing traffic, we can effectively grasp the basic characteristics of network behavior, diagnose problems in network operation, and detect network attacks. In view of the typical heavy-tailed characteristics of the traffic distribution in the network, that is, a large number of flows are small flows, a few flows are large flows, and most of the traffic distribution is concentrated in these small number of large flows, this phenomenon is called the middle of the network. Rat Icon Mode...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to propose a software-defined network-oriented solution for the current SDN network-oriented large-flow measurement mechanism, which still has the defects that the measurement accuracy is overly dependent on the sampling rate setting, the TCAM resources are limited, it is difficult to carry out measurement on demand, and the measurement mechanism is difficult to deploy flexibly. The two-stage large flow measurement model of

Images