Whole network flow measurement method and system in data center network and packet loss detection method

A data center network and measurement method technology, applied in the field of network measurement, can solve the problems of inability to cope with the diversification of measurement applications, the inability of traffic information to reflect network status, and the inability to effectively store network traffic.

Active Publication Date: 2021-05-18
XI AN JIAOTONG UNIV +1
View PDF1 Cites 3 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, Sketch cannot effectively store all network traffic, and the loss of some traffic information caused by sampling will cause the traffic information to fail to reflect the real network status.
[0007]...

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Whole network flow measurement method and system in data center network and packet loss detection method
  • Whole network flow measurement method and system in data center network and packet loss detection method
  • Whole network flow measurement method and system in data center network and packet loss detection method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0086] This embodiment will show the processing flow of data packets in the flow information measurement method and the change of flow information in the process by way of specific examples. Such as Figure 6 As shown in , it shows when a data packet arrives, the buffer space and the flow information change in the counting table. From left to right in the figure is the order in which data packets arrive in sequence. The cache space size is set to 2, and the compression parameter b is set to 1.002.

[0087] When the first packet of stream 1 (packet size 1000 bytes) arrives, its information will be parsed. The buffer module will check whether there is a piece of traffic data that hits the data packet information, and the result of the search is that there is no cache record that meets the conditions, because there is no traffic information in the cache space at this time. Therefore, a new flow information (flow 1) will be generated based on the above-mentioned parsed data pack...

Embodiment 2

[0091] This embodiment will illustrate the detection of packet loss in the network based on the DFlow network flow measurement system.

[0092] Figure 7 An example of network topology is shown, where Si represents the i-th switch, and lines represent links between different switches. Taking the transmission of flow 1 from S1 to S2 as an example for illustration, it is assumed that the data packets of flow 1 will pass through the network along the path of S1→S4→S2.

[0093] In the traditional packet loss detection process, the controller must first obtain the transmission path of flow 1. Afterwards, the degree of loss of the flow 1 during transmission is judged according to the flow information of the flow 1 collected by the switch on the transmission path. The judgment method is: on the transmission path, compare the difference between the data packet quantity field in the flow 1 flow information collected by the directly adjacent upstream switch and the downstream switch i...

Embodiment 3

[0097] This embodiment will illustrate the detection of Heavy Hitter in the network based on the DFlow network flow measurement system.

[0098] Heavy Hitter generally refers to the situation that there is a flow whose traffic size (data size or number of packets) exceeds a preset threshold. The measurement of Heavy Hitter is a relatively common network measurement application. The main method is to read the traffic count information in the traffic collection information and compare it with the preset threshold. If the former is greater than the latter, it is considered that Heavy Hitter has occurred.

[0099] The DFlow collector supports capturing the total byte count information of the flow, so the Heavy Hitter detection tool based on the DFlow system can detect based on the data volume of the flow. Compared with many heavy hitter detection tools based only on the number of flow packets, the Heavy Hitter detection tool based on the DFlow system can analyze the throughput of ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention discloses a whole network flow measurement method and system in a data center network and a packet loss detection method. The system comprises a collector used for supporting to capture all flows flowing through a switch, and the collector comprises a data packet analysis module, a buffer module and a counting table module; the data packet analysis module is used for analyzing quintuple information and data packet size information of a received data packet; the buffer module is used for establishing and maintaining a buffer space to buffer the traffic information and maintaining the traffic information in the buffer space; the buffer module updates the traffic information in an incremental operation mode; the counting table module is used for establishing and maintaining a counting table to store the flow information and maintaining the flow information in the counting table; and when the cache space of the cache module is not spare, the traffic information in the cache space is compressed and then is updated or added to the corresponding position of the counting table. The real-time performance, the accuracy and the comprehensiveness of the flow information can be ensured, and diversified network measurement requirements are supported.

Description

technical field [0001] The invention belongs to the technical field of network measurement, and in particular relates to a method and system for measuring network-wide traffic in a data center network and a packet loss detection method. Background technique [0002] The continuous expansion of the network scale makes the network structure of the data center gradually complicated. At the same time, the user's requirements for the network service quality and the company's requirements for the refined management of network equipment are also becoming higher and higher. The traditional way of manual troubleshooting is slow in response and low in processing efficiency, and cannot adapt to these new changes and new requirements. Therefore, normalized network traffic measurement is the basic means to manage and monitor these complex data center networks. Diversified network problems, such as network black holes, packet loss, and routing errors, have brought more and more challenge...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): H04L12/26H04L29/06
CPCH04L43/04H04L43/0876H04L69/04
Inventor 齐琪唐亚哲焦利彬朱晓明胡成臣李瑞杏梁爽曹丽惠吴晨
Owner XI AN JIAOTONG UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap