A distributed control load balancing system and method based on SDN

A distributed control and load balancing technology, which is applied in the field of communication networks, can solve the problems of uneven traffic, uneven load, and distributed expansion of large-scale single-data center networks. Achieve the effect of overcoming fine-grained traffic balance, overcoming overhead, and overcoming distributed expansion

Active Publication Date: 2020-11-10
XIDIAN UNIV +1
View PDF5 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

For the load balancing scheme based on the SDN idea, there are mainly two schemes in the existing technology, namely, the distributed multi-controller load balancing scheme for large-scale multi-data center networks and the centralized single-controller load balancing scheme for small-scale single data center networks. The controller load balancing solution, a distributed multi-controller load balancing solution for large-scale multi-data center networks, solves the problem of uneven load among multi-data center networks, but does not solve the problem of fine-grained internal large-scale single data center networks The problem of uneven traffic
The centralized single-controller load balancing solution for small-scale single-data center networks solves the problem of uneven load within single-data center networks, but does not solve the problem of distributed expansion of large-scale single-data center networks, and cannot adapt to large-scale data center networks. application requirements, as well as the network overhead caused by large-scale network global control and the load problem of a centralized single controller

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
  • A distributed control load balancing system and method based on SDN
  • A distributed control load balancing system and method based on SDN
  • A distributed control load balancing system and method based on SDN

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0046] There are two main types of existing load balancing solutions, namely the distributed multi-controller load balancing solution for large-scale multi-data center networks and the centralized single-controller load balancing solution for small-scale single data center networks. The distributed multi-controller load balancing solution for data center networks solves the problem of uneven load among multi-data center networks, but does not solve the problem of uneven traffic within a large-scale single data center network caused by fine-grained granularity. The centralized single-controller load balancing solution for small-scale single-data center networks solves the problem of uneven load within the single-data center network, but does not solve the problem of distributed expansion of large-scale single-data center networks, and cannot adapt to large-scale data centers Network application requirements. In order to adapt to the application requirements of the new large-sca...

Embodiment 2

[0049] The overall composition of the SDN-based distributed control load balancing system is the same as that in Embodiment 1. The link information module in the SDN controller of the present invention is used to store the link information of the area network; the routing table module is used to store the routing information of the flow. ; The routing calculation module is used to query the information stored in the link information module and the routing table module, and calculate the path for the flow according to the queried information; for the monitoring and collection of link information in the link information module, the SDN controller will First, send LLDP (Link Discovery Protocol) data packets to all SDN switches connected to it through a Packet_out message. After receiving the message, these SDN switches will send LLDP data packets to all their ports. If an SDN switch receives These sent LLDP data packets will send the link information between the source and destina...

Embodiment 3

[0052]The overall composition of the SDN-based distributed control load balancing system is the same as in Embodiment 1-2, and the SDN controller 1 of the present invention is used to obtain all link information between the edge layer and the aggregation layer, that is, the area network 1, and to control all The SDN switch at the edge layer delivers the flow table and performs centralized control on area network 1. The SDN controller 2 of the present invention is used to obtain all link information between the convergence layer and the core layer, that is, the area network 2, and issue flow tables to all the convergence layer SDN switches and perform centralized control on the area network 2; for The functions of the SDN controllers. The two SDN controllers respectively perform link discovery, link information management, routing policy formulation, and flow table item delivery to their respective control area networks through the southbound interface protocol (OpenFlow protoco...

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 an SDN-based distributed control load balancing system and method, which mainly solve the problems of distributed expansion, network overhead, uneven load and link congestion in the existing small-scale single data center network. The system of the present invention includes a three-layer Clos underlying network and two SDN controllers. Link information, a routing table and a routing calculation module are added to the SDN controller, and the above modules act together as a data flow calculation path. The load balancing method adopts two-step path selection, divides the data flow into flow slices in the buffer of the end host, and uses two controllers to monitor and collect local link information of the two area networks respectively, and each port of the switch The queuing delay is used as the link information to optimize the routing link. The present invention adopts distributed multi-controllers and two-step path selection in a three-layer Clos network, has remarkable network balancing effect, and is used for load balancing control of a large-scale single data center network.

Description

technical field [0001] The invention belongs to the technical field of communication networks, and relates to a distributed control load balancing technology, in particular to an SDN-based distributed control load balancing system and method, which can be used in a Clos-structured data center network. Background technique [0002] In recent years, with the rapid growth of Internet services, data center network operations have become more complex. Currently, traffic management operations in a data center network are mainly divided into two strategies. One is a strategy related to user services, such as web browsing, search query, and other user services; this strategy is mainly aimed at mouse traffic, which is generated through user services, and for user-specific SLA services, mouse traffic often transmits A short duration, ie a short Flow Completion Time (FCT), is required. Another type of policy-operated traffic is generated by virtual machine migration, data backup, and...

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
Patent Type & Authority Patents(China)
IPC IPC(8): H04L12/803H04L12/709H04L12/727H04L12/951H04L45/121H04L45/243H04L47/43
CPCH04L45/121H04L45/24H04L47/125H04L49/9057
Inventor 顾华玺刘勇郭彦涛余晓杉王琨
Owner XIDIAN 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