CN103986663B - Data center and data processing implemented method and the network controller - Google Patents

Data center and data processing implemented method and the network controller Download PDF

Info

Publication number
CN103986663B
CN103986663B CN201410192914.2A CN201410192914A CN103986663B CN 103986663 B CN103986663 B CN 103986663B CN 201410192914 A CN201410192914 A CN 201410192914A CN 103986663 B CN103986663 B CN 103986663B
Authority
CN
China
Prior art keywords
data
packet
network
forwarding
port
Prior art date
Application number
CN201410192914.2A
Other languages
Chinese (zh)
Other versions
CN103986663A (en
Inventor
王淑玲
张云勇
房秉毅
Original Assignee
中国联合网络通信集团有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中国联合网络通信集团有限公司 filed Critical 中国联合网络通信集团有限公司
Priority to CN201410192914.2A priority Critical patent/CN103986663B/en
Publication of CN103986663A publication Critical patent/CN103986663A/en
Application granted granted Critical
Publication of CN103986663B publication Critical patent/CN103986663B/en

Links

Abstract

本发明公开了一种数据中心及其实现数据处理的方法和网络控制器,在基于SDN的数据中心中,服务器根据自身支持的协议生成相应的数据包;接收到数据包的交换机确定数据包为所属的数据流中的第一个数据包,网络控制器根据数据包转发端口的类型,生成转发路径;数据包所属数据流中的其他数据包采用生成的转发路径进行转发。 The present invention discloses a data center and data processing implemented method and a network controller, SDN-based data center server generates the corresponding data packet according to the protocol supported by itself; the received packet to the packet switch determines data flow belongs first packet, the network controller forwards the data packet type of the port, generating a transfer path; other data packets of the packet flow relevant to the use of the generated forwarding path for forwarding. 本发明方法通过数据包转发端口的类型识别数据流的类型,规划了适合该数据流的最优转发路径,而且为不同的数据流匹配了不同的服务等级质量,避免了数据传输的瓶颈,同时减小了维护开销。 Data type identification method of the present invention by a port forwarding packet type flow plan for optimal forwarding path of the data stream, and a different service levels to match the quality of different data streams, to avoid the bottleneck of data transmission, while reduced maintenance overhead. 而且基于SDN的数据中心降低了原有数据中心复杂度。 And SDN-based data center to reduce the existing data center complexity.

Description

数据中心及其实现数据处理的方法和网络控制器 Data center and data processing implemented method and the network controller

技术领域 FIELD

[0001] 本发明涉及软件定义网络(SDN,Software Defined Network)技术,尤指一种基于SDN的数据中心及其实现数据处理的方法和网络控制器。 [0001] The present invention relates to a software-defined network (SDN, Software Defined Network) technology, particularly to a data processing implemented method and a network controller based on the data center and SDN.

背景技术 Background technique

[0002]目前,为了实现数据中心内和数据中心间服务器的通信,需要构建网络结构,图1 为现有数据中心的网络结构示意图,如图1所示,以计算服务器为界,数据中心内网络可以分为前端网络和后端网络。 [0002] Currently, in order to achieve communication between the data center and data center servers, need to construct a network structure, FIG. 1 is a schematic diagram of a network structure of an existing data center, shown in Figure 1, to calculate the community server, the data center network may be divided into front and rear network network. 其中,前端网络主要承载着数据中心内计算服务器之间以及与数据中心外的服务器之间的通信,是基于TCP/IP网络体系架构,实现技术为IP+以太网(Ethernet)方式。 Wherein the front end of the main network carries communications between the server outside between computing servers within the data center and the data center is based on TCP / IP network architecture, Ethernet IP + implemented technology (Ethernet) mode. 后端网络则承载着服务器至存储系统的通信。 The backend network carries communication to the server storage system. 由于存储设备间通过存储区域网络(SAN,Storage Area Network)技术构建存储系统,是一种有别于IP+Ethernet的f术,但是,由于Ethernet技术拥有的冲突丢包的天然缺陷,使得后端网络的I/O同步问题、 同速低延迟等基本需求无法得到如足。 Since the storage system constructed between storage devices through a storage area network (SAN, Storage Area Network) technology, is a distinct IP + Ethernet, f technique, however, packet loss due to collision of the Ethernet technology has natural defects, such that the rear end network I / O synchronization problems with basic needs such as speed low latency can not get enough. 因而,目目! Thus, the head head! J,后端网络基本上米用基于光纤通彳目的SAN(FC SAN,Fiber Channel SAN)技术,实现计算系统与存储系统的通信。 J, m backend network with fiber-based substantially on left foot object of SAN (FC SAN, Fiber Channel SAN) technology, a computing system for communication with the storage system.

[0003] FC SAN技术的引入,使得数据中心的网络存在以下几个问题:多张网络,也就是说,对计算服务器而言,处于Ethernet和FC SAN两张网络中;多种端口协议,由于计算服务器存在于多个网络中,为了适应不同网络的需求,计算服务器上需要具备支持多个协议的网口,增加了计算服务器的复杂性;以及成本问题,由于FC SAN采用的是光纤通信,链路的设备成本较高。 FC SAN technology is introduced [0003], so that the data center network there are several problems: a plurality of networks, that is, for calculating the server, and in FC SAN two Ethernet network; multiple port protocol, since calculating a plurality of network servers exist in order to adapt to different networks, the server computing requires multiple protocols to support network ports, an increase computational complexity server; and cost, since the FC SAN is used in optical fiber communication, higher equipment costs link.

[0004]为此,在数据中心的后端网络中引入了基于光纤通信的以太网(FCoE,Fiber Channel over Ethernet)技术,通过对原有的FC数据包进行Ethernet封装,实现了前端网络与后端网络的融合,图2为现有基于FCoE技术的数据中心网络结构示意图,如图2所示,计算服务器上配备CAN网卡,计算服务器的任何通信均需求通过该CAN网卡发送给FCoE交换机。 [0004] For this reason, the introduction of the Ethernet (FCoE, Fiber Channel over Ethernet) technology at the rear end of fiber optic communication network in the data center, by the original Ethernet packet encapsulated FC achieved after front-end network end of any communication network integration, FIG 2 is a conventional diagram of a data center network architecture based FCoE art, shown in Figure 2, with the server computing CAN network card, server computing needs are sent to the FCoE switch through the CAN card. 位于FCoE交换机中的数据转发单兀(FCF,Fiber Channel Forwarder)(图2中未示出)会对数据进行识别,如果是正常的Ethernet数据包,则将其转发至IP/Ethernet网络中;如果是封装成Ethernet包的FC数据包,则将其转发至FC网络中。 FCoE switch in the data forwarding single Wu (FCF, Fiber Channel Forwarder) (not shown in FIG. 2) will identify the data, if it is normal Ethernet packet, it is forwarded to the IP / Ethernet network; if FC packets are encapsulated into Ethernet packets, and then forwards it to the FC network.

[0005] ^成技术提供了利用Ethernet技术实现计算服务器与后端存储网络通信需求的方案,但是仍然存在一些问题: [0005] ^ technology provides a solution to the use of Ethernet technology computing servers and back-end storage network communication needs, but there are still some problems:

[0006] 一方面,为了保证计算服务器与存储系统之间高速的数据传输,同时为了使存储系统之间的数据传输不占用以太网的资源,在数据中心,将计算服务器的通信网络与存储系统的通信网络分离开来,从而构建了两张网络,引来了多网络的问题:前端网络采用IP+ Ethernet技术进行组网,而存储网络采用SAN技术。 [0006] In one aspect, in order to ensure high-speed data transfer between the calculated and the storage system, and in order to make the transfer of data between storage systems do not use the Ethernet resources in the data center, the calculated communication network server storage system separated communications network, to construct the two networks, led to problems of multi-network: IP + Ethernet network using the front end of the networking technology, stored network using SAN technology. FCoE技术通过数据包封装的方式,实现了用以太网技术连接计算服务器和存储设备,但是,存储设备之间的网络仍然没有发生改变,也就是说,在数据中心之间仍然是多张网络的格局。 FCoE data packet encapsulation techniques, enabling a calculation Ethernet connection servers and storage devices, however, between the network storage device still does not change, that is, between a plurality of data centers still network pattern.

[0007]另一方面,目前在FCoE技术的部署方案中,基本上是在计算服务器的通信网以及存储网的边缘部署FCoE交换机。 [0007] On the other hand, in the current art FCoE deployment, deploying FCoE switch substantially at the edge of the communication network and storage network compute servers. 随着云计算的发展,计算服务器与存储系统之间的信息通信需求将日渐强烈。 With the development of cloud computing, information and communication needs between the computing servers and storage systems will become more intense. 为了承载大量的数据流动,FCoE交换机的部署范围将日益拓展,甚至FCoE交换机的部署返回会扩展至整个存储网络。 In order to carry large amounts of data flow, the deployment range FCoE switch will become increasingly expanded, deployed even FCoE switch returns will be extended to the entire storage network. 但是,FCoE交换机属于专用交换机,适用范围较窄;而且,FCoE交换机是基于光纤通信的,因此,基于FCoE技术的数据中心网络的成本显然是较高的,而且,后期存储网络的建设和改造将引发较大的资金投入。 However, FCoE switches are private branch exchange, a narrow scope of application; Further, FCoE switch is based on optical fiber communication, and therefore, the cost of data center network based FCoE art is clearly higher, and late storage network construction and transformation to lead to greater capital investment. 更重要的是,在云计算及互联网需求的推动下,40/100G网络技术在数据中心内得到了较好的发展,而且40/100G技术也将在数据中心中大行其道。 More importantly, cloud computing and the Internet to promote the needs of the next, 40 / 100G network technology has been good development in the data center, and 40 / 100G technology will be popular in the data center. 但是,静观FC SAN网络近些年来的发展,一直维持在10G以下的速度。 However, observing the development of FC SAN network in recent years, the rate has been maintained at less than 10G. 因而,在数据中心之间仍然是多张网络的格局中,必然会导致前端网络和后端网络速度不匹配的问题,从而导致了数据传输的瓶颈。 Thus, between data centers is still the pattern of multiple networks, will inevitably lead to problems front-end and back-end network does not match the speed of the network, resulting in a bottleneck of data transmission.

发明内容 SUMMARY

[0008] 为了解决上述技术问题,本发明提供了一种数据中心及其实现数据处理的方法和网络控制器,能够降低数据中心复杂度、避免数据传输的瓶颈,同时减小维护开销。 [0008] To solve the above problems, the present invention provides a data center and a network controller and implement a data processing method, it is possible to reduce the complexity of the data center, the data transmission to avoid bottlenecks, while reducing maintenance costs.

[0009] 为了达到本发明目的,本发明提供了一种网络控制器,至少包括策略处理模块、流表处理模块、路由处理模块,以及用于维护网络结构视图的拓扑处理模块;其中, [0009] To achieve the object of the present invention, the present invention provides a network controller, comprising at least policy processing module, the processing module flow table, the routing processing module, and processing means for maintaining a network topology view of the structure; wherein

[0010]策略处理模块,用于将业务层的用户需求转换为相应数据流的网络需求,并将这些网络需求通知给路由处理模块和流表处理模块; [0010] policy processing module, for converting the business needs of the user demand for network layer corresponding data stream and notifies the network needs to route processing module and the module processing flow table;

[0011] 路由处理模块,接收来自策略处理模块的网络需求,并根据拓扑管理模块维护的网络结构视图,为数据流规划最符合要求的转发路由; [0011] The route processing module, receiving a network policy requirements from the processing module, and according to the network topology management module maintains a structural view of the data stream forwarding route plan that best meet the requirements;

[0012] 流表处理模块,用于接收来自策略处理模块的网络需求,按照路由处理模块规划的转发路由生成相应的OpenFlow流表,并下发至OpenFlow交换机。 [0012] The flow table processing module, for receiving a demand from the network policy processing module, generating a corresponding flow table OpenFlow forwarding route processing module according to the routing plan, and issued to OpenFlow switch.

[0013]所述拓扑处理模块,还用于通过收集所述设备层中的各服务器的信息,在所述网络控制器中构建各服务器之间的逻辑拓扑图并存储。 [0013] The topology of the processing module is further for collecting server through the information of each layer in the device, construct a logical topology between the server and stored in the network controller.

[0014] 所述网络控制器还包括:设备处理模块,用于对所述网络数据转发层中的网络交换机进行管理、配置。 [0014] The network controller further comprising: a device processing module, for managing the data forwarding network layer network switch configuration.

[0015] 本发明又公开了一种数据中心,至少包括设备层、数据转发层,以及网络控制层; 其中, [0015] The present invention further discloses a data center, comprising at least a device layer, a data forwarding layer, and a network control layer; wherein,

[0016] 设备层包括一个或一个以上服务器,服务器包括计算服务器和存储服务器,用于依据自身支持的协议生成相应的数据包,并通过网卡连接到数据转发层,实现计算服务器之间、计算服务器与存储服务器之间、存储服务器与存储服务器之间的数据交换; [0016] The device layer includes one or more servers, including server computing servers and storage servers, for generating a corresponding self-supporting packet based protocol, and is connected through the network card forwards the data to the layer, to achieve between computing servers, computing servers between the server and the storage data between the storage server and the storage server exchange;

[0017]数据转发层由若干个支持Openf low协议的网络交换机组成,包括接入层和汇聚/ 核心层,用于按照网络控制层的控制对来自服务器的数据包进行数据转发; [0017] Data forwarded layer supported by a plurality of network switches Openf low protocol, including the access and aggregation layer / core layer, under the control of a network control layer packet forwarding data from the server;

[0018]网络控制层,包括网络控制器,用于控制数据的转发。 [0018] The network control layer, comprising a network controller, for forwarding control data.

[0019] 所述接入交换机还用于,预留预设数量个高速上行接口,用于保证所述接入交换机至汇聚/核心交换机的网络速度;将剩余的高速端口优先分配给所述存储服务器。 [0019] The access switch further configured to reserve a predetermined number of high-speed uplink interface, the access network for ensuring convergence to switch speed / core switch; the remaining port priority assigned to the high-speed memory server.

[0020] 本发明还公开了一种数据中心实现数据处理的方法,包括:服务器依据自身支持的协议生成相应的数据包; [0020] The present invention also discloses a method for data processing in a data center, comprising: a server generates a corresponding data packet according to the protocol supported by the user;

[0021] 接收到数据包的交换机确定数据包为所属的数据流中的第一个数据包,网络控制器根据数据包转发端口的类型,生成转发路径; [0021] received packet switch determines that the first data packet is a data packet stream belongs, the network controller forwards the packet in accordance with the type of port, generating a forwarding path;

[0022] 数据包所属数据流中的其他数据包采用生成的转发路径进行转发。 Other data packet [0022] data packet belongs in the data stream generated using the forwarding path for forwarding.

[0023] 所述确定数据包是否为其所属的数据流中的第一个数据包包括: [0023] The determining whether the data stream for which the data packet belongs to a first packet comprises:

[0024] 根据所述数据包的表示端口类型信息的字段,匹配交换机中的流表项,如果在交换机的流表中不存在相关的匹配信息,则确定所述数据包是所属的数据流中的第一个数据包,否则所述数据包不是第一数据包。 [0024] The type field indicates the port information of the data packet, the matching entry flow switch, if the relevant information does not exist in the matching flow table of the switch, it is determined that the data packet is a data stream belongs a first data packet, otherwise the packet is not a first packet.

[0025] 所述网络控制器根据数据包转发端口的类型,生成转发路径之前,该方法还包括: Before [0025] The network controller according to the type of packet forwarding port, generating a forwarding path, the method further comprising:

[0026] 所述交换机根据接收数据包的端口的类型生成执行相应的策略,在数据流来自高速的转发端口时,将该数据流标示为存储流,上报给所述网络控制器;所述控制器根据约束最短路由优先CSPF算法为该数据流规划路径。 When [0026] the switch generates the appropriate policy based on the type of port receiving data packets, the data stream from the high-speed forwarding port, designated to store the data stream flow, reported to the network controller; wherein the control according to the CSPF algorithm constraint shortest path First planned path for that data stream.

[0027] 与现有技术相比,本发明包括:在基于SDN的数据中心中,服务器根据自身支持的协议生成相应的数据包;确定数据包为所属的数据流中的第一个数据包,根据数据包转发端口的类型,生成转发路径;数据包所属数据流中的其他数据包采用生成的转发路径进行转发。 [0027] Compared with the prior art, the present invention comprises: SDN-based data centers, server generates a corresponding data packet according to the protocol supported by the; determining the first data packet is a data packet stream belongs, Depending on the type of packet forwarding port, generating a transfer path; other data packets of the packet flow relevant to the use of the generated forwarding path for forwarding. 本发明方法通过数据包转发端口的类型识别数据流的类型,规划了适合该数据流的最优转发路径,具体地,为在高速端口间转发的数据流设置高的服务等级质量,选择质量好、转发速度快的路径配置数据流转发路由,而为在低速端口间转发的数据流设置低的服务等级质量,选择质量不是最好、转发速度不是最快的路径配置数据流转发路由,为不同的数据流匹配了不同的服务等级质量,避免了数据传输的瓶颈,同时减小了维护开销。 Data type identification method of the present invention by a port forwarding packet type flow plan for optimal forwarding path of the data stream, specifically, a high-speed port forwarding between the high-level data flow quality of service provided, to select good quality fast forwarding speed configuration data forwarding path routing, and a low grade of service quality of the data stream forwarding port disposed between the low speed, the quality is not the best choice, not the fastest forward speed of the data forwarding path routing configuration for different the data stream matches the different levels of service quality and avoid the bottleneck of data transmission, while reducing maintenance costs. 而且基于SDN的数据中心降低了原有数据中心复杂度。 And SDN-based data center to reduce the existing data center complexity.

[0028] 本发明的其它特征和优点将在随后的说明书中阐述,并且,部分地从说明书中变得显而易见,或者通过实施本发明而了解。 [0028] Other features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or learned by practice of the present invention. 本发明的目的和其他优点可通过在说明书、权利要求书以及附图中所特别指出的结构来实现和获得。 The objectives and other advantages of the present invention can be in the specification, the drawings, and particularly pointed out in the structure realized and attained by the claims.

附图说明 BRIEF DESCRIPTION

[0029] 附图用来提供对本发明技术方案的进一步理解,并且构成说明书的一部分,与本申请的实施例一起用于解释本发明的技术方案,并不构成对本发明技术方案的限制。 [0029] The accompanying drawings provide a further understanding of aspect of the present invention, and constitute part of the specification, embodiments of the present application, serve to explain the technical solution of the present invention, not to limit the technical solution of the present invention.

[0030] 图1为现有数据中心的网络结构示意图; [0030] FIG. 1 is a schematic diagram of a network structure of a conventional data center;

[0031] 图2为现有基于FCoE技术的数据中心网络结构示意图; [0031] FIG 2 is a conventional data center network architecture based on a schematic FCoE technology;

[0032]图3为现有SDN架构的示意图; [0032] FIG. 3 is a diagram of a conventional architecture SDN;

[0033] 图4为本发明数据中心的网络结构示意图; [0033] FIG. 4 is a schematic structure of a data center network of the present invention;

[0034] 图5为本发明数据中心中不同数据流的流向示意图; [0034] FIG. 5 is a schematic flow different data streams in the data center to the present invention;

[0035] 图6为本发明数据中心中的网络控制器的逻辑组成结构示意图; [0035] FIG 6 schematic diagram illustrating a logical structure data center network controller of the present invention;

[0036] 图7为本发明实现数据处理的方法的流程图。 [0036] FIG. 7 is a data processing flowchart of a method to achieve the present invention.

具体实施方式 Detailed ways

[0037] 为使本发明的目的、技术方案和优点更加清楚明白,下文中将结合附图对本发明的实施例进行详细说明。 [0037] To make the objectives, technical solutions, and advantages of the present invention will become apparent from, the accompanying drawings hereinafter in conjunction with embodiments of the present invention will be described in detail. 需要说明的是,在不冲突的情况下,本申请中的实施例及实施例中的特征可以相互任意组合。 Incidentally, in the case of no conflict, embodiments and features of the embodiments of the present application may be arbitrarily combined with each other.

[0038] 在附图的流程图示出的步骤可以在诸如一组计算机可执行指令的计算机系统中执行。 [0038] The steps shown in the flowchart drawings can be executed in a computer system a set of computer executable instructions. 并且,虽然在流程图中示出了逻辑顺序,但是在某些情况下,可以以不同于此处的顺序执行所示出或描述的步骤。 In addition, although in the flowchart shown in a logical order, but in some cases, the steps shown or may be performed in a different order than described herein.

[0039] 软件定义网络(SDN,Software Defined Network)是控制与转发分离并直接可编程的网络架构,图3为现有SDN架构的示意图,如图3所示,其核心思想是将传统网络设备紧耦合的网络架构解耦成应用、控制、数据(转发)三层分离的架构,并通过标准化实现网络的集中管控和网络应用的可编程。 [0039] software-defined network (SDN, Software Defined Network) is the control and forwarding are separated and directly programmable network architecture, FIG. 3 is a schematic SDN conventional architecture shown in Figure 3, the core idea is the traditional network device tightly coupled to the network architecture decoupling applications, control data (forwarding) Layer separation architecture and programmable centralized control and a network through standardized network application. 在SDN架构下,开放和标准化是核心关键点:标准化数据面与控制面的接口,又称为南向接口,屏蔽了网络基础设施资源在类型、支持的协议等方面的异构性,使得数据面的网络资源设施资源能够无障碍地接收控制面的指令,承载网络中的数据转发业务;标准化控制层和应用层的接口,又称为北向接口,为上层应用提供统一的管理视图和编程接口,使得用户可以通过软件从逻辑上来定义网络控制和网络服务。 In the SDN architecture, open and standardized core of key points: the interface standardized data plane and control plane, also known as the southbound interface, shielding the network infrastructure resources heterogeneity in terms of type, supported protocols, etc., so that data network resource facility resources accessible surface capable of receiving the instruction of the control plane, the network carrying data forwarding service; interface standardization control layer and application layer, also known as NBI, for the unified management view and the upper application programming interface , so that the user can define the network control and network software services through logical up.

[0040]图4为本发明数据中心的网络结构示意图,如图4所示,至少包括设备层、数据转发层,以及网络控制层;其中, [0040] FIG. 4 is a schematic structural diagram of a data center network of the present invention, shown in Figure 4, the device comprising at least a layer, a data forwarding layer, and a network control layer; wherein,

[0041]设备层,其中包括一个或一个以上服务器,这里,服务器包括计算服务器和存储服务器,计算服务器、存储服务器均用于,依据自身支持的协议生成相应的数据包,并通过网卡连接到数据转发层的接入交换机,并通过接入交换机实现计算服务器之间、计算服务器与存储服务器之间、存储服务器与存储服务器之间的数据交换。 [0041] The device layers including one or more servers, where the server includes a server computing and storage servers, computing servers, storage servers are used to generate corresponding data packets based on their own protocols supported by the card and connected to the data the access switch forwarding layer, and is achieved by an access switch between computing server, between the server and the storage server computing, data between the storage server and the storage server exchange.

[0042]数据转发层包括接入层和汇聚/核心层,由若干个支持〇penf l〇w协议的网络交换机组成,用于按照网络控制层的控制对来自服务器的数据包进行数据转发。 [0042] The data forwarding comprises an access and aggregation layer / core layer, supported by a plurality of network switches 〇penf l〇w protocol, and is used under the control of the network control layer of the data packets from the server to forward data.

[0043]在数据转发层中,所有的网络交换机均通过统一、标准的协议,g卩Openfl〇ff协议, 与网络控制层中的网络控制器进行通信。 [0043] In the data forwarding layer, all of the network switches are unified, standard protocol, g Jie Openfl〇ff protocol, communicates with the network control layer of the network controller. 这样,一方面,标准的协议为网络控制器屏蔽了设备层中计算服务器、存储服务器等设备的异构性,也就是说,设备层中的设备对网络控制器来说是不存在差异的,从而降低了对不同厂家设备的依赖程度;另一方面,网络的本质就是数据转发,因而会涉及到大量的数据转发设备,在本发明的数据中心中,数据转发设备的成本仅在于转发芯片,更容易实现量产,使得网络成本的降低具有了较大的优势。 Thus, on the one hand, the standard protocol for the network controller shielding layer heterogeneity of computing device servers, storage servers and other devices, i.e., devices in the device layer network controller is no difference, and thereby reducing the reliance on equipment from different vendors; on the other hand, the nature of the network is the data forwarding, and therefore involve a large amount of data forwarding devices, in a data center according to the present invention, the cost of the data transfer apparatus only in forward chip, easier to mass production, so as to reduce network costs have a greater advantage.

[0044]在本发明数据中心中,为了尽可能地降低物理端口速度对数据流速度的限制,在构建计算/存储服务器与接入交换机、接入交换机与汇聚/核心交换机之间的连接方式时, 首先,接入交换机预留预设数量如2个高速上行接口,用于保证接入交换机至汇聚/核心交换机的网络速度;其次,如果还有高速端口剩余,那么,遵循优先分配给存储服务器的原则。 [0044] In the present invention, the data center in order to reduce as much as possible the physical port speed limit data stream velocity, when building calculation / storage server and the access switches, access switches and the connection between the aggregation / core switch number, first, the preset reservation access switches such as the high speed uplink interfaces 2, to guarantee the access switch to the aggregation network speed / core switch; secondly, if there is remaining high-speed ports, then follow preferentially allocated to the storage server the rules. 需要说明的是,这里并不是一定要保证所有的交换机之间都能以高速端口连接,但是,按照本发^方法,会尽量满足高速连接需求;如图5所示,图5为本发明数据中心中不同数据流的流向示意图,图5中,以接入交换机、汇聚/核心交换为〇penFl〇w交换机、网络控制器为SDN控制器f例,斜线阴影小方框所示端口表示高速的交换机端口,斜小方格阴影小方框所示端口表示低速的交换机端口。 Incidentally, this does not make sure that all the switches between high-speed ports can be connected to, however, the method according to the invention ^, will try to meet the needs of high-speed connections; FIG. 5, FIG. 5 of the present invention, data Center flow schematic of the different data streams, in FIG. 5, the access switch, aggregation / core exchanged for 〇penFl〇w switch controller f SDN network controller embodiment, shown diagonally shaded small box represents a high-speed port switch port, obliquely hatched small square small blocks shown represent low-port switch port.

[0045]网络控制层,包括网络控制器,用于控制数据的转发,以及数据中心的管理。 [0045] The network control layer, comprising a network controller, for forwarding control data, and data center management. 图6为本发明数据中心中的网络控制器的逻辑组成结构示意图,如图6所示,主要包括策略处理模块、流表处理模块、路由处理模块,以及用于维护网络结构视图的拓扑处理模块;其中, FIG 6 is a schematic diagram illustrating a logical structure of a data center network controller of the invention, as shown in Figure 6, including the policy processing module, the processing module flow table, the routing processing module, and processing means for maintaining a network topology view of the structure ;among them,

[0046]策略处理模块,用于将业务层的用户需求转换为相应数据流的网络需求,比如网络延迟要求、重传率要求等,并将这些网络需求通知给路由处理模块和流表处理模块。 [0046] policy processing module, for converting a user needs the service layer data flow corresponding to network requirements, such as network delay requirement, requirements retransmission rate, and notifies the network needs to route processing module processing module and a flow meter . 具体地,用于为计算服务器与计算服务器、计算服务器与存储服务器、存储服务器之间的转发需求设置不同的网络服务质量。 Specifically, the calculation server is used to compute servers, storage server computing servers, storage servers forward demand between different network quality of service provided.

[0047]路由处理模块,接收来自策略处理模块的网络需求,并根据拓扑管理模块维护的丨糾络结构视图,为数据流规划最符合要求的转发路由。 [0047] routing processing module receives the network needs from policy processing module, according to Shu and correct network topology management module maintains a structural view of the data stream forwarding route plan best meets the requirements.

[0048]流表处理模块,用于接收来自策略处理模块的网络需求,按照路由处理模块规划的转发路由生成相应的OpenFlow流表,并下发至0penFlow交换机。 [0048] The flow table processing module, for receiving a demand from the network policy processing module, generating a corresponding flow table OpenFlow forwarding route processing module according to the routing plan, and issued to 0penFlow switch. 这样,当来自服务器的数据包经过这些OpenFlow交换机时,按照0penFlow交换机中存储的流表进行转发。 Thus, when the data packets from the server through these OpenFlow switch, according 0penFlow forward flow table stored in the switch.

[0049]本发明拓扑处理模块,还用于通过收集设备层中的各服务器的信息,在网络控制器中构建各服务器之间的逻辑拓扑图并存储,这样,在生成转发路径时,可以作为依据。 [0049] The topology of the processing module of the present invention, further information for each server apparatus through the collecting layer, constructing the logical topology between the server and stored in the network controller, so that, when generating the forwarding path, as in accordance with. 其中,收集的信息包括但不限于:服务器的端口、服务器的邻居节点信息以及邻居与端口的对应关系、服务器类型(如计算服务器、或者存储服务器)等。 Wherein the collected information includes but is not limited to: the port, and the correspondence between the neighbor node information, and the neighbor port of the server type of the server the server (e.g., computing servers, storage, or server) and the like. 具体如何构建逻辑拓扑图,属于本领域技术人员的惯用技术手段,不用于限定本发明的保护范围,这里不再赘述。 Specifically, how to build a logical topology, conventional techniques within the skill of the art, not intended to limit the scope of the present invention and will not be repeated here.

[0050]_本发明网络控制器还包括:设备处理模块,用于对网络数据转发层中的网络交换机进行状态监控、转发端口设置、VLAN设置等,以实现网络控制器对数据转发层中的网络交换机的管理、配置。 [0050] The present invention further _ network controller comprising: a device processing module for the network layer data forwarding condition monitoring network switch, the forwarding port, VLAN and other settings to implement data forwarding network controller layer management network switch configuration.

[0051]在本发明数据中心的网络控制器中,将网络控制器的各项功能采用直观的可视化界面,展现给网络管理人员,网络管理人员通过可视化界面设置各项用户需求,并提供给策略处理模块。 [0051] In the data center network controller of the present invention, the various functions of the network controller intuitive visual interface, to show to the network management, network management is provided by the user needs visual interface, and provided to the policy processing module.

[0052]本发明数据中心还包括:网络应用层,包括如防火墙、网络监控等应用,利用网络控制层开放的API接口,由开发人员自主开发,其具体实现不在本发明的保护范围之内,这里不再赘述。 [0052] The present invention is a data center further comprises: a network application layer, including applications such as firewalls, network monitoring, etc., using an open network control layer API interface, developed by the developer, within the scope of the specific implementation is not of the present invention, not repeat them here.

[0053]图4所示的本发明基于SDN的数据中心网络架构中,在设备层,计算服务器与计算服务器之间的数据通信遵循IP/Ethernet协议,计算服务器与存储服务器之间,以及存储服务器之间的通信遵循FC SAN协议。 The present invention shown in [0053] FIG. 4 data center network architecture based SDN, data communications between the device layer, with the calculation server computing server follows IP / Ethernet protocol between the server and the storage server computing and storage servers communication between FC SAN follow protocol. 相对于TCP/IP架构来说,本发明基于SDN的数据中心网络架构充分利用现有网络的设备(如计算服务器、存储服务器等),不需要对原有设备升级,设备依据自身支持的协议生成数据报文即可,本发明只需要一张网即可承载传统的以太网、 存储网的数据流量,降低了数据中心复杂度;本发明数据中心中的网络交换机是支持Openf low协议的OpenFlow交换机,所有的网络交换机均通过统一、标准的协议与网络控制层中的网络控制器进行通信,并且只保留了转发面,使得网络交换机可以量产,降低了网络成本,而且标准的协议为网络控制器屏蔽了设备层中各种设备的异构性,使得设备层中的设备对网络控制器来说是不存在差异的,从而降低了对不同厂家设备的依赖程度。 TCP / IP architecture is relative, the present invention is based on the data center network architecture SDN fully utilize the existing network devices (e.g., computing servers, storage servers, etc.), no need to upgrade existing equipment, based on equipment supported by the protocol generator data packets can, according to the present invention requires an Ethernet network data traffic, the network can be stored in a conventional carrier, reducing the complexity of the data center; in the present invention, a data center network switch is supported protocol OpenFlow switch Openf low All network switches are performed by a unified, standard protocol and the network control layer of the network communications controller, and retaining only the forwarding plane so that the network can switch the amount of production, reducing the cost of the network, and network control protocol standard shielding the heterogeneity of the various devices in the device layer, such that the device layer to the network controller device is no difference, thereby reducing the reliance on equipment from different vendors. 更重要的是,本发明网络控制器在确定需求策略时,区分不同数据流设置不同的服务等级质量,以生成适合该数据流的最优转发路径,避免了数据传输的瓶颈,同时减少了维护开销。 More importantly, the present invention is a network controller in determining policy needs to distinguish between different data flows of different quality of service level is provided to generate the optimal route for forwarding the data stream to avoid the bottleneck of data transmission, while reducing maintenance overhead.

[0054]图7为本发明实现数据处理的方法的流程图,如图7所示,包括: [0054] FIG. 7 is a data processing flowchart of a method of realization of the present invention, shown in Figure 7, comprising:

[0055]步骤700:服务器依据自身支持的协议生成相应的数据包。 [0055] Step 700: The server generates a corresponding data packet according to the protocol supported by the user. 本步骤的具体实现属于本领域技术人员的惯用技术手段,这里不再赘述。 Specific implementation of this step belongs to the usual techniques of those skilled in the art, it will not be repeated here.

[0056]步骤701:接收到数据包的交换机确定数据包为所属的数据流中的第一个数据包, 网络控制器根据数据包转发端口的类型,生成转发路径。 [0056] Step 701: Receive the packet switch determines that the first data packet is a data packet stream belongs, the type of network interface controller forwarding of data packets according to generate forwarding path.

[0057]本步骤还包括:如果数据包不是其所属数据流中的第一个数据包,那么,按照之前下发的转发路径进行转发即可,结束本流程。 [0057] The present step further comprises: if the first packet is not a packet data stream to which it belongs, then forwarded according to the forwarding path before issued, the process ends.

[0058]本步骤中,确定数据包是否为其所属的数据流中的第一个数据包包括:根据数据包的L2/L3/L4、端口号、VLAN等表示端口类型信息的字段,匹配交换机中的流表项,如果在交换机的流表中不存在相关的匹配信息,则确定该数据包是所属的数据流中的第一个数据包,否则该数据包不是第一数据包。 [0058] In this step, it is determined whether the data packet belongs in the data stream for the first data packet comprising: information field indicating the type of the port, the switch according to the data packet matches L2 / L3 / L4, port number, VLAN, etc. the flow table entry, if the relevant information does not exist in the matching flow table of the switch, it is determined that the packet is the first packet in the data stream belongs, otherwise the packet is not the first data packet.

[0059] 本步骤中,交换机根据接收数据包的端口的类型生成执行相应的策略如果数据流来自高速的转发端口,则将该数据流标示为存储流,上报给SDN控制器;SDN控制器中的路由转发模块会根据CSPF算法为该数据流规划合适的路径,比如带宽大、链路利用率较低的路径。 [0059] In this step, the switch port is generated according to the type of the received packet if the appropriate policy from the high speed data stream forwarding port is designated to store the data stream flow, SDN reported to the controller; SDN controller the routing and forwarding module CSPF algorithm for that data stream suitable route planning, such as large bandwidth, a low link utilization path. 这里,使用何种策略与网络配置相关,可通过设定CSPF中的约束条件来达到。 Here, what policies associated with the network configuration, can be achieved by setting constraints in CSPF.

[0060] 这里,生成转发路由的算法可以采用现有的约束最短路由优先(CSPF)算法,并根据自身需求对算法中的约束条件进行修改和匹配。 [0060] Here, the routing algorithm can be generated using forward existing constraint Shortest Path First (the CSPF) algorithm, and the modified matching algorithm and constraints according to their needs. 由于存储数据流的链路占用时间长、数据流量大等特点,因此,其中的约束条件可以具体化为:链路累计占用率<7〇%,以使得存储数据流选择非关键繁忙路径,这里约束条件的具体规则与数据流的需求、网管人员的设计等有关。 Since the length of the data stream stored link occupation time, large data traffic characteristics, therefore, which may be embodied constraints: the cumulative link occupancy rate <7〇%, so that the stored data stream is a non-critical path peak where demand specific rules and data flow constraints, network management personnel to design and so on. CSPF算法的使用属于本领域技术人员的常用技术,这里不再赘述。 CSPF algorithm using conventional techniques belong to the skilled person, it will not be repeated here.

[0061] 步骤702:数据包所属数据流中的其他数据包采用生成的转发路径进行转发。 [0061] Step 702: the other packets of the packet flow relevant to a data transfer path using the generated forward.

[0062] 举个例子来说,本发明实现数据处理的方法包括: [0062] For example, the data processing method of the present invention is implemented comprises:

[0063] 在设备层中的服务器依据自身支持的协议生成相应的数据包,并通过相应的网卡将数据包发送给网络交换机;在数据转发层,网络交换机则根据L2/L3/L4、端口号、VLAN等字段匹配收到的数据包,以确定该数据包是否是所属的数据流中的第一个数据包;如果是, 则将该数据包转发给网络控制层中的网络控制器,网络控制器通过全局的统一视图以及数据包的格式和协议等需求策略,生成相应的转发路由;并将相应的转发路由发送给相关的网络交换机,这样,以后收到的该条数据流的其他数据包将采用生成的转发路由进行转发; 如果不是第一个数据包,则按照之前网络控制器下发的转发路径进行转发即可。 [0063] The server device layer is generated based on its own protocol supported by the corresponding data packet, and sends the packet to the network switch via a respective card; forwarding data layer, network switches is according L2 / L3 / L4, port number , VLAN and other fields match the received data packet to determine whether the packet is a first packet in a data stream belongs; if so, forwarding the data packet to the network control layer of the network controller, the network controller needs global policy and a unified view of data packet formats and protocols, etc., to form the corresponding forwarding route; and transmits corresponding forward routing to the appropriate network switch, so that other piece of data received after the data stream the packet forwarding route generated using forwards; if not the first packet, forwarding is performed according to previously delivered by the network controller transfer path. 其中,生成相应的转发路由具体包括:如果数据流来自于高速的转发端口并且要去往高速的转发接口时,则认为该数据流是存储服务器与存储服务器之间的数据流,或者是对网络服务质量要求高的数据流。 Wherein, to form the corresponding forwarding route comprises: if the data stream from the fast forwarding port and forwarding destined for high-speed interfaces, it is considered that the data stream is a data flow between the server and stored in the storage server, or the network high quality of service requirements of the data stream. 这时,网络控制器会给这部分数据流设置高的服务等级质量,而在配置数据流转发路由时,会选择质量好、转发速度快的路径。 At this time, the network controller will give this part of the data stream is set high service level quality, and in the configuration of a data stream routing, choose good quality, fast forwarding speed path.

[0064]本发明方法通过数据包转发端口的类型区分不同数据流,设置不同的服务等级质量,生成了适合该数据流的最优转发路径,即为在高速端口间转发的数据流设置高的服务等级质量,选择质量好、转发速度快的路径配置数据流转发路由,而为在低速端口间转发的数据流设置低的服务等级质量,选择质量不是最好、转发速度不是最快的路径配置数据流转发路由,避免了数据传输的瓶颈,同时减少了维护开销。 [0064] The method of the present invention, the type of the data packet forwarding port distinguish different data streams provided different service quality levels, generating the optimal path for forwarding the data stream, that is, between the high-speed data flow port forwarding setting service quality level, a good choice for quality, fast forwarding speed of configuration data forwarding path routing, and is low during low-speed data flow port forwarding setting service quality level, the quality is not the best choice, not the fastest path forwarding speed configuration data Traffic routes, avoiding the bottleneck of data transmission, while reducing maintenance costs.

[0065]虽然本发明所揭露的实施方式如上,但所述的内容仅为便于理解本发明而采用的实施方式,并非用以限定本发明。 [0065] While the disclosed embodiment of the present invention described above, but the embodiment is provided only to facilitate understanding of the invention embodiment is employed, the present invention is not limited thereto. 任何本发明所属领域内的技术人员,在不脱离本发明所揭露的精神和范围的前提下,可以在实施的形式及细节上进行任何的修改与变化,但本发明的专利保护范围,仍须以所附的权利要求书所界定的范围为准。 Any skilled person in the art the present invention pertains, without departing from the spirit and scope of the disclosed invention may be made any modifications and changes in form and detail of the embodiments, but the scope of the present invention patent, still in the appended claims define the scope of equivalents.

Claims (2)

1. 一种数据中心实现数据处理的方法,其特征在于,包括:服务器依据自身支持的协议生成相应的数据包; 接收到数据包的交换机确定数据包为所属的数据流中的第一个数据包,网络控制器根据数据包转发端口的类型,生成转发路径; 数据包所属数据流中的其他数据包采用生成的转发路径进行转发; 所述网络控制器根据数据包转发端口的类型,生成转发路径之前,该方法还包括: 所述交换机根据接收数据包的端口的类型生成执行相应的策略,在数据流来自高速的转发端口时,将该数据流标示为存储流,上报给所述网络控制器;所述控制器根据约束最短路由优先CSPF算法为该数据流规划路径。 1. A data center for data processing implemented method, characterized by comprising: generating a server based on protocols supported by the respective data packet; receiving the packet switch determines the first data packet is a data flow belongs packet, the network controller forwards the packet in accordance with the type of port, generating a forwarding path; other data packets of the packet flow relevant to the use of the generated forwarding path for forwarding; the network controller port forwarding data packets according to the type, form a transfer before the path, the method further comprising: generating said switch port according to a type of the received packet the appropriate policy, from the high speed data stream forwarding port, designated to store the data stream flow, reported to the network control ; said controller according to the constrained shortest path First algorithm CSPF planned path for the data stream.
2. 根据权利要求1所述的方法,其特征在于,所述确定数据包为其所属的数据流中的第一个数据包包括: 根据所述数据包的表示端口类型信息的字段,匹配交换机中的流表项,如果在交换机的流表中不存在相关的匹配信息,则确定所述数据包是所属的数据流中的第一个数据包, 否则所述数据包不是第一数据包。 2. The method according to claim 1, wherein said data flow determining for a packet belongs to a first packet data comprising: the port type field indicates the information of the data packet, matching the switch the flow table entry, if the relevant information does not exist in the matching flow table of the switch, it is determined that the data packet is the first packet data stream belongs, otherwise the data packet is not the first data packet.
CN201410192914.2A 2014-05-08 2014-05-08 Data center and data processing implemented method and the network controller CN103986663B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410192914.2A CN103986663B (en) 2014-05-08 2014-05-08 Data center and data processing implemented method and the network controller

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410192914.2A CN103986663B (en) 2014-05-08 2014-05-08 Data center and data processing implemented method and the network controller

Publications (2)

Publication Number Publication Date
CN103986663A CN103986663A (en) 2014-08-13
CN103986663B true CN103986663B (en) 2018-03-27

Family

ID=51278490

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410192914.2A CN103986663B (en) 2014-05-08 2014-05-08 Data center and data processing implemented method and the network controller

Country Status (1)

Country Link
CN (1) CN103986663B (en)

Families Citing this family (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105704054A (en) * 2014-11-27 2016-06-22 英业达科技有限公司 Data center network flow migration method and system thereof
CN105847161B (en) * 2015-01-15 2019-07-23 南宁富桂精密工业有限公司 It controls equipment and its controls the method that interchanger updates flow table
CN104717098B (en) * 2015-04-09 2017-12-29 北京邮电大学 A data processing method and apparatus
CN105024853A (en) * 2015-07-01 2015-11-04 中国科学院信息工程研究所 SDN resource matching and service path discovering method based on rumor spreading mechanism
CN105406992B (en) * 2015-10-28 2018-11-09 浙江工商大学 A kind of business demand conversion and dispositions method towards SDN
CN105337872B (en) * 2015-11-18 2018-05-04 东北大学 One kind of energy-efficient method of controlling priority levels of the network-based partitioning
CN105357310A (en) * 2015-11-20 2016-02-24 中国石油天然气股份有限公司华北油田分公司 Energy cloud platform and data interaction method based on energy cloud platform
CN105515844A (en) * 2015-12-01 2016-04-20 国家电网公司 System for improving software defined network reliability
CN105703960A (en) * 2016-04-25 2016-06-22 刘昱 Network function management system based on SDN and method thereof
CN105897611B (en) * 2016-06-24 2019-03-29 武汉绿色网络信息服务有限责任公司 A kind of system and method that business identification and flow scheduling is realized using DPI technology based on SDN
CN106161457B (en) * 2016-07-26 2019-09-27 刘昱 Network domains isolating device and method based on SDN
CN106302254A (en) * 2016-07-29 2017-01-04 山东大学苏州研究院 SDN-based Internet of Things network data forwarding system
CN106357534A (en) * 2016-08-25 2017-01-25 江苏省未来网络创新研究院 Network flow monitoring system and method based on SDN

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102546351A (en) * 2012-03-15 2012-07-04 北京邮电大学 System and method for interconnecting openflow network and conventional Internet protocol (IP) network
CN102843299A (en) * 2012-09-12 2012-12-26 盛科网络(苏州)有限公司 Method and system for realizing Openflow multi-stage flow tables on basis of ternary content addressable memory (TCAM)
CN102916887A (en) * 2012-10-23 2013-02-06 清华大学 OpenFlow out-of-band networking method based on in-band virtual channel
CN102957624A (en) * 2012-12-12 2013-03-06 中国联合网络通信集团有限公司 Content routing method and device
CN103179046A (en) * 2013-04-15 2013-06-26 昆山天元昌电子有限公司 Data center flow control method and data center flow control system based on openflow
WO2013139298A1 (en) * 2012-03-22 2013-09-26 Huawei Technologies Co., Ltd. Supporting software defined networking with application layer traffic optimization
CN103338163A (en) * 2013-07-16 2013-10-02 清华大学 Software-defined network controller supporting scheduling of dynamic elastic resource
CN103414626A (en) * 2013-08-28 2013-11-27 盛科网络(苏州)有限公司 Message processing method and device based on network virtualization
WO2014055912A1 (en) * 2012-10-05 2014-04-10 Huawei Technologies Co., Ltd. Software defined network virtualization utilizing service specific topology abstraction and interface
KR20140050463A (en) * 2012-10-19 2014-04-29 에스케이텔레콤 주식회사 Method and apparatus to implement differential networks based on virtual network

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102546351A (en) * 2012-03-15 2012-07-04 北京邮电大学 System and method for interconnecting openflow network and conventional Internet protocol (IP) network
WO2013139298A1 (en) * 2012-03-22 2013-09-26 Huawei Technologies Co., Ltd. Supporting software defined networking with application layer traffic optimization
CN102843299A (en) * 2012-09-12 2012-12-26 盛科网络(苏州)有限公司 Method and system for realizing Openflow multi-stage flow tables on basis of ternary content addressable memory (TCAM)
WO2014055912A1 (en) * 2012-10-05 2014-04-10 Huawei Technologies Co., Ltd. Software defined network virtualization utilizing service specific topology abstraction and interface
KR20140050463A (en) * 2012-10-19 2014-04-29 에스케이텔레콤 주식회사 Method and apparatus to implement differential networks based on virtual network
CN102916887A (en) * 2012-10-23 2013-02-06 清华大学 OpenFlow out-of-band networking method based on in-band virtual channel
CN102957624A (en) * 2012-12-12 2013-03-06 中国联合网络通信集团有限公司 Content routing method and device
CN103179046A (en) * 2013-04-15 2013-06-26 昆山天元昌电子有限公司 Data center flow control method and data center flow control system based on openflow
CN103338163A (en) * 2013-07-16 2013-10-02 清华大学 Software-defined network controller supporting scheduling of dynamic elastic resource
CN103414626A (en) * 2013-08-28 2013-11-27 盛科网络(苏州)有限公司 Message processing method and device based on network virtualization

Also Published As

Publication number Publication date
CN103986663A (en) 2014-08-13

Similar Documents

Publication Publication Date Title
CN103346922B (en) The controller determines the state of the network and is determined based sdn
CN104780099B (en) A kind of method and controller network equipment for establishing path
CN105681231B (en) Autonomous networks and in autonomous networks handle data grouping method
CN101313528B (en) Multi-domain route computation method and system
US9450817B1 (en) Software defined network controller
US8867349B2 (en) Regulation of network traffic in virtual private networks
CN103051565B (en) A hierarchical system architecture software-defined network controller and implementation method
US8737394B2 (en) Route computation method and system, and path computation element
CN102549972B (en) Data center floo
CN1890943B (en) A method for optical networking infrastructure and between servers and storage area network
US7215640B2 (en) Method and apparatus for path configuration in networks
CN104426766B (en) The end-to-end network path of dynamic across multiple network layers is established
Azodolmolky et al. Integrated OpenFlow–GMPLS control plane: an overlay model for software defined packet over optical networks
EP2577911B1 (en) Communication available transport network bandwidth to l2 ethernet nodes
CN105229968B (en) For the method and system of the bandwidth aware service arrangement of service chaining
CN104272653B (en) Congestion control packet data networking in
CN100490420C (en) Packet transfer apparatus
US9319324B2 (en) Method and system of service placement for service chaining
CN103179046B (en) Openflow data center based flow control method and system
CN104702522B (en) The computer-implemented method in a software-defined network routing data, means, the controller
US8606105B2 (en) Virtual core router and switch systems and methods with a hybrid control architecture
US7912087B2 (en) Capacity variable link apparatus and capacity variable link setting method
JP2007532070A (en) Distinction transfer in the address-type carrier network
CN1731757A (en) Method and system for providing QoS assurance in edge router
CN1518278A (en) System for realizing resource distribution in network communication and its method

Legal Events

Date Code Title Description
C06 Publication
C10 Entry into substantive examination
GR01