WO2017084448A1 - 一种网络系统及网络运行方法 - Google Patents

一种网络系统及网络运行方法 Download PDF

Info

Publication number
WO2017084448A1
WO2017084448A1 PCT/CN2016/101394 CN2016101394W WO2017084448A1 WO 2017084448 A1 WO2017084448 A1 WO 2017084448A1 CN 2016101394 W CN2016101394 W CN 2016101394W WO 2017084448 A1 WO2017084448 A1 WO 2017084448A1
Authority
WO
WIPO (PCT)
Prior art keywords
controller
network system
data packet
network
switches
Prior art date
Application number
PCT/CN2016/101394
Other languages
English (en)
French (fr)
Inventor
翟跃
Original Assignee
上海斐讯数据通信技术有限公司
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 上海斐讯数据通信技术有限公司 filed Critical 上海斐讯数据通信技术有限公司
Publication of WO2017084448A1 publication Critical patent/WO2017084448A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/12—Shortest path evaluation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
    • H04L41/12—Discovery or management of network topologies

Definitions

  • the present invention relates to the field of electronic communications, and in particular to a network system and a network operation method.
  • the traditional OSPF Open Shortest Path First
  • OSPF Open Shortest Path First
  • LSA Link-State Advertisements
  • an object of the present invention is to provide a network system and a network operation method for solving the problem of complicated short-path priority configuration and complicated management operation and low efficiency in the prior art.
  • the present invention provides a network system, where the network system includes a controller and at least one switch, and each of the switches prestores a corresponding flow table, and the flow table includes a packet header. And performing an action; each of the switches sends its basic information message to the controller, where the basic information message includes a MAC address, port information, and link state information of the switch itself; a routing algorithm, acquiring a network topology of the network system, and generating a link state database according to the received link state information sent by all the switches in the network system, and generating a zone according to the link state database a weighted directed graph, and according to the weighted directed graph, each of the switches is used as a root node to respectively generate a corresponding minimum spanning tree; and when the switch receives a data with a destination IP address When the message is received, it is determined whether the data packet matches the corresponding flow table of the switch, and when not, the data is Sending to the controller, cause the controller to query the respective minimum
  • the controller generates a weighted directed graph according to the link state database, and according to the weighted directed graph, each of the switches is a root node, respectively A minimum spanning tree corresponding to each of the switches is generated.
  • the switch is a software defined network switch
  • the controller is a software defined network controller
  • each of the software-defined network switches communicates with the software-defined network controller via an OpenFlow channel.
  • the basic information message is a private experimenter message.
  • the packet header of the flow table includes a destination IP address, and when the destination IP address in the data packet is the same as the destination IP address included in the packet header of the flow table, the data is The packet matches the flow table. When the destination IP address in the data packet is different from the destination IP address included in the packet header of the flow table, the data packet does not match the flow table.
  • the controller acquires a network topology of the network system according to the shortest path algorithm and a link layer discovery protocol.
  • the controller maintains a set of states for each of the switches.
  • the set of states includes an initialization state, a load state, an interaction state, and a completion state.
  • a link overhead formula is predefined in the controller, and the controller generates the weighted directed graph according to the link state database and the link cost formula.
  • the present invention further provides a network operation method, running in the network system according to any one of the preceding claims, the method comprising the steps of: each of the switches transmitting to the controller
  • the basic information packet of the network includes the MAC address, port information, and link state information of the switch itself;
  • the controller acquires the network topology of the network system according to a shortest path algorithm, and Generating a link state database according to the received link state information sent by all the switches in the network system;
  • the controller generates a weighted directed graph according to the link state database, and according to the To the figure, each of the switches is used as a root node to generate a corresponding minimum spanning tree.
  • a switch When a switch receives a data packet, it determines whether the data packet matches the flow table. And when no, sending the data packet to the controller, where the data packet has a corresponding destination IP address; the controller root Destination IP address of the data packet corresponding to the minimum spanning tree of the query to obtain the forwarding path, the data packet is forwarded accordingly.
  • each of the switches sends its basic information message to the controller, so that the controller acquires the network system according to a shortest path algorithm.
  • a network topology and generating a link state database according to the received link state information sent by all the switches in the network system;
  • the controller generates a weighted directed graph according to the link state database, and And generating, according to the weighted directed graph, a corresponding minimum spanning tree;
  • a switch receives a data packet, determining whether the data packet matches the flow table, and if yes, according to the Performing, by the flow table, the data packet to be forwarded; if not, sending the data packet to the controller, where the data packet has a corresponding destination IP address;
  • the destination IP address in the data packet is queried to the corresponding minimum spanning tree to obtain a forwarding path, and the data packet is forwarded accordingly.
  • the configuration of all the switches of the present invention is centralized on the controller, which can simplify the configuration and management of the network
  • FIG. 1 shows a block diagram of a network system of the present invention in a specific embodiment.
  • FIG. 2 is a schematic diagram showing a weighted directed graph in accordance with an embodiment of the present invention.
  • FIG. 3 shows a schematic diagram of a minimum spanning tree in accordance with an embodiment of the present invention.
  • FIG. 4 shows a schematic diagram of a minimum spanning tree in accordance with an embodiment of the present invention.
  • FIG. 5 is a schematic diagram showing the state transition of a switch in an embodiment of the present invention.
  • FIG. 6 is a schematic diagram showing the workflow of a controller in an embodiment of the present invention.
  • FIG. 7 is a schematic diagram showing the workflow of a switch according to an embodiment of the present invention.
  • FIG. 8 is a schematic diagram showing the application of the network system of the present invention in a specific embodiment.
  • FIG. 9 is a flow chart showing a network operation method of the present invention in a specific embodiment.
  • the Open Shortest Path First (OSPF) protocol is the Open Shortest Path First Protocol. It is a dynamic routing protocol based on the link state of the autonomous system. It exchanges link state information between the three layers of switches to form a link state database, and then generates a routing table based on the database using the shortest path first algorithm.
  • Traditional OSPF is an internal gateway protocol used to route decisions within a single autonomous system. It is a routing protocol based on link status. The link status refers to the parameters of the router interface or link. The OSPF router is not telling other routers which networks and distances they can reach, but telling them the status of the network link, the network to which these interfaces are connected, and the cost of using these interfaces. Each router has its own link state called the local link state.
  • the present invention provides a network system, where the network system includes a controller and at least one switch, each of which has a corresponding flow table pre-stored, the flow table includes a packet header and an execution action;
  • the switch sends its own basic information packet to the controller, where the basic information packet includes the MAC address, port information, and link state information of the switch itself; the controller acquires the information according to a shortest path algorithm.
  • a network topology structure of the network system and generating a link state database according to the received link state information sent by all the switches in the network system, and generating a weighted directed graph according to the link state database, And generating, according to the weighted directed graph, each of the switches as a root node, respectively generating a corresponding minimum spanning tree; and when a switch receives a data packet having a destination IP address, determining Whether the data packet matches the corresponding flow table of the switch, and if so, the number is performed according to the execution action of the flow table Transmitting the packet to the controller, so that the controller queries the corresponding minimum spanning tree according to the destination IP address in the data packet to obtain the forwarding path. And forwarding the data packet accordingly.
  • FIG. 1 is a schematic diagram of a module in a specific embodiment of the network system of the present invention.
  • the network system 1 includes a controller 11 and at least one switch 12, wherein the controller 11 is an SDN (Software Defined Network) controller, and the switch 12 is an SDN (Software Defined Network).
  • the network) switch, and the SDN controller 11 and the SDN switch 12 communicate through the OpenFlow channel.
  • Each of the switches 12 prestores a corresponding flow table, the flow table includes a packet header and an execution action; the packet header has a destination IP address, and each of the switches 12 sends its own basic information report to the controller.
  • the basic information packet includes a MAC address, port information, and link state information of the switch itself, and the basic information packet may be a private experimenter (Experimenter) message.
  • Experimenter the format of the private extension Experimenter message from the SDN switch to the controller is shown in Table 1.
  • a Experimenter value of 255 requires an application to the ONF organization.
  • a Experimenter type value of 1 indicates that it is from the direction of the SDN switch to the controller.
  • the maximum number of reported port numbers is 128.
  • the link cost is shown in Table 1.
  • the link type is 1- Ethernet link 2-PON link.
  • the controller 11 preferably acquires the network topology of the network system 1 according to a shortest path algorithm and a link layer discovery protocol (LLDP).
  • the Shotest Routing algorithm is an easy-to-understand and widely used technique. Its basic idea is to create a subnet diagram in which each node represents a router, each arc represents a communication line (link), and the number on the arc represents the weight of the line. In order to select a routing path between a given pair of routers, the routing algorithm only needs to find the shortest path between the pair of nodes in the figure. There are several methods for path length measurement. One method is to calculate the number of stations. In addition, distance, channel bandwidth, average traffic, communication overhead, queue length, and propagation delay can be calculated.
  • the controller 11 further generates a link state database according to the received link state information sent by all the switches in the network system 1, and generates a weighted directed graph according to the link state database for clearing The path of signal transmission between each switch 12 and the link cost of the corresponding path are identified.
  • the controller 11 is predefined in the A link overhead formula, the controller generates the weighted directed graph according to the link state database and the link cost formula.
  • the network system 1 includes switches S0, S1, S2, and S3, and the weighted directed graph is shown in FIG. 2.
  • the link cost between S0 and S1 is 20, the link cost between S0 and S2 is 200, the link cost between S1 and S2 is 20000, and the link cost between S2 and S3 is 2000.
  • the controller 11 generates a corresponding minimum spanning tree according to each of the switches as a root node according to the weighted directed graph; for example, the switch S0 is used as a root node, and the controller 11 is generated as shown in FIG. 3
  • the minimum spanning tree shown, with the S1 as the root node, generates a minimum spanning tree as shown in FIG.
  • the switch 12 when the switch 12 receives a data packet having a destination IP address, determining whether the data packet matches the corresponding flow table of the switch 12, and if so, preferably according to the flow Executing the data packet to forward the data packet; if not, sending the data packet to the controller 11 to enable the controller 11 to query according to the destination IP address in the data packet A corresponding minimum spanning tree is obtained to obtain a forwarding path, and the data packet is forwarded accordingly.
  • the packet header of the flow table includes a destination IP address. When the destination IP address in the data packet is the same as the destination IP address included in the packet header of the flow table, the data packet is compared with the flow table. Matching; when the destination IP address in the data packet is different from the destination IP address included in the packet header of the flow table, the data packet does not match the flow table.
  • the controller 11 maintains a set of states for each of the switches 12.
  • the state set includes an initialization state (Init), a loading state (Loading), an interaction state (Exchange), and a completion state (Full).
  • Initialization state Init
  • Loading loading state
  • Exchange interaction state
  • Full completion state
  • FIG. 5 For the relationship between the states, refer to FIG. 5, in which, when the switch in the Init state receives the link state of the switch, it is converted into the Loading state, and when the switch in the Full state detects that the link state of the switch occurs.
  • the process flow diagram of the controller 11 is specifically referred to FIG. 6, specifically: extracting the MAC address, port, and link state information of the switch from the Experimenter message, and determining the setting in the controller.
  • the shortest path of the SDN and link status is turned on by the algorithm switch. If it is off, the operation ends; if it is on, the following operations are performed:
  • the controller generates a minimum spanning tree with each forwarding node as a root node
  • the process flow diagram of the switch 12 is specifically referred to FIG. 7.
  • the processing procedure is relatively simple, specifically: determining whether the received packet matches the Liu table. If yes, forward according to the flow table. If not, the message will be sent to the controller.
  • FIG. 8 a schematic diagram of an application of the network system of the present invention is shown, where the switch includes S0-S3.
  • the traffic sent from Host A to Host B is forwarded through the path of S1-S0-S2-S3 according to the network system.
  • the link from S0 to S2 is replaced with the link of 10M
  • Host A sends it to
  • the traffic of Host B will be forwarded through the path of S1-S3 (that is, the shortest path).
  • the present invention further provides a network operation method.
  • a flow chart of the network operation method in a specific embodiment is shown.
  • the method operates in the network system 1 as shown in FIG. 1, and a description about the technical solution of the network system 1 can be applied to the present embodiment.
  • the method includes the following steps:
  • Each of the switches sends its own basic information packet to the controller, where the basic information packet includes a MAC address, port information, and link state information of the switch itself.
  • S12 The controller acquires a network topology of the network system according to a shortest path algorithm, and generates a link state database according to the received link state information sent by all the switches in the network system.
  • the controller generates a weighted directed graph according to the link state database, and generates a corresponding minimum spanning tree according to each of the switches as a root node according to the weighted directed graph.
  • a switch When a switch receives a data packet, it determines whether the data packet matches the flow table, and if yes, preferably performs step S15; if not, executes step S16.
  • S15 Forward the data packet according to an execution action of the flow table.
  • S16 Send the data packet to the controller, where the data packet has a corresponding destination IP address, and the controller queries the corresponding minimum spanning tree according to the destination IP address in the data packet. The forwarding packet is obtained, and the data packet is forwarded accordingly.
  • the network system and the network operation method of the present invention enable each of the switches to send its own basic information message to the controller, so that the controller acquires the network system according to a shortest path algorithm.
  • a network topology and generating a link state database according to the received link state information sent by all the switches in the network system;
  • the controller generates a weighted directed graph according to the link state database, And generating, according to the weighted directed graph, a corresponding minimum spanning tree; when a switch receives a data packet, determining whether the data packet matches the flow table, and if so, according to the Performing an action of the flow table to forward the data packet; if not, sending the data message to the controller, where the data packet has a corresponding destination IP address;
  • the destination IP address in the data packet is queried to the corresponding minimum spanning tree to obtain a forwarding path, and the data packet is forwarded accordingly.
  • the configuration of all the switches of the present invention is centralized on the controller, which can simplify the configuration and management of

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本发明提供一种网络系统及网络运行方法，令每个交换机向控制器发送自身的基本信息报文，令控制器根据一最短路由算法，获取网络系统的网路拓扑结构，并生成相应的链路状态数据库；控制器根据链路状态数据库生成一带权有向图，且根据带权有向图，生成相应的最小生成树；当一交换机接收到一数据报文时，判断数据报文是否与流表相匹配，若是，则根据流表的执行动作对数据报文进行转发；若否，则将数据报文发送至控制器；控制器查询相应的最小生成树，以获取转发路径，将数据报文进行相应的转发。本发明的所有交换机的配置都集中在控制器上，且网络运行操作基本上由控制器执行，可以简化网络的配置以及管理，提高网络运行的效率。

Description

一种网络系统及网络运行方法
本申请要求2015年11月17日提交的申请号为：201510791993.3、发明名称为“一种网络系统及网络运行方法”的中国专利申请的优先权，其全部内容合并在此。
技术领域
本发明涉及电子通信领域，特别是涉及一种网络系统及网络运行方法。
背景技术
传统的OSPF(Open Shortest Path First，开放式最短路径优先)是一个内部网关协议，用于在单一自治系统内决策路由，它是基于链路状态的路由协议。OSPF协议是分布式的，配置相对复杂。由于网络区域划分和网络属性的复杂性，需要网络分析员有较高的网络知识水平才能配置和管理OSPF网络。OSPF需要交互很多种LSA(Link-State Advertisement，链路状态广播)报文，比较复杂并占用资源。
发明内容
鉴于以上所述现有技术的缺点，本发明的目的在于提供一种网络系统及网络运行方法，用于解决现有技术中开放式最短路径优先的配置以及管理操作复杂、效率低的问题。
为实现上述目的及其他相关目的，本发明提供一种网络系统，所述网络系统中包括一控制器以及至少一交换机，每个所述交换机中预存有相应的流表，所述流表包括包头以及执行动作；每个所述交换机向所述控制器发送自身的基本信息报文，所述基本信息报文包括交换机自身的MAC地址、端口信息、链路状态信息；所述控制器根据一最短路由算法，获取所述网络系统的网路拓扑结构，并根据接收的所述网络系统中的所有所述交换机发送的链路状态信息生成链路状态数据库，且根据所述链路状态数据库生成一带权有向图，并根据所述带权有向图，分别以每个所述交换机为根节点，分别生成相应的最小生成树；且，当一所述交换机接收到一具有目的IP地址的数据报文时，判断所述数据报文是否与相应的所述交换机的流表相匹配，且当为否时，将所述数据报文发送至所述控制器，以令所述控制器根据所述数据报文中的目的IP地址查询相应的最小生成树，以获取转发路径，将所述数据报文进行相应的转发。
于本发明的一具体实施例中，当判断所述数据报文与相应的所述交换机的流表相匹配时， 根据所述流表的执行动作对所述数据报文进行转发。
于本发明的一具体实施例中，所述控制器根据所述链路状态数据库生成一带权有向图，并根据所述带权有向图，分别以每个所述交换机为根节点，分别生成与每个所述交换机相应的最小生成树。
于本发明的一具体实施例中，所述交换机为软件定义网络交换机，所述控制器为软件定义网络控制器。
于本发明的一具体实施例中，每个所述软件定义网络交换机与所述软件定义网络控制器通过OpenFlow通道进行通信。
于本发明的一具体实施例中，所述基本信息报文为私有的实验者报文。
于本发明的一具体实施例中，所述流表的包头包括目的IP地址，当所述数据报文中的目的IP地址与所述流表的包头包括的目的IP地址相同时，所述数据报文与所述流表相匹配；当所述数据报文中的目的IP地址与所述流表的包头包括的目的IP地址不相同时，所述数据报文与所述流表不匹配。
于本发明的一具体实施例中，所述控制器根据所述最短路由算法以及一链路层发现协议获取所述网络系统的网路拓扑结构。
于本发明的一具体实施例中，所述控制器维护每个所述交换机的状态集。
于本发明的一具体实施例中，所述状态集包括初始化状态、加载状态、交互状态、以及完成状态。
于本发明的一具体实施例中，所述控制器中预先定义有一链路开销公式，所述控制器根据所述链路状态数据库以及所述链路开销公式生成所述带权有向图。
为实现上述目的及其他相关目的，本发明还提供一种网络运行方法，运行于如上任一项所述的网络系统中，所述方法包括以下步骤：每个所述交换机向所述控制器发送自身的基本信息报文，所述基本信息报文包括交换机自身的MAC地址、端口信息、链路状态信息；所述控制器根据一最短路由算法，获取所述网络系统的网路拓扑结构，并根据接收的所述网络系统中的所有所述交换机发送的链路状态信息生成链路状态数据库；所述控制器根据所述链路状态数据库生成一带权有向图，且根据所述带权有向图，分别以每个所述交换机为根节点，分别生成相应的最小生成树；当一所述交换机接收到一数据报文时，判断所述数据报文是否与所述流表相匹配，且当为否时，将所述数据报文发送至所述控制器，其中所述数据报文具有相应的目的IP地址；所述控制器根据所述数据报文中的目的IP地址查询相应的最小生成树，以获取转发路径，将所述数据报文进行相应的转发。
如上所述，本发明的网络系统及网络运行方法，令每个所述交换机向所述控制器发送自身的基本信息报文，令所述控制器根据一最短路由算法，获取所述网络系统的网路拓扑结构，并根据接收的所述网络系统中的所有所述交换机发送的链路状态信息生成链路状态数据库；所述控制器根据所述链路状态数据库生成一带权有向图，且根据所述带权有向图，生成相应的最小生成树；当一所述交换机接收到一数据报文时，判断所述数据报文是否与所述流表相匹配，若是，则根据所述流表的执行动作对所述数据报文进行转发；若否，则将所述数据报文发送至所述控制器，其中所述数据报文具有相应的目的IP地址；所述控制器根据所述数据报文中的目的IP地址查询相应的最小生成树，以获取转发路径，将所述数据报文进行相应的转发。本发明的所有交换机的配置都集中在控制器上，可以简化网络的配置以及管理，提高网络运行的效率。
附图说明
图1显示为本发明网络系统在一具体实施例中的模块示意图。
图2显示为本发明一具体实施例中的带权有向图的示意图。
图3显示为本发明一具体实施例中的最小生成树的示意图。
图4显示为本发明一具体实施例中的最小生成树的示意图。
图5显示为本发明一具体实施例中的一交换机的状态转化示意图。
图6显示为本发明一具体实施例中的控制器的工作流程示意图。
图7显示为本发明一具体实施例中的交换机的工作流程示意图。
图8显示为本发明的网络系统在一具体实施例中的应用示意图。
图9显示为本发明的网络运行方法在一具体实施例中的流程示意图。
元件标号说明
1               网络系统
11              控制器
12、S0～S3      交换机
S11～S16        步骤
具体实施方式
以下通过特定的具体实例说明本发明的实施方式，本领域技术人员可由本说明书所揭露的内容轻易地了解本发明的其他优点与功效。本发明还可以通过另外不同的具体实施方式加 以实施或应用，本说明书中的各项细节也可以基于不同观点与应用，在没有背离本发明的精神下进行各种修饰或改变。需说明的是，在不冲突的情况下，以下实施例及实施例中的特征可以相互组合。
需要说明的是，以下实施例中所提供的图示仅以示意方式说明本发明的基本构想，遂图示中仅显示与本发明中有关的组件而非按照实际实施时的组件数目、形状及尺寸绘制，其实际实施时各组件的型态、数量及比例可为一种随意的改变，且其组件布局型态也可能更为复杂。
OSPF(Open Shortest Path First)协议即“开放最短路径优先协议”。它是一种基于链路状态的自治系统内部的动态路由协议，它通过三层交换机间交换链路状态信息来组成一个链路状态数据库，然后基于这个数据库用最短路径优先算法生成路由表。传统的OSPF是一个内部网关协议，用于在单一自治系统内决策路由。它是基于链路状态的路由协议。链路状态是指路由器接口或链路的参数。OSPF路由器不是告知其他路由器可以到达哪些网络及距离是多少，而是告知它的网络链路状态，这些接口所连的网络及使用这些接口的费用。各个路由器都有其自身的链路状态称为本地链路状态。这些本地链路状态在OSPF路由域内传播直到所有的OSPF路由器都有完整而等同的链路状态数据库为止。一旦每个路由器都接收到所有的链路状态，每个路由器可以构造一棵树以它自己为根而分支表示到网络系统中所有网络的最短的或费用最低的路由。传统的OSPF协议是分布式的，配置相对复杂。由于网络区域划分和网络属性的复杂性，需要网络分析员有较高的网络知识水平才能配置和管理OSPF网络。OSPF需要交互很多种LSA报文，比较复杂并占用资源。针对这些缺点，本发明提出一种网络系统，所述网络系统中包括一控制器以及至少一交换机，每个所述交换机中预存有相应的流表，所述流表包括包头以及执行动作；每个所述交换机向所述控制器发送自身的基本信息报文，所述基本信息报文包括交换机自身的MAC地址、端口信息、链路状态信息；所述控制器根据一最短路由算法，获取所述网络系统的网路拓扑结构，并根据接收的所述网络系统中的所有所述交换机发送的链路状态信息生成链路状态数据库，且根据所述链路状态数据库生成一带权有向图，并根据所述带权有向图，分别以每个所述交换机为根节点，分别生成相应的最小生成树；且，当一所述交换机接收到一具有目的IP地址的数据报文时，判断所述数据报文是否与相应的所述交换机的流表相匹配，若是，则根据所述流表的执行动作对所述数据报文进行转发；若否，则将所述数据报文发送至所述控制器，以令所述控制器根据所述数据报文中的目的IP地址查询相应的最小生成树，以获取转发路径，将所述数据报文进行相应的转发。
具体的，请参阅图1，显示为本发明网络系统在一具体实施例中的模块示意图。所述网络系统1包括一控制器11以及至少一交换机12，其中，所述控制器11为SDN(Software Defined Network，软件定义网络)控制器，所述交换机12为SDN(Software Defined Network，软件定义网络)交换机，且，SDN控制器11与SDN交换机12通过OpenFlow通道进行通信。
每个所述交换机12中预存有相应的流表，所述流表包括包头以及执行动作；所述包头中具有目的IP地址，每个所述交换机12向所述控制器发送自身的基本信息报文，所述基本信息报文包括交换机自身的MAC地址、端口信息、链路状态信息，所述基本信息报文可以为私有的实验者(Experimenter)报文。例如，从SDN交换机到控制器的私有扩展Experimenter报文的格式如表1所示。Experimenter值为255需要向ONF组织申请。Experimenter type值为1表明是从SDN交换机方向到控制器。上报的端口号最大支持128个，链路开销如表一所示，其中，链路类型：1-以太网链路2-PON链路。
Figure PCTCN2016101394-appb-000001
表1
所述控制器11优选根据一最短路由算法以及一链路层发现协议(LLDP)获取所述网络系统1的网路拓扑结构。最短路由(Shotest Routing)算法是一种简单易懂而应用广泛的技术。它的基本思想是：建立一个子网图，图中每一个节点代表一台路由器，每条弧线代表一条通信线路(链路)，弧上的数字代表该线路的权重。为了在一对给定的路由器之间选择一条路由路径，路由算法只需在图中找到这对节点之间的最短路径即可。对于路径长度测量有多种方法，一种方法是计算站点数量，另外也可以计算距离、信道带宽、平均通信量、通信开销、队列长度、传播时延等。
所述控制器11还根据接收的所述网络系统1中的所有所述交换机发送的链路状态信息生成链路状态数据库，且根据所述链路状态数据库生成一带权有向图，用于清楚的标识各交换机12之间信号传输的路径以及相应路径的链路开销。优选的，所述控制器11中预先定义有 一链路开销公式，所述控制器根据所述链路状态数据库以及所述链路开销公式生成所述带权有向图。所述链路开销公式为“链路开销＝200Gbps/链路速率”，链路速率与链路开销根据所述公式形成的对应的数据表格如表1所示。于一具体实施例中，所述网络系统1中包括交换机S0、S1、S2、以及S3，所述带权有向图参阅图2所示。其中，S0与S1之间的链路开销为20，S0与S2之间的链路开销为200，S1与S2之间的链路开销为20000，S2与S3之间的链路开销为2000。
链路速率 链路开销
10M 20000
100M 2000
1G 200
10G 20
表1
且，所述控制器11根据所述带权有向图，分别以每个所述交换机为根节点，分别生成相应的最小生成树；例如以所述交换机S0为根节点，生成如图3所示的最小生成树，以所述S1为根节点，生成如图4所示的最小生成树。
且，当一所述交换机12接收到一具有目的IP地址的数据报文时，判断所述数据报文是否与相应的所述交换机12的流表相匹配，若是，则优选的根据所述流表的执行动作对所述数据报文进行转发；若否，则将所述数据报文发送至所述控制器11，以令所述控制器11根据所述数据报文中的目的IP地址查询相应的最小生成树，以获取转发路径，将所述数据报文进行相应的转发。其中，所述流表的包头包括目的IP地址，当所述数据报文中的目的IP地址与所述流表的包头包括的目的IP地址相同时，所述数据报文与所述流表相匹配；当所述数据报文中的目的IP地址与所述流表的包头包括的目的IP地址不相同时，所述数据报文与所述流表不匹配。
于本发明的一具体实施例中，所述控制器11维护每个所述交换机12的状态集。其中，所述状态集包括初始化状态(Init)、加载状态(Loading)、交互状态(Exchange)、以及完成状态(Full)。所述状态之间的关系参阅图5，其中，当处于Init状态的交换机接收到本交换机的链路状态时，转化为Loading状态，且当处于Full状态的交换机检测到到本交换机链路状态发生变化时，转化为Loading状态，且当处于Loading状态的交换机进行交互链路状态的操作时，转化为Exchange状态，且当处于Exchange状态的交换机完成所有链路状态的交互时，转化为Full状态，图5清楚的显示了各交换机的所有状态，且各状态间转化的条件。
于本发明的具体应用中，所述控制器11的处理流程示意图具体参阅图6，具体为：从Experimenter报文中提取交换机的MAC地址、端口、和链路状态信息，判断控制器内设定的SDN和链路状态的最短路由算法开关是否打开，若关闭，则结束操作；若打开，则执行以下操作：
根据LLDP协议计算网络拓扑结构；
根据全网的链路状态生成链路状态数据库；
根据链路状态数据库生成带权有向图；
控制器以每个转发节点为根节点生成最小生成树；
当有未匹配流表的报文时，将其根据报文的目的IP地址查询最小生成树，获得转发路径以进行转发，并下发流表。
于本发明的具体应用中，所述交换机12的处理流程示意图具体参阅图7，处理过程相对简单，具体为：判断接收的报文与刘表是否匹配。若是，根据流表进行转发。若否，报文将被送往控制器。
且，进一步参阅图8，显示为本发明的网络系统的应用示意图，所述交换机包括S0～S3。其中，从Host A发往Host B的流量根据本网络系统，会经过S1-S0-S2-S3的路径进行转发；当将S0到S2的链路替换为10M的链路时，Host A发往Host B的流量将经过S1-S3的路径(即最短路径)进行转发。
为实现上述目的及其他相关目的，本发明还提供一种网络运行方法，请参阅图9，显示为所述网络运行方法在一具体实施例中的流程示意图。所述方法运行于如图1所示的网络系统1中，且关于所述网络系统1的技术方案的描述均可应用于本实施例中。所述方法包括以下步骤：
S11：每个所述交换机向所述控制器发送自身的基本信息报文，所述基本信息报文包括交换机自身的MAC地址、端口信息、链路状态信息。
S12：所述控制器根据一最短路由算法，获取所述网络系统的网路拓扑结构，并根据接收的所述网络系统中的所有所述交换机发送的链路状态信息生成链路状态数据库。
S13：所述控制器根据所述链路状态数据库生成一带权有向图，且根据所述带权有向图，分别以每个所述交换机为根节点，分别生成相应的最小生成树。
S14：当一所述交换机接收到一数据报文时，判断所述数据报文是否与所述流表相匹配，若是，则优选的执行步骤S15；若否，则执行步骤S16。
S15：根据所述流表的执行动作对所述数据报文进行转发。
S16：将所述数据报文发送至所述控制器，其中所述数据报文具有相应的目的IP地址；所述控制器根据所述数据报文中的目的IP地址查询相应的最小生成树，以获取转发路径，将所述数据报文进行相应的转发。
综上所述，本发明的网络系统及网络运行方法，令每个所述交换机向所述控制器发送自身的基本信息报文，令所述控制器根据一最短路由算法，获取所述网络系统的网路拓扑结构，并根据接收的所述网络系统中的所有所述交换机发送的链路状态信息生成链路状态数据库；所述控制器根据所述链路状态数据库生成一带权有向图，且根据所述带权有向图，生成相应的最小生成树；当一所述交换机接收到一数据报文时，判断所述数据报文是否与所述流表相匹配，若是，则根据所述流表的执行动作对所述数据报文进行转发；若否，则将所述数据报文发送至所述控制器，其中所述数据报文具有相应的目的IP地址；所述控制器根据所述数据报文中的目的IP地址查询相应的最小生成树，以获取转发路径，将所述数据报文进行相应的转发。本发明的所有交换机的配置都集中在控制器上，可以简化网络的配置以及管理，提高网络运行的效率。所以，本发明有效克服了现有技术中的种种缺点而具高度产业利用价值。
上述实施例仅例示性说明本发明的原理及其功效，而非用于限制本发明。任何熟悉此技术的人士皆可在不违背本发明的精神及范畴下，对上述实施例进行修饰或改变。因此，举凡所属技术领域中具有通常知识者在未脱离本发明所揭示的精神与技术思想下所完成的一切等效修饰或改变，仍应由本发明的权利要求所涵盖。

Claims (12)

  1. 一种网络系统，其特征在于，所述网络系统中包括一控制器以及至少一交换机，每个所述交换机中预存有相应的流表，所述流表包括包头以及执行动作；
    每个所述交换机向所述控制器发送自身的基本信息报文，所述基本信息报文包括交换机自身的MAC地址、端口信息、链路状态信息；
    所述控制器根据一最短路由算法，获取所述网络系统的网路拓扑结构，并根据接收的所述网络系统中的所有所述交换机发送的链路状态信息生成链路状态数据库，且根据所述链路状态数据库计算生成与每个所述交换机相应的最小生成树；
    且，当一所述交换机接收到一具有目的IP地址的数据报文时，判断所述数据报文是否与相应的所述交换机的流表相匹配，且当为否时，将所述数据报文发送至所述控制器，以令所述控制器根据所述数据报文中的目的IP地址查询相应的最小生成树，以获取转发路径，将所述数据报文进行相应的转发。
  2. 根据权利要求1所述的网络系统，其特征在于：当判断所述数据报文与相应的所述交换机的流表相匹配时，根据所述流表的执行动作对所述数据报文进行转发。
  3. 根据权利要求1所述的网络系统，其特征在于：所述控制器根据所述链路状态数据库生成一带权有向图，并根据所述带权有向图，分别以每个所述交换机为根节点，分别生成与每个所述交换机相应的最小生成树。
  4. 根据权利要求1所述的网络系统，其特征在于：所述交换机为软件定义网络交换机，所述控制器为软件定义网络控制器。
  5. 根据权利要求4所述的网络系统，其特征在于：每个所述软件定义网络交换机与所述软件定义网络控制器通过OpenFlow通道进行通信。
  6. 根据权利要求4所述的网络系统，其特征在于：所述基本信息报文为私有的实验者报文。
  7. 根据权利要求1所述的网络系统，其特征在于：所述流表的包头包括目的IP地址，当所述数据报文中的目的IP地址与所述流表的包头包括的目的IP地址相同时，所述数据报文与所述流表相匹配；当所述数据报文中的目的IP地址与所述流表的包头包括的目的IP地址不相同时，所述数据报文与所述流表不匹配。
  8. 根据权利要求1所述的网络系统，其特征在于：所述控制器根据所述最短路由算法以及一链路层发现协议获取所述网络系统的网路拓扑结构。
  9. 根据权利要求1所述的网络系统，其特征在于：所述控制器维护每个所述交换机的状态集。
  10. 根据权利要求9所述的网络系统，其特征在于：所述状态集包括初始化状态、加载状态、交互状态、以及完成状态。
  11. 根据权利要求1所述的网络系统，其特征在于：所述控制器中预先定义有一链路开销公式，所述控制器根据所述链路状态数据库以及所述链路开销公式生成所述带权有向图。
  12. 一种网络运行方法，其特征在于：运行于如权利要求1～11中任一项所述的网络系统中，所述方法包括以下步骤：
    每个所述交换机向所述控制器发送自身的基本信息报文，所述基本信息报文包括交换机自身的MAC地址、端口信息、链路状态信息；
    所述控制器根据一最短路由算法，获取所述网络系统的网路拓扑结构，并根据接收的所述网络系统中的所有所述交换机发送的链路状态信息生成链路状态数据库；
    所述控制器根据所述链路状态数据库生成一带权有向图，且根据所述带权有向图，分别以每个所述交换机为根节点，分别生成相应的最小生成树；
    当一所述交换机接收到一数据报文时，判断所述数据报文是否与所述流表相匹配，且当为否时，将所述数据报文发送至所述控制器，其中所述数据报文具有相应的目的IP地址；
    所述控制器根据所述数据报文中的目的IP地址查询相应的最小生成树，以获取转发路径，将所述数据报文进行相应的转发。
PCT/CN2016/101394 2015-11-17 2016-09-30 一种网络系统及网络运行方法 WO2017084448A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510791993.3 2015-11-17
CN201510791993.3A CN105450447B (zh) 2015-11-17 2015-11-17 一种网络系统及网络运行方法

Publications (1)

Publication Number Publication Date
WO2017084448A1 true WO2017084448A1 (zh) 2017-05-26

Family

ID=55560253

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/101394 WO2017084448A1 (zh) 2015-11-17 2016-09-30 一种网络系统及网络运行方法

Country Status (2)

Country Link
CN (1) CN105450447B (zh)
WO (1) WO2017084448A1 (zh)

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105450447B (zh) * 2015-11-17 2018-05-18 上海斐讯数据通信技术有限公司 一种网络系统及网络运行方法
CN105871624B (zh) * 2016-05-24 2019-07-16 中国电子科技集团公司第三十研究所 不依赖于控制专网的动态sdn控制信令带内传输方法
CN106533940B (zh) * 2016-10-25 2019-10-11 上海斐讯数据通信技术有限公司 一种基于sdn的等价路由哈希选路方法、装置和系统
CN106993036B (zh) * 2017-03-30 2020-02-21 联想(北京)有限公司 运行环境信息处理方法及交换机
CN109039959B (zh) * 2018-07-27 2021-04-16 广东工业大学 一种sdn网络规则的一致性判断方法及相关装置
CN111865400B (zh) * 2020-08-06 2022-03-25 四川安迪科技实业有限公司 一种网状卫星网络抑制转发方法及装置

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1783836A (zh) * 2004-11-30 2006-06-07 Ut斯达康通讯有限公司 改进最短路径生成树算法的方法及其设备
CN101572674A (zh) * 2009-06-12 2009-11-04 杭州华三通信技术有限公司 一种路由计算方法及装置
CN102594689A (zh) * 2012-02-22 2012-07-18 中兴通讯股份有限公司 一种分布式网络控制方法及装置
CN103326943A (zh) * 2012-03-23 2013-09-25 日电（中国）有限公司 数据流重路由方法和控制器
CN103986661A (zh) * 2014-05-30 2014-08-13 上海珉智信息科技有限公司 一种支持网络编码的组播路由协议
CN104486218A (zh) * 2014-12-16 2015-04-01 上海斐讯数据通信技术有限公司 一种基于sdn的mpls网络控制系统及方法
CN104901890A (zh) * 2015-06-24 2015-09-09 上海斐讯数据通信技术有限公司 一种sdn的路由生成、匹配方法和系统
CN104954226A (zh) * 2015-07-28 2015-09-30 上海斐讯数据通信技术有限公司 一种基于SDN的支持QoS的通信隧道建立方法及系统
CN105049351A (zh) * 2015-07-07 2015-11-11 上海斐讯数据通信技术有限公司 基于sdn的多链接透明互联算法
CN105450447A (zh) * 2015-11-17 2016-03-30 上海斐讯数据通信技术有限公司 一种网络系统及网络运行方法

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100596117C (zh) * 2007-11-09 2010-03-24 杭州华三通信技术有限公司 一种路由信息传播方法和路由器
US8559314B2 (en) * 2011-08-11 2013-10-15 Telefonaktiebolaget L M Ericsson (Publ) Implementing OSPF in split-architecture networks
CN103001887B (zh) * 2012-11-22 2018-01-05 中兴通讯股份有限公司 一种链路保活方法、控制器及交换机

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1783836A (zh) * 2004-11-30 2006-06-07 Ut斯达康通讯有限公司 改进最短路径生成树算法的方法及其设备
CN101572674A (zh) * 2009-06-12 2009-11-04 杭州华三通信技术有限公司 一种路由计算方法及装置
CN102594689A (zh) * 2012-02-22 2012-07-18 中兴通讯股份有限公司 一种分布式网络控制方法及装置
CN103326943A (zh) * 2012-03-23 2013-09-25 日电（中国）有限公司 数据流重路由方法和控制器
CN103986661A (zh) * 2014-05-30 2014-08-13 上海珉智信息科技有限公司 一种支持网络编码的组播路由协议
CN104486218A (zh) * 2014-12-16 2015-04-01 上海斐讯数据通信技术有限公司 一种基于sdn的mpls网络控制系统及方法
CN104901890A (zh) * 2015-06-24 2015-09-09 上海斐讯数据通信技术有限公司 一种sdn的路由生成、匹配方法和系统
CN105049351A (zh) * 2015-07-07 2015-11-11 上海斐讯数据通信技术有限公司 基于sdn的多链接透明互联算法
CN104954226A (zh) * 2015-07-28 2015-09-30 上海斐讯数据通信技术有限公司 一种基于SDN的支持QoS的通信隧道建立方法及系统
CN105450447A (zh) * 2015-11-17 2016-03-30 上海斐讯数据通信技术有限公司 一种网络系统及网络运行方法

Also Published As

Publication number Publication date
CN105450447A (zh) 2016-03-30
CN105450447B (zh) 2018-05-18

Similar Documents

Publication Publication Date Title
US10673741B2 (en) Control device discovery in networks having separate control and forwarding devices
WO2017084448A1 (zh) 一种网络系统及网络运行方法
US10491519B2 (en) Routing method, device, and system
US9379975B2 (en) Communication control system, control server, forwarding node, communication control method, and communication control program
US20200396162A1 (en) Service function chain sfc-based communication method, and apparatus
KR102018395B1 (ko) 분할 아키텍쳐 네트워크에서의 패킷 방송 메카니즘
EP3474502B1 (en) Reduced configuration for multi-stage network fabrics
CN112448900B (zh) 一种数据传输方法及装置
US8934490B2 (en) Accelerated MAC address resolution for IPv6 traffic with IS-IS protocol
US8982733B2 (en) System and method for managing topology changes in a network environment
US9807035B1 (en) Using available bandwidths of an IP fabric to intelligently distribute data
CN106656793B (zh) 一种sdn网络与ip网络的数据交互方法
JP2017028698A (ja) トラフィック切り替え方法、デバイス、およびシステム
US20180324097A1 (en) Method for Implementing Load Balancing, Apparatus, and Network System
CN107294852B (zh) 一种使用拓扑分散短路径集的网络路由方法
CN105721321B (zh) 一种等价多路径的出接口更新方法及装置
US11290394B2 (en) Traffic control in hybrid networks containing both software defined networking domains and non-SDN IP domains
WO2015117412A1 (zh) 基于isis的洪泛方法及装置
CN105471747A (zh) 一种智能路由器选路方法和装置
US20150288595A1 (en) Control apparatus, communication system, control information creation method, and program
CN110380966A (zh) 一种发现转发路径的方法及其相关设备
US20210067438A1 (en) Multicast transmissions management
TWI556599B (zh) 網路控制方法及網路控制器
US9450835B2 (en) Method for turning off routers in a communications network and router implementing this method
CN105591940B (zh) 一种trill网络分发树选择方法和trill网络节点

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16865626

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 16865626

Country of ref document: EP

Kind code of ref document: A1