WO2015114714A1 - Network topology detection system, method and program, and control apparatus - Google Patents

Network topology detection system, method and program, and control apparatus Download PDF

Info

Publication number
WO2015114714A1
WO2015114714A1 PCT/JP2014/006140 JP2014006140W WO2015114714A1 WO 2015114714 A1 WO2015114714 A1 WO 2015114714A1 JP 2014006140 W JP2014006140 W JP 2014006140W WO 2015114714 A1 WO2015114714 A1 WO 2015114714A1
Authority
WO
WIPO (PCT)
Prior art keywords
connection
port
group
detection
protocol packet
Prior art date
Application number
PCT/JP2014/006140
Other languages
French (fr)
Japanese (ja)
Inventor
市川　富彦
Original Assignee
日本電気株式会社
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 日本電気株式会社 filed Critical 日本電気株式会社
Publication of WO2015114714A1 publication Critical patent/WO2015114714A1/en

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00—Data switching networks
    • H04L12/64—Hybrid switching systems
    • H04L12/6418—Hybrid transport

Definitions

  • the present invention relates to a network topology detection system, a network topology detection method, a network topology detection program, and a control device that detect connection of a network.
  • Patent Document 1 An example of a system that automatically detects a network topology and detects an erroneous connection is described in Patent Document 1.
  • Patent Document 1 describes a method of manually inputting connection information between devices of a network and a method of using a unique protocol provided by a device developer for network topology detection.
  • a general OpenFlow (OpenFlow) control device uses a protocol (connection detection protocol) that detects the connection state between OFS (OpenFlow Switch, OpenFlow switch) when automatically acquiring network topology information. To do. When a packet of the connection detection protocol is transmitted to port 1 of OFS1, if a corresponding packet is received from port 2 of OFS2, it is determined that there is a connection (link) from port 1 of OFS1 to port 2 of OFS2. .
  • connection detection protocol connection detection protocol
  • an OpenFlow control device (not shown) transmits a connection detection protocol packet to all ports of all OFS under control. Then, the OpenFlow control device confirms the connection states of all the ports based on whether or not a connection detection protocol packet has been received.
  • the OpenFlow control device transmits connection detection protocol packets to a total of 12 ports of 3 each for OFS 1, OFS 2, OFS 3, and OFS 4 under control. Since OFS1 port 1 is an external connection, the OpenFlow control device does not receive the connection detection protocol packet transmitted to OFS1 port 1 from anywhere. Therefore, it is determined that OFS1 port 1 is an external connection (1). Since the OFS 1 port 2 is connected to the OFS 2 port 1, the OpenFlow control device receives the connection detection protocol packet transmitted to the OFS 1 port 2 from the OFS 2 port 1. Therefore, it is determined that there is a connection (link) from OFS1 port 2 to OFS2 port 1 (2).
  • connection detection protocol packet transmitted to OFS2 port 1 from OFS1 port 2
  • connection detection protocol packet transmitted to OFS2 port 1 from OFS1 port 2
  • connection opposite to (2) there is also a connection from OFS2 port 1 to OFS1 port 2 (connection opposite to (2)).
  • the same determination is made for reception of the remaining connection detection protocol packets.
  • connection detection protocol is executed by transmitting a packet on a communication path used by a general user, if the frequency is high, it becomes a factor that hinders communication of the general user. Therefore, it is desirable that the number of packet transmissions of the connection detection protocol is as small as possible.
  • an object of the present invention is to provide a network topology detection system, method, program, and control device that reduce the number of times of transmission of connection detection protocol packets for detecting the topology of the network.
  • a network topology detection system includes a plurality of switches that transmit and receive packets to each other and a control device that controls the plurality of switches, and the control device does not detect a connection among ports of the plurality of switches.
  • Storage means for storing a plurality of connection undetected groups holding information indicating ports, and a plurality of connection detection groups holding information indicating ports detected to be connected among ports of a plurality of switches, and a connection undetected group
  • Topology management means for transmitting a connection detection protocol packet, which is a packet for detecting a connection state between the plurality of switches, to each port belonging to the group, and the topology management means includes a first connection undetected group. After sending a connection detection protocol packet to the first port belonging to If the second ports belonging to discovery group has received the connection detection protocol packets, and wherein the moving information indicating the information and the second port of a first port for connection detection group.
  • a control device is a control device that controls a plurality of switches that mutually transmit and receive packets, and that has a plurality of unconnected connections that hold information indicating a port from which no connection is detected among the ports of the plurality of switches.
  • a storage means for storing a plurality of connection detection groups that hold information indicating a detection detected port among the ports of the detection group and a plurality of switches, and a port belonging to the connection non-detection group between the plurality of switches.
  • Topology management means for transmitting a connection detection protocol packet, which is a packet for detecting a connection state, for each group, and the topology management means sends a connection detection protocol packet to the first port belonging to the first connection undetected group. Is transmitted from the second port belonging to the second undetected connection group.
  • a network topology detection method is a network topology detection method used in a network topology detection system including a plurality of switches that transmit and receive packets to each other and a control device that controls the plurality of switches.
  • a plurality of connection undetected groups that hold information indicating ports in which no connection is detected among switch ports, and a plurality of connection detection groups that hold information indicating ports in which connections are detected among ports of a plurality of switches
  • a connection detection protocol packet which is a packet for detecting a connection state between the plurality of switches, is transmitted for each group to a port belonging to the connection undetected group. After sending a connection detection protocol packet to port 1, When the connection detection protocol packet is received from the second port belonging to the connection undetected group, information indicating the first port and information indicating the second port are moved to the connection detection group.
  • a network topology detection program is a network topology detection program used in a network topology detection system including a plurality of switches that mutually transmit and receive packets and a control device that controls the plurality of switches.
  • a plurality of connection undetected groups that hold information indicating ports that are not detected in connection among a plurality of switch ports, and a plurality of information that indicate information indicating ports in which connections are detected among ports of a plurality of switches.
  • Run topology management processing When the connection detection protocol packet is received from the second port belonging to the second connection undetected group after transmitting the connection detection protocol packet to the first port belonging to the first connection undetected group, The information indicating the first port and the information indicating the second port are moved to the connection detection group.
  • connection detection protocol packet transmissions for detecting the topology of the network can be reduced.
  • FIG. 5 is an explanatory diagram showing a state of the system after a distribution operation after connection detection is performed on the other undetected group from the state of the system shown in FIG. 4.
  • the network topology detection system of the present embodiment uses this property to reduce the number of connection detection protocol packet transmissions.
  • FIG. 1 is a block diagram showing a configuration in an embodiment of a network topology detection system.
  • the network topology detection system of the present embodiment includes an OpenFlow control device 10 and a plurality of OFSs 20.
  • the OpenFlow control device 10 includes a topology management unit 11 that manages the network topology of the OFS 20 to be controlled, and a storage unit 13.
  • the storage unit 13 stores the topology management table 12.
  • the topology management table 12 holds network topology data.
  • the topology management table 12 includes information indicating the ports of the OFS 20, including a plurality of connection undetected groups including a port where connection is not detected, and a plurality of ports including a port where connection is detected among ports of a plurality of OpenFlow switches. Separately hold the connection detection group.
  • the topology management table 12 holds information indicating the ports of the OFS 20 in four groups of connection undetected A, connection undetected B, connection detected A, and connection detected B.
  • the topology management unit 11 transmits a connection detection protocol packet, which is a packet for detecting a connection state between the OFSs 20, to the ports of the OFS 20 belonging to the connection undetected group for each group.
  • the topology management unit 11 is realized by, for example, hardware designed to perform specific arithmetic processing or the like, or an information processing device such as a CPU (Central Processing Unit) that operates according to a program.
  • the program is stored in a non-transitory computer-readable storage medium.
  • the storage unit 13 is realized by a storage device such as an HDD (Hard disk drive).
  • the OFS 20 includes a connection detection protocol transmission unit 21 and a connection detection protocol reception unit 22.
  • the connection detection protocol transmission unit 21 transmits the connection detection protocol packet sent from the OpenFlow control device 10 to its own port to the other OFS 20.
  • connection detection protocol receiving unit 22 receives a connection detection protocol packet sent from another OFS 20 from its own port, and sends it to the OpenFlow control device 10.
  • FIG. 2 is a flowchart showing the operation of the connection detection phase in the embodiment of the network topology detection system.
  • the topology management unit 11 collectively distributes all ports of the OFS into the group of connection undetected A and connection undetected B in the order of recognition of the OFS (step A1).
  • FIG. 3 is an explanatory diagram showing the state of the system after this group distribution operation has been performed.
  • the topology management unit 11 transmits a connection detection protocol packet to the port of connection not detected A (step A2). Then, the topology management unit 11 waits for a predetermined time to receive the connection detection protocol packet (step A3). When the topology management unit 11 receives the connection detection protocol packet, the topology management unit 11 moves the transmission source port to the connection detection A, moves the reception destination port to the connection detection B, and determines that these two ports are connected. (Step A4). At this time, when the topology management unit 11 receives both directions, the topology management unit 11 determines a group based on the previously received direction. The topology management unit 11 does nothing if the connection detection protocol packet is not received.
  • FIG. 4 is an explanatory view showing a state after the distribution operation after the connection detection from step A2 to A4 is performed from the state shown in FIG.
  • connection not detected A includes six data of OFS 1 port 1, OFS 1 port 2, OFS 1 port 3, OFS 3 port 1, OFS 3 port 2, and OFS 3 port 3. Therefore, topology management unit 11 transmits a connection detection protocol packet to these six ports (step A2). Since OFS1 port 1 is an external connection, topology management unit 11 does not receive the connection detection protocol packet transmitted to OFS1 port 1 (“not received” in step A3).
  • OFS1 port 1 remains in connection undetected A as it is. Since the OFS1 port 2 is connected to the OFS2 port 1, the topology management unit 11 receives the connection detection protocol packet transmitted to the OFS1 port 2 from the OFS2 port 1 (“received” in step A3). In this case, topology management unit 11 moves OFS1 port 2 to connection detection A and moves OFS2 port 1 to connection detection B (step A4).
  • the topology management unit 11 receives the connection detection protocol packet transmitted to the OFS 1 port 3 from the OFS 3 port 3. Further, the topology management unit 11 receives the connection detection protocol packet transmitted to the OFS 3 port 3 from the OFS 1 port 3.
  • topology management unit 11 moves OFS1 port 3 to connection detection A, moves OFS3 port 3 to connection detection B, and determines that these two ports are connected. (Step A4). The same processing is performed for the reception of the remaining connection detection protocol packets, and the ports of the OFS belonging to each group are in the state shown in FIG.
  • the topology management unit 11 transmits a connection detection protocol packet to the port of connection undetected B (step A5). Then, topology management unit 11 waits for a certain period of time to receive the connection detection protocol packet (step A6). When the topology management unit 11 receives the connection detection protocol packet, the topology management unit 11 moves the transmission source port to the connection detection B, moves the reception destination port to the connection detection A, and determines that these two ports are connected. (Step A7). When the topology management unit 11 receives a connection detection protocol packet between two ports in both directions, the topology management unit 11 determines a group based on the previously received direction. The topology management unit 11 does nothing if the connection detection protocol packet is not received. Thereafter, the processing after step A2 is repeated.
  • FIG. 5 is an explanatory diagram showing the state of the system after performing the operations from steps A5 to A7 from the state shown in FIG.
  • the connection undetected B includes four of the OFS 2 port 2, the OFS 2 port 3, the OFS 4 port 2, and the OFS 4 port 3. Therefore, topology management unit 11 transmits a connection detection protocol packet to these four ports (step A5). Since OFS2 port 2 is an external connection, topology management unit 11 does not receive the connection detection protocol packet transmitted to OFS2 port 2 (“not received” in step A6). Therefore, OFS2 port 2 remains in connection undetected B as it is. Since the OFS 2 port 3 is connected to the OFS 4 port 3, the topology management unit 11 receives the connection detection protocol packet transmitted to the OFS 2 port 3 from the OFS 4 port 3 (“received” in step A 6).
  • the topology management unit 11 receives the connection detection protocol packet transmitted to the OFS 4 port 3 from the OFS 2 port 3 (“received” in step A 6). If the reception from the OFS2 port 3 is early, the topology management unit 11 assigns the OFS4 port 3 to the connection detection B, assigns the OFS2 port 3 to the connection detection A, and determines that these two ports are connected ( Step A7). The same processing is performed for reception of the remaining connection detection protocol packets, and the topology management table 12 is in the state shown in FIG.
  • the connection detection protocol packet is transmitted to all ports as in the example shown in FIG. 11, the number of packet transmissions in the connection detection phase is 12, which is the total number of ports. Therefore, the network topology detection system can reduce the number of times of connection detection protocol packet transmission compared to a system that transmits ports to all ports without grouping the ports.
  • FIG. 6 is a flowchart showing the operation in the connection maintenance confirmation phase of the network topology detection system.
  • the topology management unit 11 transmits a connection detection protocol packet to the connection detection A port (step B1). Then, topology management unit 11 waits for a certain period of time to receive the connection detection protocol packet (step B2). When the topology management unit 11 receives the connection detection protocol packet, the topology management unit 11 checks whether the destination port has changed (step B3). If there is no change, one connection maintenance confirmation is completed. When the topology management unit 11 does not receive the connection detection protocol packet from anywhere, the topology management unit 11 moves each port from the connection detection A to the connection non-detection A, and moves from the connection detection B to the connection non-detection B (step). B4).
  • topology management unit 11 moves the old destination port to the connection undetected B and moves the new destination port to the connection detection B.
  • topology management unit 11 moves the source port to connection undetected A (step B5).
  • FIG. 7 is an explanatory diagram showing a specific operation in the connection maintenance confirmation phase.
  • connection detection A since connection detection A includes four ports, OFS1 port 2, OFS1 port 3, OFS3 port 2, and OFS2 port 3, connection detection protocol packets are transmitted to these four ports (step B1). If the connection between OFS1 port 2 and OFS2 port 1 is maintained, topology management unit 11 receives the connection detection protocol packet transmitted to OFS1 port 2 from OFS2 port 1 (“received” in step B2). Since there is no change in the destination port (NO in step B3), the connection maintenance confirmation is completed. The topology management unit 11 performs the same process for the remaining three ports.
  • the topology management unit 11 receives the connection detection protocol packet transmitted to the OFS1 port 2 from the OFS2 port 1, thereby confirming that the connection between the OFS1 port 2 and the OFS2 port 1 is maintained. .
  • the detected ports are distributed to different groups in the connection detection phase, it can be considered that all connection maintenance confirmations have been made by checking only one group.
  • the network topology detection system confirms connection from the opposite direction after a predetermined period of time when confirmation of connection from one direction is performed. That is, the network topology detection system alternately performs the operation shown in FIG. 6 and the reverse operation shown in FIG.
  • an operation indicating confirmation of connection from the reverse direction in the connection maintenance confirmation phase will be described.
  • FIG. 8 is a flowchart showing the operation of the connection maintenance confirmation phase in the reverse direction.
  • the topology management unit 11 transmits a connection detection protocol packet to the connection detection B port (step B6). Then, topology management unit 11 waits for a predetermined time to receive the connection detection protocol packet (step B7). When the topology management unit 11 receives the connection detection protocol packet, the topology management unit 11 checks whether the destination port has changed (step B8). If there is no change, one connection maintenance confirmation is completed. When the topology management unit 11 does not receive the connection detection protocol packet from anywhere, the topology management unit 11 moves each port from the connection detection B to the connection non-detection B, and moves from the connection detection A to the connection non-detection A (step). B9).
  • topology management unit 11 moves the old destination port to the connection not detected A and moves the new destination port to the connection detection A.
  • topology management unit 11 moves the port to connection undetected B (step B10).
  • connection detection B includes four ports, OFS2 port 1, OFS3 port 3, OFS4 port 1, and OFS4 port 3, topology management unit 11 transmits a connection detection protocol packet to these four ports (step B6). If the connection between OFS2 port 1 and OFS1 port 2 is maintained, topology management unit 11 receives the connection detection protocol packet transmitted to OFS2 port 1 from OFS1 port 2 (“received” in step B7). Since there is no change in the destination port (NO in step B8), the connection maintenance confirmation is completed. The topology management unit 11 performs the same process for the remaining three ports.
  • the number of packet transmissions in the connection maintenance confirmation phase is 4 in the configuration example shown in FIG.
  • the number of packet transmissions in the connection maintenance confirmation phase is 8. Therefore, the network topology detection system can reduce the number of transmissions of connection detection protocol packets for connection maintenance confirmation, compared to a method of transmitting ports to all ports without grouping the ports. In particular, in the connection detection phase, since the detected ports are allocated to different groups, the number of protocol packet transmissions can be more reliably reduced in the connection maintenance confirmation phase.
  • FIG. 9 is an explanatory diagram showing another configuration example of the network topology detection system of the present embodiment.
  • the connection between OFS1 and OFS4 and the connection between OFS2 and OFS3 are increased compared to the configuration shown in FIG.
  • the effect of the connection detection phase is further increased.
  • the network topology detection system of the present embodiment can reduce the number of connection detection protocol packet transmissions by managing network topology detection system ports in a plurality of groups and performing connection detection for each group. The reason is that the network topology detection system of the present embodiment does not detect in the reverse direction, assuming that a bidirectional connection is possible if a one-way connection is detected between certain ports.
  • the way of allocating unconnected groups before detection is not limited to the way shown in step A1 and FIG. It is preferable that the storage unit 13 store the ports for which connection detection is performed in different groups as much as possible, and store the ports for which connection detection is not performed in the same group.
  • the topology management unit 11 transmits connection detection protocol packets to the two ports in a lump, and therefore the number of detections between the ports. This is because it cannot be reduced.
  • the storage unit 13 since the ports included in the same OFS are not connected, stores the ports included in the same OFS in the same connection undetected group.
  • the group is divided into four groups, but the number of groups is not necessarily four.
  • the topology management unit 11 may further divide the connection undetected A and the connection undetected B into two groups so as to be divided into different groups for each OFS. As a result, the number of detections can be further reduced. However, since the time required for the detection process increases as the number of groups increases, it is preferable to determine the number of groups in consideration of the time.
  • connection detection protocol used in the network topology detection system of this embodiment is assumed to be used as a unique protocol provided by the device developer.
  • OpenFlow a network in which only devices conforming to the same OpenFlow standard is generally configured, so that the topology of the entire network can be detected using the unique protocol.
  • the network topology detection system of the present embodiment can be used for networks other than networks using the open flow technology.
  • the Openflow device 10 may be a device that controls a switch, and is realized using a general server or the like.
  • the OFS 20 may be a switch that can switch the communication path.
  • FIG. 10 is a block diagram showing the configuration of the main part of the network topology detection system according to the present invention.
  • the network topology detection system according to the present invention includes, as main components, a plurality of switches 40 that mutually transmit and receive packets, and a control device 30 that controls the plurality of switches 40.
  • the control device 30 indicates a plurality of connection undetected groups that hold information indicating ports for which connections are not detected among the ports of the plurality of switches 40, and ports for which connections are detected among the ports of the plurality of switches 40.
  • Storage means 33 for storing a plurality of connection detection groups holding information, and connection detection protocol packets, which are packets for detecting the connection state between the plurality of switches 40, for each port belonging to the connection undetected group.
  • Topology management means 31 for transmitting to the network.
  • the topology management unit 31 receives the connection detection protocol packet from the second port belonging to the second connection undetected group after transmitting the connection detection protocol packet to the first port belonging to the first connection undetected group. In this case, the information indicating the first port and the information indicating the second port are moved to the connection detection group.
  • network topology detection systems shown in the following (1) to (4) are also disclosed.
  • a plurality of switches (for example, OFS 20) that transmit and receive packets to and from each other and a control device (for example, OpenFlow control device 10) that controls the plurality of switches.
  • a connection is detected among a plurality of connection undetected groups (for example, a group of connection undetected A, a group of connection undetected B) that holds information indicating ports for which connection is not detected, and a plurality of switch ports.
  • Storage means for example, the storage unit 13 for storing a plurality of connection detection groups (for example, a group of connection detection A and a group of connection detection B) that hold information indicating the connected ports, and ports belonging to the connection undetected group , A connection detection protocol packet that is a packet for detecting a connection state between the plurality of switches is transmitted for each group.
  • Topology management means for example, topology management unit 11
  • the topology management means is a first port (for example, OFS1 port 2) belonging to the first connection undetected group (for example, group of connection undetected A).
  • connection detection protocol packet When the connection detection protocol packet is received from the second port (for example, OFS2 port 1) belonging to the second connection undetected group (for example, the group of connection undetected B) after transmitting the connection detection protocol packet to A network topology detection system that moves information indicating the first port and information indicating the second port to a connection detection group.
  • the second port for example, OFS2 port 1
  • the second connection undetected group for example, the group of connection undetected B
  • the topology management means transmits a connection detection protocol packet to the first port (for example, OFS1 port 2) belonging to the first connection undetected group (for example, connection not detected A).
  • a second port for example, OFS2 port 1 belonging to the second connection undetected group (for example, connection not detected B)
  • information indicating the first port is displayed.
  • Move to a first connection detection group for example, connection detection A
  • move information indicating the second port to a second connection detection group for example, connection detection B
  • the first port By transmitting a connection detection protocol packet to the first port and receiving the connection detection protocol packet from the second port. It may be configured to verify that the connection between the second port is maintained.
  • the detected ports are distributed to different groups, so it can be considered that all connection maintenance confirmations have been made only by checking one group. .
  • the topology management means transmits a connection detection protocol packet to the first port (for example, OFS1 port 2), and the connection detection protocol from the second port (for example, OFS2 port 1). After a predetermined period in which the connection maintenance between the first port and the second port is confirmed by receiving a packet, the connection maintenance between the first port and the second port is performed again.
  • the connection between the first port and the second port is transmitted by transmitting a connection detection protocol packet to the second port and receiving the connection detection protocol packet from the first port. It may be configured to confirm maintenance.
  • the number of packet detections can be reduced even in a system in which a failure may occur only in one-way communication, such as when an optical cable is used for a network line.
  • a decrease in accuracy of connection maintenance confirmation can be suppressed.
  • ports of the same switch for example, OFS1, OFS2, OFS3, or OFS4 are stored in the same connection undetected group (for example, connection undetected A or connection undetected B). It may be configured as follows. According to such a network topology detection system, the number of connection detection protocol packet transmissions can be reduced more effectively.
  • the present invention can be applied to a network using OpenFlow, for example, and is not limited to OpenFlow, and can be applied to detection of a general network topology.
  • Topology management unit 10 OpenFlow control device 11 Topology management unit 12 Topology management table 13 Storage unit 20 OFS 21 Connection Detection Protocol Transmitting Unit 22 Connection Detection Protocol Receiving Unit 30 Control Device 31 Topology Management Unit 33 Storage Unit 40 Switch

Landscapes

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

Abstract

A network topology detection system is provided with a plurality of switches (40) and a control apparatus (30). The control apparatus (30) includes: a storage means (33) that stores a plurality of connection-undetected groups and a plurality of connection-detected groups; and a topology management means (31) that transmits a connection detection protocol packet to a port belonging to each of the connection-undetected groups. When the topology management means (31), after transmitting the connection detection protocol packet to a first port belonging to a first connection-undetected group, receives the packet from a second port belonging to a second connection-undetected group, the topology management means (31) causes information indicating the first port and information indicating the second port to be shifted to the connection-detected groups.

Description

ネットワークトポロジー検出システム、方法およびプログラムならびに制御装置Network topology detection system, method, program, and control apparatus
　本発明は、ネットワークの接続検出を行うネットワークトポロジー検出システム、ネットワークトポロジー検出方法およびネットワークトポロジー検出プログラムならびに制御装置に関する。 The present invention relates to a network topology detection system, a network topology detection method, a network topology detection program, and a control device that detect connection of a network.
　ネットワークトポロジーを自動で検出し、誤接続の検出を行うシステムの一例が、特許文献１に記載されている。特許文献１には、ネットワークのトポロジー検出について、ネットワークの各装置間の接続情報を手動で入力する方法、および装置の開発元により提供される独自のプロトコルを利用する方法が記載されている。 An example of a system that automatically detects a network topology and detects an erroneous connection is described in Patent Document 1. Patent Document 1 describes a method of manually inputting connection information between devices of a network and a method of using a unique protocol provided by a device developer for network topology detection.
特許第３８８８９２８号公報（段落０００２）Japanese Patent No. 3888928 (paragraph 0002)
　一般的なＯｐｅｎＦｌｏｗ（オープンフロー）制御装置は、ネットワークのトポロジー情報を自動的に取得する際には、ＯＦＳ（ＯｐｅｎＦｌｏｗ　Ｓｗｉｔｃｈ、オープンフロースイッチ）間の接続状態を検出するプロトコル（接続検出プロトコル）を使用する。接続検出プロトコルのパケットをＯＦＳ１のポート１へ送信した際に、ＯＦＳ２のポート２から該当パケットを受信した場合に、ＯＦＳ１のポート１からＯＦＳ２のポート２への接続（リンク）があると判断される。 A general OpenFlow (OpenFlow) control device uses a protocol (connection detection protocol) that detects the connection state between OFS (OpenFlow Switch, OpenFlow switch) when automatically acquiring network topology information. To do. When a packet of the connection detection protocol is transmitted to port 1 of OFS1, if a corresponding packet is received from port 2 of OFS2, it is determined that there is a connection (link) from port 1 of OFS1 to port 2 of OFS2. .
　一般的な方法では、図１１に示すように、ＯｐｅｎＦｌｏｗ制御装置（図示せず）は、制御下の全ＯＦＳの全ポートへ接続検出プロトコルのパケットを送信する。そして、ＯｐｅｎＦｌｏｗ制御装置は、その全ポートの接続状態を、接続検出プロトコルパケットの受信有無によりそれぞれ確認する。 In a general method, as shown in FIG. 11, an OpenFlow control device (not shown) transmits a connection detection protocol packet to all ports of all OFS under control. Then, the OpenFlow control device confirms the connection states of all the ports based on whether or not a connection detection protocol packet has been received.
　具体的には、図１１において、ＯｐｅｎＦｌｏｗ制御装置は、制御下のＯＦＳ１、ＯＦＳ２、ＯＦＳ３、ＯＦＳ４に各３個ずつの合計１２個のポートへ接続検出プロトコルパケットを送信する。ＯＦＳ１ポート１は外部接続であるので、ＯｐｅｎＦｌｏｗ制御装置は、ＯＦＳ１ポート１へ送信した接続検出プロトコルパケットをどこからも受信しない。したがって、ＯＦＳ１ポート１は、外部接続であると判断される（１）。ＯＦＳ１ポート２はＯＦＳ２ポート１と接続されているので、ＯｐｅｎＦｌｏｗ制御装置は、ＯＦＳ１ポート２へ送信した接続検出プロトコルパケットをＯＦＳ２ポート１から受信する。したがって、ＯＦＳ１ポート２からＯＦＳ２ポート１への接続（リンク）があると判断される（２）。逆に、ＯｐｅｎＦｌｏｗ制御装置は、ＯＦＳ２ポート１へ送信した接続検出プロトコルパケットを、ＯＦＳ１ポート２から受信するので、ＯＦＳ２ポート１からＯＦＳ１ポート２への接続（（２）と逆方向の接続）もあると判断される（４）。残りの接続検出プロトコルパケットの受信についても同様に判断される。 Specifically, in FIG. 11, the OpenFlow control device transmits connection detection protocol packets to a total of 12 ports of 3 each for OFS 1, OFS 2, OFS 3, and OFS 4 under control. Since OFS1 port 1 is an external connection, the OpenFlow control device does not receive the connection detection protocol packet transmitted to OFS1 port 1 from anywhere. Therefore, it is determined that OFS1 port 1 is an external connection (1). Since the OFS 1 port 2 is connected to the OFS 2 port 1, the OpenFlow control device receives the connection detection protocol packet transmitted to the OFS 1 port 2 from the OFS 2 port 1. Therefore, it is determined that there is a connection (link) from OFS1 port 2 to OFS2 port 1 (2). Conversely, since the OpenFlow control device receives the connection detection protocol packet transmitted to OFS2 port 1 from OFS1 port 2, there is also a connection from OFS2 port 1 to OFS1 port 2 (connection opposite to (2)). (4). The same determination is made for reception of the remaining connection detection protocol packets.
　しかし、接続検出プロトコルは、一般ユーザが使用する通信路上にパケットを送信することにより実行されるので、頻度が高いと一般ユーザの通信を阻害する要因となる。したがって、接続検出プロトコルのパケット送信回数はなるべく少ない方が望ましい。 However, since the connection detection protocol is executed by transmitting a packet on a communication path used by a general user, if the frequency is high, it becomes a factor that hinders communication of the general user. Therefore, it is desirable that the number of packet transmissions of the connection detection protocol is as small as possible.
　そこで、本発明は、ネットワークのトポロジーを検出するための接続検出プロトコルパケットの送信回数を削減するネットワークトポロジー検出システム、方法およびプログラムならびに制御装置を提供することを目的とする。 Therefore, an object of the present invention is to provide a network topology detection system, method, program, and control device that reduce the number of times of transmission of connection detection protocol packets for detecting the topology of the network.
　本発明によるネットワークトポロジー検出システムは、相互にパケットの送受信を行う複数のスイッチと、複数のスイッチを制御する制御装置とを備え、制御装置は、複数のスイッチのポートのうち接続が検出されていないポートを示す情報を保持する複数の接続未検出グループ、および複数のスイッチのポートのうち接続が検出されたポートを示す情報を保持する複数の接続検出グループを記憶する記憶手段と、接続未検出グループに属するポートに、当該複数のスイッチ間の接続状態を検出するためのパケットである接続検出プロトコルパケットをグループ毎に送信するトポロジー管理手段とを含み、トポロジー管理手段は、第１の接続未検出グループに属する第１のポートに接続検出プロトコルパケットを送信した後に、第２の接続未検出グループに属する第２のポートから当該接続検出プロトコルパケットを受信した場合に、当該第１のポートを示す情報および当該第２のポートを示す情報を接続検出グループに移動させることを特徴とする。 A network topology detection system according to the present invention includes a plurality of switches that transmit and receive packets to each other and a control device that controls the plurality of switches, and the control device does not detect a connection among ports of the plurality of switches. Storage means for storing a plurality of connection undetected groups holding information indicating ports, and a plurality of connection detection groups holding information indicating ports detected to be connected among ports of a plurality of switches, and a connection undetected group Topology management means for transmitting a connection detection protocol packet, which is a packet for detecting a connection state between the plurality of switches, to each port belonging to the group, and the topology management means includes a first connection undetected group. After sending a connection detection protocol packet to the first port belonging to If the second ports belonging to discovery group has received the connection detection protocol packets, and wherein the moving information indicating the information and the second port of a first port for connection detection group.
　本発明による制御装置は、相互にパケットの送受信を行う複数のスイッチを制御する制御装置であって、複数のスイッチのポートのうち接続が検出されていないポートを示す情報を保持する複数の接続未検出グループ、および複数のスイッチのポートのうち接続が検出されたポートを示す情報を保持する複数の接続検出グループを記憶する記憶手段と、接続未検出グループに属するポートに、当該複数のスイッチ間の接続状態を検出するためのパケットである接続検出プロトコルパケットをグループ毎に送信するトポロジー管理手段とを備え、トポロジー管理手段は、第１の接続未検出グループに属する第１のポートに接続検出プロトコルパケットを送信した後に、第２の接続未検出グループに属する第２のポートから当該接続検出プロトコルパケットを受信した場合に、当該第１のポートを示す情報および当該第２のポートを示す情報を接続検出グループに移動させることを特徴とする。 A control device according to the present invention is a control device that controls a plurality of switches that mutually transmit and receive packets, and that has a plurality of unconnected connections that hold information indicating a port from which no connection is detected among the ports of the plurality of switches. A storage means for storing a plurality of connection detection groups that hold information indicating a detection detected port among the ports of the detection group and a plurality of switches, and a port belonging to the connection non-detection group between the plurality of switches. Topology management means for transmitting a connection detection protocol packet, which is a packet for detecting a connection state, for each group, and the topology management means sends a connection detection protocol packet to the first port belonging to the first connection undetected group. Is transmitted from the second port belonging to the second undetected connection group. When receiving the Rupaketto, characterized in that moving the information indicating the information and the second port of a first port for connection detection group.
　本発明によるネットワークトポロジー検出方法は、相互にパケットの送受信を行う複数のスイッチと、複数のスイッチを制御する制御装置とを備えたネットワークトポロジー検出システムに用いられるネットワークトポロジー検出方法であって、複数のスイッチのポートのうち接続が検出されていないポートを示す情報を保持する複数の接続未検出グループ、および複数のスイッチのポートのうち接続が検出されたポートを示す情報を保持する複数の接続検出グループを記憶し、接続未検出グループに属するポートに、当該複数のスイッチ間の接続状態を検出するためのパケットである接続検出プロトコルパケットをグループ毎に送信し、第１の接続未検出グループに属する第１のポートに接続検出プロトコルパケットを送信した後に、第２の接続未検出グループに属する第２のポートから当該接続検出プロトコルパケットを受信した場合に、当該第１のポートを示す情報および当該第２のポートを示す情報を接続検出グループに移動させることを特徴とする。 A network topology detection method according to the present invention is a network topology detection method used in a network topology detection system including a plurality of switches that transmit and receive packets to each other and a control device that controls the plurality of switches. A plurality of connection undetected groups that hold information indicating ports in which no connection is detected among switch ports, and a plurality of connection detection groups that hold information indicating ports in which connections are detected among ports of a plurality of switches A connection detection protocol packet, which is a packet for detecting a connection state between the plurality of switches, is transmitted for each group to a port belonging to the connection undetected group. After sending a connection detection protocol packet to port 1, When the connection detection protocol packet is received from the second port belonging to the connection undetected group, information indicating the first port and information indicating the second port are moved to the connection detection group. And
　本発明によるネットワークトポロジー検出プログラムは、相互にパケットの送受信を行う複数のスイッチと、複数のスイッチを制御する制御装置とを備えたネットワークトポロジー検出システムに用いられるネットワークトポロジー検出プログラムであって、コンピュータに、複数のスイッチのポートのうち接続が検出されていないポートを示す情報を保持する複数の接続未検出グループ、および複数のスイッチのポートのうち接続が検出されたポートを示す情報を保持する複数の接続検出グループを記憶する記憶処理と、接続未検出グループに属するポートに、当該複数のスイッチ間の接続状態を検出するためのパケットである接続検出プロトコルパケットをグループ毎に送信するトポロジー管理処理とを実行させ、トポロジー管理処理で、第１の接続未検出グループに属する第１のポートに接続検出プロトコルパケットを送信した後に、第２の接続未検出グループに属する第２のポートから当該接続検出プロトコルパケットを受信した場合に、当該第１のポートを示す情報および当該第２のポートを示す情報を接続検出グループに移動させることを特徴とする。 A network topology detection program according to the present invention is a network topology detection program used in a network topology detection system including a plurality of switches that mutually transmit and receive packets and a control device that controls the plurality of switches. A plurality of connection undetected groups that hold information indicating ports that are not detected in connection among a plurality of switch ports, and a plurality of information that indicate information indicating ports in which connections are detected among ports of a plurality of switches Storage processing for storing a connection detection group, and topology management processing for transmitting a connection detection protocol packet, which is a packet for detecting a connection state between the plurality of switches, to a port belonging to a connection undetected group for each group. Run topology management processing When the connection detection protocol packet is received from the second port belonging to the second connection undetected group after transmitting the connection detection protocol packet to the first port belonging to the first connection undetected group, The information indicating the first port and the information indicating the second port are moved to the connection detection group.
　本発明によれば、ネットワークのトポロジーを検出するための接続検出プロトコルパケットの送信回数を削減することができる。 According to the present invention, the number of connection detection protocol packet transmissions for detecting the topology of the network can be reduced.
本発明によるネットワークトポロジー検出システムの実施形態における構成を示すブロック図である。It is a block diagram which shows the structure in embodiment of the network topology detection system by this invention. 本発明によるネットワークトポロジー検出システムの実施形態における接続検出フェーズの動作を示すフローチャートである。It is a flowchart which shows the operation | movement of the connection detection phase in embodiment of the network topology detection system by this invention. グループ振り分け動作が行われた後のシステムの状態を示す説明図である。It is explanatory drawing which shows the state of the system after group distribution operation | movement was performed. 図３に示すシステムの状態から、一方の接続未検出グループに対する接続検出後の振り分け動作が行われた後のシステムの状態を示す説明図である。It is explanatory drawing which shows the state of the system after the distribution operation after the connection detection with respect to one connection undetected group was performed from the state of the system shown in FIG. 図４に示すシステムの状態から、他方の接続未検出グループに対する接続検出後の振り分け動作が行われた後のシステムの状態を示す説明図である。FIG. 5 is an explanatory diagram showing a state of the system after a distribution operation after connection detection is performed on the other undetected group from the state of the system shown in FIG. 4. 本発明によるネットワークトポロジー検出システムの実施形態における接続維持確認フェーズの動作を示すフローチャートである。It is a flowchart which shows the operation | movement of the connection maintenance confirmation phase in embodiment of the network topology detection system by this invention. 接続維持確認フェーズの具体的な動作を示す説明図である。It is explanatory drawing which shows the specific operation | movement of a connection maintenance confirmation phase. 逆方向の接続維持確認フェーズの動作を示すフローチャートである。It is a flowchart which shows the operation | movement of the connection maintenance confirmation phase of a reverse direction. ネットワークトポロジー検出システムにおけるＯＦＳの他の構成例を示す説明図である。It is explanatory drawing which shows the other structural example of OFS in a network topology detection system. 本発明によるネットワークトポロジー検出システムの主要部の構成を示すブロック図である。It is a block diagram which shows the structure of the principal part of the network topology detection system by this invention. 一般的なネットワークトポロジー検出システムを示す説明図である。It is explanatory drawing which shows a general network topology detection system.
　以下、本発明によるネットワークトポロジー検出システムの実施形態を、図面を参照して説明する。ネットワークトポロジーのポート間のリンクは双方向であるので、片方向のリンクが確認できれば、双方向共に接続されているとみなすことができる。本実施形態のネットワークトポロジー検出システムは、この性質を利用して、接続検出プロトコルパケットの送信回数を削減する。 Hereinafter, an embodiment of a network topology detection system according to the present invention will be described with reference to the drawings. Since the link between the ports in the network topology is bidirectional, if a one-way link can be confirmed, it can be considered that both directions are connected. The network topology detection system of the present embodiment uses this property to reduce the number of connection detection protocol packet transmissions.
　図１は、ネットワークトポロジー検出システムの実施形態における構成を示すブロック図である。図１に示すように、本実施形態のネットワークトポロジー検出システムは、ＯｐｅｎＦｌｏｗ制御装置１０と、複数のＯＦＳ２０とを含む。 FIG. 1 is a block diagram showing a configuration in an embodiment of a network topology detection system. As shown in FIG. 1, the network topology detection system of the present embodiment includes an OpenFlow control device 10 and a plurality of OFSs 20.
　ＯｐｅｎＦｌｏｗ制御装置１０は、制御対象のＯＦＳ２０のネットワークトポロジーを管理するトポロジー管理部１１と、記憶部１３とを備える。記憶部１３は、トポロジー管理テーブル１２を記憶する。 The OpenFlow control device 10 includes a topology management unit 11 that manages the network topology of the OFS 20 to be controlled, and a storage unit 13. The storage unit 13 stores the topology management table 12.
　トポロジー管理テーブル１２は、ネットワークのトポロジーデータを保持する。トポロジー管理テーブル１２は、ＯＦＳ２０のポートを示す情報を、接続が検出されていないポートを含む複数の接続未検出グループと、複数のオープンフロースイッチのポートのうち接続が検出されたポートを含む複数の接続検出グループとに分けて保持する。本実施形態では、トポロジー管理テーブル１２は、ＯＦＳ２０のポートを示す情報を、接続未検出Ａ、接続未検出Ｂ、接続検出Ａ、接続検出Ｂという４つのグループに分けて保持する。 The topology management table 12 holds network topology data. The topology management table 12 includes information indicating the ports of the OFS 20, including a plurality of connection undetected groups including a port where connection is not detected, and a plurality of ports including a port where connection is detected among ports of a plurality of OpenFlow switches. Separately hold the connection detection group. In the present embodiment, the topology management table 12 holds information indicating the ports of the OFS 20 in four groups of connection undetected A, connection undetected B, connection detected A, and connection detected B.
　トポロジー管理部１１は、接続未検出グループに属するＯＦＳ２０のポートに、ＯＦＳ２０間の接続状態を検出するためのパケットである接続検出プロトコルパケットをグループ毎に送信する。 The topology management unit 11 transmits a connection detection protocol packet, which is a packet for detecting a connection state between the OFSs 20, to the ports of the OFS 20 belonging to the connection undetected group for each group.
　トポロジー管理部１１は、例えば、特定の演算処理等を行うよう設計されたハードウェア、またはプログラムに従って動作するＣＰＵ（Ｃｅｎｔｒａｌ　Ｐｒｏｃｅｓｓｉｎｇ　Ｕｎｉｔ）等の情報処理装置によって実現される。また、そのプログラムは、非一時的でコンピュータが読み取り可能な記憶媒体に記憶される。また、記憶部１３は、ＨＤＤ（Ｈａｒｄ　ｄｉｓｋ　ｄｒｉｖｅ）等の記憶装置により実現される。 The topology management unit 11 is realized by, for example, hardware designed to perform specific arithmetic processing or the like, or an information processing device such as a CPU (Central Processing Unit) that operates according to a program. The program is stored in a non-transitory computer-readable storage medium. The storage unit 13 is realized by a storage device such as an HDD (Hard disk drive).
　ＯＦＳ２０は、接続検出プロトコル送信部２１と、接続検出プロトコル受信部２２とを含む。 The OFS 20 includes a connection detection protocol transmission unit 21 and a connection detection protocol reception unit 22.
　接続検出プロトコル送信部２１は、ＯｐｅｎＦｌｏｗ制御装置１０から自身のポートへ送られた接続検出プロトコルのパケットを他のＯＦＳ２０に送信する。 The connection detection protocol transmission unit 21 transmits the connection detection protocol packet sent from the OpenFlow control device 10 to its own port to the other OFS 20.
　接続検出プロトコル受信部２２は、他のＯＦＳ２０から送られた接続検出プロトコルのパケットを自身のポートから受信し、ＯｐｅｎＦｌｏｗ制御装置１０に送る。 The connection detection protocol receiving unit 22 receives a connection detection protocol packet sent from another OFS 20 from its own port, and sends it to the OpenFlow control device 10.
　次に、本実施形態のネットワークトポロジー検出システムの動作を、接続検出フェーズ、接続維持確認フェーズのそれぞれについて説明する。まず、接続検出フェーズの動作を説明する。図２は、ネットワークトポロジー検出システムの実施形態における接続検出フェーズの動作を示すフローチャートである。 Next, the operation of the network topology detection system of this embodiment will be described for each of the connection detection phase and the connection maintenance confirmation phase. First, the operation of the connection detection phase will be described. FIG. 2 is a flowchart showing the operation of the connection detection phase in the embodiment of the network topology detection system.
　トポロジー管理部１１は、最初に、ＯＦＳの認識順に、そのＯＦＳの全ポートを一括して、接続未検出Ａと接続未検出Ｂのグループに交互に振り分ける（ステップＡ１）。図３は、このグループ振り分け動作が行われた後のシステムの状態を示す説明図である。 First, the topology management unit 11 collectively distributes all ports of the OFS into the group of connection undetected A and connection undetected B in the order of recognition of the OFS (step A1). FIG. 3 is an explanatory diagram showing the state of the system after this group distribution operation has been performed.
　トポロジー管理部１１は、接続未検出Ａのポートへ接続検出プロトコルパケットを送信する（ステップＡ２）。そして、トポロジー管理部１１は、接続検出プロトコルパケットの受信を一定時間待つ（ステップＡ３）。トポロジー管理部１１は、接続検出プロトコルパケットを受信した場合は、送信元ポートを接続検出Ａに移動させ、受信先ポートを接続検出Ｂに移動させ、これら２つのポートは接続されていると判断する（ステップＡ４）。この際、トポロジー管理部１１は、双方向ともに受信した場合には、先に受信した方向に基づいてグループを決める。トポロジー管理部１１は、接続検出プロトコルパケットを受信しなかった場合は、何もしない。 The topology management unit 11 transmits a connection detection protocol packet to the port of connection not detected A (step A2). Then, the topology management unit 11 waits for a predetermined time to receive the connection detection protocol packet (step A3). When the topology management unit 11 receives the connection detection protocol packet, the topology management unit 11 moves the transmission source port to the connection detection A, moves the reception destination port to the connection detection B, and determines that these two ports are connected. (Step A4). At this time, when the topology management unit 11 receives both directions, the topology management unit 11 determines a group based on the previously received direction. The topology management unit 11 does nothing if the connection detection protocol packet is not received.
　図３、図４を参照して、ステップＡ２からＡ４までの具体的な動作を説明する。図４は、図３に示す状態から、ステップＡ２からＡ４までの接続検出後の振り分け動作が行われた後の状態を示した説明図である。図３において、接続未検出Ａは、ＯＦＳ１ポート１、ＯＦＳ１ポート２、ＯＦＳ１ポート３、ＯＦＳ３ポート１、ＯＦＳ３ポート２、ＯＦＳ３ポート３の６つのデータを含む。そのため、トポロジー管理部１１は、これら６つのポートへ接続検出プロトコルパケットを送信する（ステップＡ２）。ＯＦＳ１ポート１は、外部接続であるので、トポロジー管理部１１は、ＯＦＳ１ポート１へ送信した接続検出プロトコルパケットを受信しない（ステップＡ３の「受信しなかった」）。 Referring to FIGS. 3 and 4, the specific operation from step A2 to A4 will be described. FIG. 4 is an explanatory view showing a state after the distribution operation after the connection detection from step A2 to A4 is performed from the state shown in FIG. In FIG. 3, connection not detected A includes six data of OFS 1 port 1, OFS 1 port 2, OFS 1 port 3, OFS 3 port 1, OFS 3 port 2, and OFS 3 port 3. Therefore, topology management unit 11 transmits a connection detection protocol packet to these six ports (step A2). Since OFS1 port 1 is an external connection, topology management unit 11 does not receive the connection detection protocol packet transmitted to OFS1 port 1 (“not received” in step A3).
　したがって、ＯＦＳ１ポート１は、そのまま接続未検出Ａに残る。ＯＦＳ１ポート２はＯＦＳ２ポート１と接続されているので、トポロジー管理部１１は、ＯＦＳ１ポート２へ送信した接続検出プロトコルパケットをＯＦＳ２ポート１から受信する（ステップＡ３の「受信した」）。この場合、トポロジー管理部１１は、ＯＦＳ１ポート２を接続検出Ａに移動させ、ＯＦＳ２ポート１を接続検出Ｂに移動させる（ステップＡ４）。 Therefore, OFS1 port 1 remains in connection undetected A as it is. Since the OFS1 port 2 is connected to the OFS2 port 1, the topology management unit 11 receives the connection detection protocol packet transmitted to the OFS1 port 2 from the OFS2 port 1 (“received” in step A3). In this case, topology management unit 11 moves OFS1 port 2 to connection detection A and moves OFS2 port 1 to connection detection B (step A4).
　ＯＦＳ１ポート３は、ＯＦＳ３ポート３と接続されているので、トポロジー管理部１１は、ＯＦＳ１ポート３へ送信した接続検出プロトコルパケットを、ＯＦＳ３ポート３から受信する。さらに、トポロジー管理部１１は、ＯＦＳ３ポート３へ送信した接続検出プロトコルパケットを、ＯＦＳ１ポート３から受信する。トポロジー管理部１１は、ＯＦＳ３ポート３からの受信が早い場合、ＯＦＳ１ポート３を接続検出Ａに移動させ、ＯＦＳ３ポート３は接続検出Ｂに移動させ、これら２つのポートは接続されていると判断する（ステップＡ４）。また、残りの接続検出プロトコルパケットの受信についても同様に処理が行われ、各グループに属するＯＦＳのポートは図４に示す状態になる。 Since the OFS 1 port 3 is connected to the OFS 3 port 3, the topology management unit 11 receives the connection detection protocol packet transmitted to the OFS 1 port 3 from the OFS 3 port 3. Further, the topology management unit 11 receives the connection detection protocol packet transmitted to the OFS 3 port 3 from the OFS 1 port 3. When the reception from OFS3 port 3 is early, topology management unit 11 moves OFS1 port 3 to connection detection A, moves OFS3 port 3 to connection detection B, and determines that these two ports are connected. (Step A4). The same processing is performed for the reception of the remaining connection detection protocol packets, and the ports of the OFS belonging to each group are in the state shown in FIG.
　次に、トポロジー管理部１１は、接続未検出Ｂのポートへ接続検出プロトコルパケットを送信する（ステップＡ５）。そして、トポロジー管理部１１は、接続検出プロトコルパケットの受信を一定時間待つ（ステップＡ６）。トポロジー管理部１１は、接続検出プロトコルパケットを受信した場合は、送信元ポートを接続検出Ｂに移動させ、受信先ポートを接続検出Ａに移動させ、それら２つのポートは接続されていると判断する（ステップＡ７）。トポロジー管理部１１は、２つのポート間の接続検出プロトコルパケットを、双方向ともに受信した場合には、先に受信した方向に基づいてグループを決める。トポロジー管理部１１は、接続検出プロトコルパケットを受信しなかった場合は、何もしない。そして、その後はステップＡ２以降の処理が繰り返される。 Next, the topology management unit 11 transmits a connection detection protocol packet to the port of connection undetected B (step A5). Then, topology management unit 11 waits for a certain period of time to receive the connection detection protocol packet (step A6). When the topology management unit 11 receives the connection detection protocol packet, the topology management unit 11 moves the transmission source port to the connection detection B, moves the reception destination port to the connection detection A, and determines that these two ports are connected. (Step A7). When the topology management unit 11 receives a connection detection protocol packet between two ports in both directions, the topology management unit 11 determines a group based on the previously received direction. The topology management unit 11 does nothing if the connection detection protocol packet is not received. Thereafter, the processing after step A2 is repeated.
　図４および図５を参照して、ステップＡ５からＡ７までの具体的な動作を説明する。図５は、図４に示す状態から、ステップＡ５からＡ７までの動作を行った後のシステムの状態を示す説明図である。図４において、接続未検出Ｂは、ＯＦＳ２ポート２、ＯＦＳ２ポート３、ＯＦＳ４ポート２、ＯＦＳ４ポート３の４つを含む。よって、トポロジー管理部１１は、これら４つのポートへ接続検出プロトコルパケットを送信する（ステップＡ５）。ＯＦＳ２ポート２は外部接続であるので、トポロジー管理部１１は、ＯＦＳ２ポート２へ送信した接続検出プロトコルパケットを受信しない（ステップＡ６の「受信しなかった」）。したがって、ＯＦＳ２ポート２は、そのまま接続未検出Ｂに残る。ＯＦＳ２ポート３はＯＦＳ４ポート３と接続されているので、トポロジー管理部１１は、ＯＦＳ２ポート３へ送信した接続検出プロトコルパケットを、ＯＦＳ４ポート３から受信する（ステップＡ６の「受信した」）。 With reference to FIG. 4 and FIG. 5, the specific operation | movement from step A5 to A7 is demonstrated. FIG. 5 is an explanatory diagram showing the state of the system after performing the operations from steps A5 to A7 from the state shown in FIG. In FIG. 4, the connection undetected B includes four of the OFS 2 port 2, the OFS 2 port 3, the OFS 4 port 2, and the OFS 4 port 3. Therefore, topology management unit 11 transmits a connection detection protocol packet to these four ports (step A5). Since OFS2 port 2 is an external connection, topology management unit 11 does not receive the connection detection protocol packet transmitted to OFS2 port 2 (“not received” in step A6). Therefore, OFS2 port 2 remains in connection undetected B as it is. Since the OFS 2 port 3 is connected to the OFS 4 port 3, the topology management unit 11 receives the connection detection protocol packet transmitted to the OFS 2 port 3 from the OFS 4 port 3 (“received” in step A 6).
　さらに、トポロジー管理部１１は、ＯＦＳ４ポート３へ送信した接続検出プロトコルパケットを、ＯＦＳ２ポート３から受信する（ステップＡ６の「受信した」）。ＯＦＳ２ポート３からの受信が早かった場合、トポロジー管理部１１は、ＯＦＳ４ポート３を接続検出Ｂに振り分け、ＯＦＳ２ポート３を接続検出Ａに振り分け、これら２つのポートが接続されていると判断する（ステップＡ７）。残りの接続検出プロトコルパケットの受信についても同様に処理が行われ、トポロジー管理テーブル１２は図５に示す状態になる。 Further, the topology management unit 11 receives the connection detection protocol packet transmitted to the OFS 4 port 3 from the OFS 2 port 3 (“received” in step A 6). If the reception from the OFS2 port 3 is early, the topology management unit 11 assigns the OFS4 port 3 to the connection detection B, assigns the OFS2 port 3 to the connection detection A, and determines that these two ports are connected ( Step A7). The same processing is performed for reception of the remaining connection detection protocol packets, and the topology management table 12 is in the state shown in FIG.
　以上のように、本実施形態のネットワークトポロジー検出システムを用いた場合、図３～図５に示した構成例では、接続検出プロトコルのパケット送信回数は、ステップＡ２で６回、ステップＡ５で４回であるため、６＋４＝１０となる。一方、図１１に示した例のように、接続検出プロトコルパケットを、全ポートへ送信した場合、接続検出フェーズのパケット送信回数は全ポート数の１２である。よって、ネットワークトポロジー検出システムは、ポートをグループ分けせずに全ポートへ送信するシステムと比べて、接続検出プロトコルパケットの送信回数を削減することができる。 As described above, when the network topology detection system of the present embodiment is used, in the configuration examples shown in FIGS. 3 to 5, the number of packet transmissions of the connection detection protocol is 6 times in step A2 and 4 times in step A5. Therefore, 6 + 4 = 10. On the other hand, when the connection detection protocol packet is transmitted to all ports as in the example shown in FIG. 11, the number of packet transmissions in the connection detection phase is 12, which is the total number of ports. Therefore, the network topology detection system can reduce the number of times of connection detection protocol packet transmission compared to a system that transmits ports to all ports without grouping the ports.
　次に、ネットワークトポロジー検出システムの接続維持確認フェーズの動作を説明する。図６は、ネットワークトポロジー検出システムの接続維持確認フェーズの動作を示すフローチャートである。 Next, the operation of the connection maintenance confirmation phase of the network topology detection system will be described. FIG. 6 is a flowchart showing the operation in the connection maintenance confirmation phase of the network topology detection system.
　トポロジー管理部１１は、接続検出Ａのポートへ接続検出プロトコルパケットを送信する（ステップＢ１）。そして、トポロジー管理部１１は、接続検出プロトコルパケットの受信を一定時間待つ（ステップＢ２）。トポロジー管理部１１は、接続検出プロトコルパケットを受信した場合は、受信先ポートが変化したかどうかを確認する（ステップＢ３）。変化しなければ１回の接続維持確認が終了する。トポロジー管理部１１は、接続検出プロトコルパケットをどこからも受信しなかった場合は、それぞれのポートを、接続検出Ａから接続未検出Ａに移動させ、接続検出Ｂから接続未検出Ｂに移動させる（ステップＢ４）。トポロジー管理部１１は、接続検出プロトコルパケットの受信先ポートが変化した場合、旧受信先ポートを接続未検出Ｂに移動させ、新受信先ポートを接続検出Ｂに移動させる。トポロジー管理部１１は、新受信先ポートに対応する送信元ポートが接続検出Ａにある場合には、その送信元ポートを接続未検出Ａに移動させる（ステップＢ５）。 The topology management unit 11 transmits a connection detection protocol packet to the connection detection A port (step B1). Then, topology management unit 11 waits for a certain period of time to receive the connection detection protocol packet (step B2). When the topology management unit 11 receives the connection detection protocol packet, the topology management unit 11 checks whether the destination port has changed (step B3). If there is no change, one connection maintenance confirmation is completed. When the topology management unit 11 does not receive the connection detection protocol packet from anywhere, the topology management unit 11 moves each port from the connection detection A to the connection non-detection A, and moves from the connection detection B to the connection non-detection B (step). B4). When the destination port of the connection detection protocol packet changes, the topology management unit 11 moves the old destination port to the connection undetected B and moves the new destination port to the connection detection B. When the source port corresponding to the new destination port is in connection detection A, topology management unit 11 moves the source port to connection undetected A (step B5).
　次に接続維持確認フェーズの動作を、図を用いて具体的に説明する。図７は、接続維持確認フェーズの具体的な動作を示す説明図である。図７において、接続検出Ａは、ＯＦＳ１ポート２、ＯＦＳ１ポート３、ＯＦＳ３ポート２、ＯＦＳ２ポート３の４つのポートを含むので、これら４つのポートへ接続検出プロトコルパケットを送信する（ステップＢ１）。ＯＦＳ１ポート２とＯＦＳ２ポート１の接続が維持されていれば、トポロジー管理部１１は、ＯＦＳ１ポート２へ送信した接続検出プロトコルパケットを、ＯＦＳ２ポート１から受信する（ステップＢ２の「受信した」）。受信先ポートに変化はないので（ステップＢ３のＮＯ）、接続維持確認は終了である。トポロジー管理部１１は、残り３つのポートに関しても同様に処理を行う。 Next, the operation of the connection maintenance confirmation phase will be specifically described with reference to the drawings. FIG. 7 is an explanatory diagram showing a specific operation in the connection maintenance confirmation phase. In FIG. 7, since connection detection A includes four ports, OFS1 port 2, OFS1 port 3, OFS3 port 2, and OFS2 port 3, connection detection protocol packets are transmitted to these four ports (step B1). If the connection between OFS1 port 2 and OFS2 port 1 is maintained, topology management unit 11 receives the connection detection protocol packet transmitted to OFS1 port 2 from OFS2 port 1 (“received” in step B2). Since there is no change in the destination port (NO in step B3), the connection maintenance confirmation is completed. The topology management unit 11 performs the same process for the remaining three ports.
　以上のように、トポロジー管理部１１は、ＯＦＳ１ポート２へ送信した接続検出プロトコルパケットを、ＯＦＳ２ポート１から受信することにより、ＯＦＳ１ポート２とＯＦＳ２ポート１の接続が維持されていることを確認する。本実施形態のネットワークトポロジー検出システムは、接続検出フェーズにおいて、接続検出したポート同士を異なるグループに振り分けているので、一方のグループのみの確認だけで全ての接続維持確認をしたとみなすことができる。 As described above, the topology management unit 11 receives the connection detection protocol packet transmitted to the OFS1 port 2 from the OFS2 port 1, thereby confirming that the connection between the OFS1 port 2 and the OFS2 port 1 is maintained. . In the network topology detection system according to the present embodiment, since the detected ports are distributed to different groups in the connection detection phase, it can be considered that all connection maintenance confirmations have been made by checking only one group.
　なお、ネットワークの回線に例えば光ケーブルを用いた場合、片方向の通信のみ障害が発生する可能性がある。本実施形態のネットワークトポロジー検出システムは、そのような可能性を考慮して、一方向からの接続の確認を行った所定期間後に、逆方向からの接続の確認を行う。すなわち、ネットワークトポロジー検出システムは、定期的に接続維持確認フェーズの動作を行う場合、図６に示す動作と後述する図８に示す逆方向の動作を交互に行う。以下、接続維持確認フェーズにおける逆方向からの接続の確認を示す動作を説明する。 Note that, for example, when an optical cable is used for a network line, there is a possibility that a failure occurs only in one-way communication. In consideration of such a possibility, the network topology detection system according to the present embodiment confirms connection from the opposite direction after a predetermined period of time when confirmation of connection from one direction is performed. That is, the network topology detection system alternately performs the operation shown in FIG. 6 and the reverse operation shown in FIG. Hereinafter, an operation indicating confirmation of connection from the reverse direction in the connection maintenance confirmation phase will be described.
　図８は、逆方向の接続維持確認フェーズの動作を示すフローチャートである。トポロジー管理部１１は、接続検出Ｂのポートへ接続検出プロトコルパケットを送信する（ステップＢ６）。そして、トポロジー管理部１１は、接続検出プロトコルパケットの受信を一定時間待つ（ステップＢ７）。トポロジー管理部１１は、接続検出プロトコルパケットを受信した場合は、受信先ポートが変化したかどうかを確認する（ステップＢ８）。変化しなければ１回の接続維持確認が終了する。トポロジー管理部１１は、接続検出プロトコルパケットをどこからも受信しなかった場合には、それぞれのポートを接続検出Ｂから接続未検出Ｂに移動させ、接続検出Ａから接続未検出Ａに移動させる（ステップＢ９）。接続検出プロトコルパケットの受信先ポートが変化した場合には、トポロジー管理部１１は、旧受信先ポートを接続未検出Ａに移動させ、新受信先ポートを接続検出Ａに移動させる。新受信先ポートに対応する送信元ポートが接続検出Ｂにある場合には、トポロジー管理部１１は、そのポートを接続未検出Ｂに移動させる（ステップＢ１０）。 FIG. 8 is a flowchart showing the operation of the connection maintenance confirmation phase in the reverse direction. The topology management unit 11 transmits a connection detection protocol packet to the connection detection B port (step B6). Then, topology management unit 11 waits for a predetermined time to receive the connection detection protocol packet (step B7). When the topology management unit 11 receives the connection detection protocol packet, the topology management unit 11 checks whether the destination port has changed (step B8). If there is no change, one connection maintenance confirmation is completed. When the topology management unit 11 does not receive the connection detection protocol packet from anywhere, the topology management unit 11 moves each port from the connection detection B to the connection non-detection B, and moves from the connection detection A to the connection non-detection A (step). B9). When the destination port of the connection detection protocol packet changes, the topology management unit 11 moves the old destination port to the connection not detected A and moves the new destination port to the connection detection A. When the source port corresponding to the new destination port is in connection detection B, topology management unit 11 moves the port to connection undetected B (step B10).
　次に逆方向からの接続維持確認フェーズの動作を、図７を用いて具体的に説明する。接続検出Ｂは、ＯＦＳ２ポート１、ＯＦＳ３ポート３、ＯＦＳ４ポート１、ＯＦＳ４ポート３の４つのポートを含むので、トポロジー管理部１１は、これら４つのポートへ接続検出プロトコルパケットを送信する（ステップＢ６）。ＯＦＳ２ポート１とＯＦＳ１ポート２の接続が維持されていれば、トポロジー管理部１１は、ＯＦＳ２ポート１へ送信した接続検出プロトコルパケットを、ＯＦＳ１ポート２から受信する（ステップＢ７の「受信した」）。受信先ポートに変化はないので（ステップＢ８のＮＯ）、接続維持確認は終了である。トポロジー管理部１１は、残り３つのポートに関しても同様に処理を行う。 Next, the operation of the connection maintenance confirmation phase from the reverse direction will be specifically described with reference to FIG. Since connection detection B includes four ports, OFS2 port 1, OFS3 port 3, OFS4 port 1, and OFS4 port 3, topology management unit 11 transmits a connection detection protocol packet to these four ports (step B6). . If the connection between OFS2 port 1 and OFS1 port 2 is maintained, topology management unit 11 receives the connection detection protocol packet transmitted to OFS2 port 1 from OFS1 port 2 (“received” in step B7). Since there is no change in the destination port (NO in step B8), the connection maintenance confirmation is completed. The topology management unit 11 performs the same process for the remaining three ports.
　以上のように、本実施形態のネットワークトポロジー検出システムを用いた場合、図７に示した構成例において、接続維持確認フェーズにおけるパケット送信回数は４である。一方、接続検出プロトコルパケットを、接続検出できた全ポートへ送信した場合、接続維持確認フェーズにおけるパケット送信回数は８である。よって、ネットワークトポロジー検出システムは、ポートをグループ分けせずに全ポートへ送信する方法と比べて、接続維持確認のための接続検出プロトコルパケットの送信回数を削減することができる。特に、接続検出フェーズにおいて、接続検出したポート同士を異なるグループに振り分けていたことによって、接続維持確認フェーズにおいてより確実にプロトコルパケットの送信回数を削減することができる。 As described above, when the network topology detection system of this embodiment is used, the number of packet transmissions in the connection maintenance confirmation phase is 4 in the configuration example shown in FIG. On the other hand, when the connection detection protocol packet is transmitted to all the ports for which connection has been detected, the number of packet transmissions in the connection maintenance confirmation phase is 8. Therefore, the network topology detection system can reduce the number of transmissions of connection detection protocol packets for connection maintenance confirmation, compared to a method of transmitting ports to all ports without grouping the ports. In particular, in the connection detection phase, since the detected ports are allocated to different groups, the number of protocol packet transmissions can be more reliably reduced in the connection maintenance confirmation phase.
　図９は、本実施形態のネットワークトポロジー検出システムの他の構成例を示す説明図である。図９に示す例では、図３等に示した構成と比べて、ＯＦＳ１とＯＦＳ４との間の接続、ＯＦＳ２とＯＦＳ３との間の接続が増えている。この場合、接続検出プロトコルパケットを全ポートへ送信した場合の送信回数は１６であるが、本実施形態のＯｐｅｎＦｌｏｗ制御装置の制御を用いると８＋４＝１２となる。このように、ＯｐｅｎＦｌｏｗ制御装置１０の制御対象であるＯＦＳ間の接続数が多い場合、接続検出フェーズの効果がより大きくなる。 FIG. 9 is an explanatory diagram showing another configuration example of the network topology detection system of the present embodiment. In the example shown in FIG. 9, the connection between OFS1 and OFS4 and the connection between OFS2 and OFS3 are increased compared to the configuration shown in FIG. In this case, the number of transmissions when the connection detection protocol packet is transmitted to all ports is 16, but 8 + 4 = 12 when the control of the OpenFlow control device of this embodiment is used. As described above, when the number of connections between OFSs to be controlled by the OpenFlow control device 10 is large, the effect of the connection detection phase is further increased.
　本実施形態のネットワークトポロジー検出システムは、ネットワークトポロジー検出システムポートを複数のグループで管理し、グループ毎に接続検出を行うことにより、接続検出プロトコルパケットの送信回数を削減することができる。その理由は、本実施形態のネットワークトポロジー検出システムは、あるポート間において片方向の接続を検出した場合は双方向の接続が可能であるとみなして逆方向の検出を行わないからである。 The network topology detection system of the present embodiment can reduce the number of connection detection protocol packet transmissions by managing network topology detection system ports in a plurality of groups and performing connection detection for each group. The reason is that the network topology detection system of the present embodiment does not detect in the reverse direction, assuming that a bidirectional connection is possible if a one-way connection is detected between certain ports.
　なお、検出前における接続未検出グループの振り分け方は、ステップＡ１および図３に示した振り分け方に限られない。記憶部１３は、可能な限り、接続検出を行う対象のポート同士を異なるグループに記憶し、接続検出を行わないポート同士を同じグループに記憶することが好ましい。接続検出を行う対象の２つのポートが同じグループに属している場合、トポロジー管理部１１は、その２つのポートに接続検出プロトコルパケットを一括して送信してしまうため、そのポート間に関しては検出回数を削減することができないからである。例えば、同じＯＦＳに含まれるポートは接続されることはないため、記憶部１３は、同じＯＦＳに含まれるポートを同じ接続未検出グループに記憶する。 Note that the way of allocating unconnected groups before detection is not limited to the way shown in step A1 and FIG. It is preferable that the storage unit 13 store the ports for which connection detection is performed in different groups as much as possible, and store the ports for which connection detection is not performed in the same group. When the two ports subject to connection detection belong to the same group, the topology management unit 11 transmits connection detection protocol packets to the two ports in a lump, and therefore the number of detections between the ports. This is because it cannot be reduced. For example, since the ports included in the same OFS are not connected, the storage unit 13 stores the ports included in the same OFS in the same connection undetected group.
　また、本実施形態では、グループは４つに分けられているが、グループの数は必ずしも４つでなくてもよい。例えば、図３に示す例において、トポロジー管理部１１は、接続未検出Ａと接続未検出Ｂをさらに２つのグループに分けることにより、ＯＦＳ毎に異なるグループに分けてもよい。これにより検出回数をさらに削減することができる。ただし、グループの数が増えるほど、検出処理にかかる時間が増加するため、その時間を考慮してグループ数を決定することが好ましい。 In this embodiment, the group is divided into four groups, but the number of groups is not necessarily four. For example, in the example illustrated in FIG. 3, the topology management unit 11 may further divide the connection undetected A and the connection undetected B into two groups so as to be divided into different groups for each OFS. As a result, the number of detections can be further reduced. However, since the time required for the detection process increases as the number of groups increases, it is preferable to determine the number of groups in consideration of the time.
　なお、本実施形態のネットワークトポロジー検出システムで利用する接続検出プロトコルは、装置の開発元により提供される独自のプロトコルとして用いられることが想定される。しかし、ＯｐｅｎＦｌｏｗを用いたシステムにおいては、一般的に同じＯｐｅｎＦｌｏｗの規格に則った装置のみを接続したネットワークが構成されるので、その独自プロトコルで、ネットワーク全体のトポロジーを検出することができる。 Note that the connection detection protocol used in the network topology detection system of this embodiment is assumed to be used as a unique protocol provided by the device developer. However, in a system using OpenFlow, a network in which only devices conforming to the same OpenFlow standard is generally configured, so that the topology of the entire network can be detected using the unique protocol.
　また、本実施形態のネットワークトポロジー検出システムは、オープンフロー技術を用いたネットワーク以外のネットワークにも用いることができる。その場合、Ｏｐｅｎｆｌｏｗ装置１０は、スイッチを制御する装置であればよく、一般的なサーバ等を用いて実現される。また、ＯＦＳ２０は、通信経路を切り替えることができるスイッチであればよい。 In addition, the network topology detection system of the present embodiment can be used for networks other than networks using the open flow technology. In this case, the Openflow device 10 may be a device that controls a switch, and is realized using a general server or the like. The OFS 20 may be a switch that can switch the communication path.
　図１０は、本発明によるネットワークトポロジー検出システムの主要部の構成を示すブロック図である。図１０に示すように、本発明によるネットワークトポロジー検出システムは、主要な構成要素として、相互にパケットの送受信を行う複数のスイッチ４０と、複数のスイッチ４０を制御する制御装置３０とを備える。制御装置３０は、複数のスイッチ４０のポートのうち接続が検出されていないポートを示す情報を保持する複数の接続未検出グループ、および複数のスイッチ４０のポートのうち接続が検出されたポートを示す情報を保持する複数の接続検出グループを記憶する記憶手段３３と、接続未検出グループに属するポートに、当該複数のスイッチ４０間の接続状態を検出するためのパケットである接続検出プロトコルパケットをグループ毎に送信するトポロジー管理手段３１とを含む。トポロジー管理手段３１は、第１の接続未検出グループに属する第１のポートに接続検出プロトコルパケットを送信した後に、第２の接続未検出グループに属する第２のポートから当該接続検出プロトコルパケットを受信した場合に、当該第１のポートを示す情報および当該第２のポートを示す情報を接続検出グループに移動させる。 FIG. 10 is a block diagram showing the configuration of the main part of the network topology detection system according to the present invention. As shown in FIG. 10, the network topology detection system according to the present invention includes, as main components, a plurality of switches 40 that mutually transmit and receive packets, and a control device 30 that controls the plurality of switches 40. The control device 30 indicates a plurality of connection undetected groups that hold information indicating ports for which connections are not detected among the ports of the plurality of switches 40, and ports for which connections are detected among the ports of the plurality of switches 40. Storage means 33 for storing a plurality of connection detection groups holding information, and connection detection protocol packets, which are packets for detecting the connection state between the plurality of switches 40, for each port belonging to the connection undetected group. Topology management means 31 for transmitting to the network. The topology management unit 31 receives the connection detection protocol packet from the second port belonging to the second connection undetected group after transmitting the connection detection protocol packet to the first port belonging to the first connection undetected group. In this case, the information indicating the first port and the information indicating the second port are moved to the connection detection group.
　また、上記の各実施形態には、下記の（１）～（４）に示すネットワークトポロジー検出システムも開示されている。 In each of the above embodiments, network topology detection systems shown in the following (1) to (4) are also disclosed.
（１）相互にパケットの送受信を行う複数のスイッチ（例えば、ＯＦＳ２０）と、複数のスイッチを制御する制御装置（例えば、ＯｐｅｎＦｌｏｗ制御装置１０）とを備え、制御装置は、複数のスイッチのポートのうち接続が検出されていないポートを示す情報を保持する複数の接続未検出グループ（例えば、接続未検出Ａのグループ、接続未検出Ｂのグループ）、および複数のスイッチのポートのうち接続が検出されたポートを示す情報を保持する複数の接続検出グループ（例えば、接続検出Ａのグループ、接続検出Ｂのグループ）を記憶する記憶手段（例えば、記憶部１３）と、接続未検出グループに属するポートに、当該複数のスイッチ間の接続状態を検出するためのパケットである接続検出プロトコルパケットをグループ毎に送信するトポロジー管理手段（例えば、トポロジー管理部１１）とを含み、トポロジー管理手段は、第１の接続未検出グループ（例えば、接続未検出Ａのグループ）に属する第１のポート（例えば、ＯＦＳ１ポート２）に接続検出プロトコルパケットを送信した後に、第２の接続未検出グループ（例えば、接続未検出Ｂのグループ）に属する第２のポート（例えば、ＯＦＳ２ポート１）から当該接続検出プロトコルパケットを受信した場合に、当該第１のポートを示す情報および当該第２のポートを示す情報を接続検出グループに移動させるネットワークトポロジー検出システム。 (1) A plurality of switches (for example, OFS 20) that transmit and receive packets to and from each other and a control device (for example, OpenFlow control device 10) that controls the plurality of switches. A connection is detected among a plurality of connection undetected groups (for example, a group of connection undetected A, a group of connection undetected B) that holds information indicating ports for which connection is not detected, and a plurality of switch ports. Storage means (for example, the storage unit 13) for storing a plurality of connection detection groups (for example, a group of connection detection A and a group of connection detection B) that hold information indicating the connected ports, and ports belonging to the connection undetected group , A connection detection protocol packet that is a packet for detecting a connection state between the plurality of switches is transmitted for each group. Topology management means (for example, topology management unit 11), and the topology management means is a first port (for example, OFS1 port 2) belonging to the first connection undetected group (for example, group of connection undetected A). When the connection detection protocol packet is received from the second port (for example, OFS2 port 1) belonging to the second connection undetected group (for example, the group of connection undetected B) after transmitting the connection detection protocol packet to A network topology detection system that moves information indicating the first port and information indicating the second port to a connection detection group.
（２）ネットワークトポロジー検出システムは、トポロジー管理手段が、第１の接続未検出グループ（例えば、接続未検出Ａ）に属する第１のポート（例えば、ＯＦＳ１ポート２）に接続検出プロトコルパケットを送信し、第２の接続未検出グループ（例えば、接続未検出Ｂ）に属する第２のポート（例えば、ＯＦＳ２ポート１）から当該接続検出プロトコルパケットを受信した場合に、前記第１のポートを示す情報を第１の接続検出グループ（例えば、接続検出Ａ）に移動させ、前記第２のポートを示す情報を第２の接続検出グループ（例えば、接続検出Ｂ）に移動させ、再度、前記第１のポートに接続検出プロトコルパケットを送信し、前記第２のポートから当該接続検出プロトコルパケットを受信することにより前記第１のポートと前記第２のポートとの間の接続が維持されていることを確認するように構成されていてもよい。このようなネットワークトポロジー検出システムによれば、接続検出フェーズにおいて、接続検出したポート同士を異なるグループに振り分けているので、一方のグループのみの確認だけで全ての接続維持確認をしたとみなすことができる。 (2) In the network topology detection system, the topology management means transmits a connection detection protocol packet to the first port (for example, OFS1 port 2) belonging to the first connection undetected group (for example, connection not detected A). When the connection detection protocol packet is received from a second port (for example, OFS2 port 1) belonging to the second connection undetected group (for example, connection not detected B), information indicating the first port is displayed. Move to a first connection detection group (for example, connection detection A), move information indicating the second port to a second connection detection group (for example, connection detection B), and then again to the first port By transmitting a connection detection protocol packet to the first port and receiving the connection detection protocol packet from the second port. It may be configured to verify that the connection between the second port is maintained. According to such a network topology detection system, in the connection detection phase, the detected ports are distributed to different groups, so it can be considered that all connection maintenance confirmations have been made only by checking one group. .
（３）ネットワークトポロジー検出システムは、トポロジー管理手段が、第１のポート（例えば、ＯＦＳ１ポート２）に接続検出プロトコルパケットを送信し、第２のポート（例えば、ＯＦＳ２ポート１）から当該接続検出プロトコルパケットを受信することにより前記第１のポートと前記第２のポートとの間の接続維持を確認した所定期間後に、再度、前記第１のポートと前記第２のポートとの間の接続維持を確認する場合、前記第２のポートに接続検出プロトコルパケットを送信し、前記第１のポートから当該接続検出プロトコルパケットを受信することにより前記第１のポートと前記第２のポートとの間の接続維持を確認するように構成されていてもよい。このようなネットワークトポロジー検出システムによれば、ネットワークの回線に例えば光ケーブルを用いた場合のように、片方向の通信のみ障害が発生する可能性があるシステムであっても、パケット検出回数を減らしつつ接続維持確認の精度の低下を抑制することができる。 (3) In the network topology detection system, the topology management means transmits a connection detection protocol packet to the first port (for example, OFS1 port 2), and the connection detection protocol from the second port (for example, OFS2 port 1). After a predetermined period in which the connection maintenance between the first port and the second port is confirmed by receiving a packet, the connection maintenance between the first port and the second port is performed again. When confirming, the connection between the first port and the second port is transmitted by transmitting a connection detection protocol packet to the second port and receiving the connection detection protocol packet from the first port. It may be configured to confirm maintenance. According to such a network topology detection system, the number of packet detections can be reduced even in a system in which a failure may occur only in one-way communication, such as when an optical cable is used for a network line. A decrease in accuracy of connection maintenance confirmation can be suppressed.
（４）ネットワークトポロジー検出システムは、同一のスイッチ（例えば、ＯＦＳ１、ＯＦＳ２、ＯＦＳ３またはＯＦＳ４）のポートが、同一の接続未検出グループ（例えば、接続未検出Ａまたは接続未検出Ｂ）に記憶されるように構成されていてもよい。このようなネットワークトポロジー検出システムによれば、接続検出プロトコルパケットの送信回数の削減をより効果的にすることができる。 (4) In the network topology detection system, ports of the same switch (for example, OFS1, OFS2, OFS3, or OFS4) are stored in the same connection undetected group (for example, connection undetected A or connection undetected B). It may be configured as follows. According to such a network topology detection system, the number of connection detection protocol packet transmissions can be reduced more effectively.
　以上、実施形態を参照して本願発明を説明したが、本願発明は上記実施形態に限定されるものではない。本願発明の構成や詳細には、本願発明のスコープ内で当業者が理解し得る様々な変更をすることができる。 The present invention has been described above with reference to the embodiments, but the present invention is not limited to the above embodiments. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.
　この出願は、２０１４年２月３日に出願された日本出願特願２０１４－０１８１８５を基礎とする優先権を主張し、その開示の全てをここに取り込む。 This application claims priority based on Japanese Patent Application No. 2014-018185 filed on February 3, 2014, the entire disclosure of which is incorporated herein.
産業上の利用の可能性Industrial applicability
　本発明は、例えばＯｐｅｎＦｌｏｗを用いたネットワークに適用することが可能であり、ＯｐｅｎＦｌｏｗに限らず一般的なネットワークのトポロジーの検出に適用することが可能である。 The present invention can be applied to a network using OpenFlow, for example, and is not limited to OpenFlow, and can be applied to detection of a general network topology.
１０　ＯｐｅｎＦｌｏｗ制御装置
１１　トポロジー管理部
１２　トポロジー管理テーブル
１３　記憶部
２０　ＯＦＳ
２１　接続検出プロトコル送信部
２２　接続検出プロトコル受信部
３０　制御装置
３１　トポロジー管理手段
３３　記憶手段
４０　スイッチ 10 OpenFlow control device 11 Topology management unit 12 Topology management table 13 Storage unit 20 OFS
21 Connection Detection Protocol Transmitting Unit 22 Connection Detection Protocol Receiving Unit 30 Control Device 31 Topology Management Unit 33 Storage Unit 40 Switch

Claims (10)

  1. 　相互にパケットの送受信を行う複数のスイッチと、
    　前記複数のスイッチを制御する制御装置とを備え、
    　前記制御装置は、
    　前記複数のスイッチのポートのうち接続が検出されていないポートを示す情報を保持する複数の接続未検出グループ、および前記複数のスイッチのポートのうち接続が検出されたポートを示す情報を保持する複数の接続検出グループを記憶する記憶手段と、
    　前記接続未検出グループに属するポートに、当該複数のスイッチ間の接続状態を検出するためのパケットである接続検出プロトコルパケットをグループ毎に送信するトポロジー管理手段とを含み、
    　前記トポロジー管理手段は、
    　第１の接続未検出グループに属する第１のポートに前記接続検出プロトコルパケットを送信した後に、第２の接続未検出グループに属する第２のポートから当該接続検出プロトコルパケットを受信した場合に、当該第１のポートを示す情報および当該第２のポートを示す情報を前記接続検出グループに移動させる
    　ことを特徴とするネットワークトポロジー検出システム。     Multiple switches that send and receive packets to and from each other,
    A control device for controlling the plurality of switches,
    The control device includes:
    A plurality of connection undetected groups that hold information indicating ports in which connection is not detected among the ports of the plurality of switches, and a plurality that hold information indicating ports in which connections are detected among the ports of the plurality of switches. Storage means for storing the connection detection group of
    Topology management means for transmitting a connection detection protocol packet, which is a packet for detecting a connection state between the plurality of switches, to a port belonging to the connection undetected group for each group;
    The topology management means includes
    When the connection detection protocol packet is received from the second port belonging to the second connection undetected group after transmitting the connection detection protocol packet to the first port belonging to the first connection undetected group, A network topology detection system, wherein information indicating a first port and information indicating a second port are moved to the connection detection group.
  2. 　トポロジー管理手段は、
    　第１の接続未検出グループに属する第１のポートに接続検出プロトコルパケットを送信し、第２の接続未検出グループに属する第２のポートから当該接続検出プロトコルパケットを受信した場合に、前記第１のポートを示す情報を第１の接続検出グループに移動させ、前記第２のポートを示す情報を第２の接続検出グループに移動させ、
    　再度、前記第１のポートに接続検出プロトコルパケットを送信し、前記第２のポートから当該接続検出プロトコルパケットを受信することにより前記第１のポートと前記第２のポートとの間の接続が維持されていることを確認する
    　請求項１記載のネットワークトポロジー検出システム。     Topology management means
    When the connection detection protocol packet is transmitted to the first port belonging to the first connection non-detection group and the connection detection protocol packet is received from the second port belonging to the second connection non-detection group, the first The information indicating the second port is moved to the first connection detection group, the information indicating the second port is moved to the second connection detection group,
    The connection detection protocol packet is transmitted to the first port again, and the connection between the first port and the second port is maintained by receiving the connection detection protocol packet from the second port. The network topology detection system according to claim 1, wherein the network topology detection system is confirmed.
  3. 　トポロジー管理手段は、
    　第１のポートに接続検出プロトコルパケットを送信し、第２のポートから当該接続検出プロトコルパケットを受信することにより前記第１のポートと前記第２のポートとの間の接続維持を確認した所定期間後に、再度、前記第１のポートと前記第２のポートとの間の接続維持を確認する場合、前記第２のポートに接続検出プロトコルパケットを送信し、前記第１のポートから当該接続検出プロトコルパケットを受信することにより前記第１のポートと前記第２のポートとの間の接続維持を確認する
    　請求項２記載のネットワークトポロジー検出システム。     Topology management means
    A predetermined period in which the connection maintenance between the first port and the second port is confirmed by transmitting the connection detection protocol packet to the first port and receiving the connection detection protocol packet from the second port. When the connection maintenance between the first port and the second port is confirmed again later, a connection detection protocol packet is transmitted to the second port, and the connection detection protocol is transmitted from the first port. The network topology detection system according to claim 2, wherein the connection maintenance between the first port and the second port is confirmed by receiving a packet.
  4. 　同一のスイッチのポートは、同一の接続未検出グループに記憶される
    　請求項１から請求項３のうちのいずれか１項記載のネットワークトポロジー検出システム。     The network topology detection system according to any one of claims 1 to 3, wherein ports of the same switch are stored in the same unconnected detection group.
  5. 　相互にパケットの送受信を行う複数のスイッチを制御する制御装置であって、
    　前記複数のスイッチのポートのうち接続が検出されていないポートを示す情報を保持する複数の接続未検出グループ、および前記複数のスイッチのポートのうち接続が検出されたポートを示す情報を保持する複数の接続検出グループを記憶する記憶手段と、
    　前記接続未検出グループに属するポートに、当該複数のスイッチ間の接続状態を検出するためのパケットである接続検出プロトコルパケットをグループ毎に送信するトポロジー管理手段とを備え、
    　前記トポロジー管理手段は、
    　第１の接続未検出グループに属する第１のポートに前記接続検出プロトコルパケットを送信した後に、第２の接続未検出グループに属する第２のポートから当該接続検出プロトコルパケットを受信した場合に、当該第１のポートを示す情報および当該第２のポートを示す情報を前記接続検出グループに移動させる
    　ことを特徴とする制御装置。     A control device that controls a plurality of switches that transmit and receive packets to and from each other,
    A plurality of connection undetected groups that hold information indicating ports in which connection is not detected among the ports of the plurality of switches, and a plurality that hold information indicating ports in which connections are detected among the ports of the plurality of switches. Storage means for storing the connection detection group of
    Topology management means for transmitting a connection detection protocol packet, which is a packet for detecting a connection state between the plurality of switches, to a port belonging to the connection undetected group for each group;
    The topology management means includes
    When the connection detection protocol packet is received from the second port belonging to the second connection undetected group after transmitting the connection detection protocol packet to the first port belonging to the first connection undetected group, A control apparatus, wherein information indicating a first port and information indicating a second port are moved to the connection detection group.
  6. 　トポロジー管理手段は、
    　第１の接続未検出グループに属する第１のポートに接続検出プロトコルパケットを送信し、第２の接続未検出グループに属する第２のポートから当該接続検出プロトコルパケットを受信した場合に、前記第１のポートを示す情報を第１の接続検出グループに移動させ、前記第２のポートを示す情報を第２の接続検出グループに移動させ、
    　再度、前記第１のポートに接続検出プロトコルパケットを送信し、前記第２のポートから当該接続検出プロトコルパケットを受信することにより前記第１のポートと前記第２のポートとの間の接続が維持されていることを確認する
    　請求項５記載の制御装置。     Topology management means
    When the connection detection protocol packet is transmitted to the first port belonging to the first connection non-detection group and the connection detection protocol packet is received from the second port belonging to the second connection non-detection group, the first The information indicating the second port is moved to the first connection detection group, the information indicating the second port is moved to the second connection detection group,
    The connection detection protocol packet is transmitted to the first port again, and the connection between the first port and the second port is maintained by receiving the connection detection protocol packet from the second port. The control apparatus according to claim 5, wherein the control apparatus is configured to confirm that the control is performed.
  7. 　相互にパケットの送受信を行う複数のスイッチと、前記複数のスイッチを制御する制御装置とを備えたネットワークトポロジー検出システムに用いられるネットワークトポロジー検出方法であって、
    　前記複数のスイッチのポートのうち接続が検出されていないポートを示す情報を保持する複数の接続未検出グループ、および前記複数のスイッチのポートのうち接続が検出されたポートを示す情報を保持する複数の接続検出グループを記憶し、
    　前記接続未検出グループに属するポートに、当該複数のスイッチ間の接続状態を検出するためのパケットである接続検出プロトコルパケットをグループ毎に送信し、
    　第１の接続未検出グループに属する第１のポートに前記接続検出プロトコルパケットを送信した後に、第２の接続未検出グループに属する第２のポートから当該接続検出プロトコルパケットを受信した場合に、当該第１のポートを示す情報および当該第２のポートを示す情報を前記接続検出グループに移動させる
    　ことを特徴とするネットワークトポロジー検出方法。     A network topology detection method used in a network topology detection system comprising a plurality of switches that transmit and receive packets to each other and a control device that controls the plurality of switches,
    A plurality of connection undetected groups that hold information indicating ports in which connection is not detected among the ports of the plurality of switches, and a plurality that hold information indicating ports in which connections are detected among the ports of the plurality of switches. Remember the connection detection group of
    A connection detection protocol packet, which is a packet for detecting a connection state between the plurality of switches, is transmitted to a port belonging to the connection undetected group for each group.
    When the connection detection protocol packet is received from the second port belonging to the second connection undetected group after transmitting the connection detection protocol packet to the first port belonging to the first connection undetected group, A network topology detection method, comprising: moving information indicating a first port and information indicating the second port to the connection detection group.
  8. 　第１の接続未検出グループに属する第１のポートに接続検出プロトコルパケットを送信し、第２の接続未検出グループに属する第２のポートから当該接続検出プロトコルパケットを受信した場合に、前記第１のポートを示す情報を第１の接続検出グループに移動させ、前記第２のポートを示す情報を第２の接続検出グループに移動させ、
    　再度、前記第１のポートに接続検出プロトコルパケットを送信し、前記第２のポートから当該接続検出プロトコルパケットを受信することにより前記第１のポートと前記第２のポートとの間の接続が維持されていることを確認する
    　請求項７記載のネットワークトポロジー検出方法。     When the connection detection protocol packet is transmitted to the first port belonging to the first connection non-detection group and the connection detection protocol packet is received from the second port belonging to the second connection non-detection group, the first The information indicating the second port is moved to the first connection detection group, the information indicating the second port is moved to the second connection detection group,
    The connection detection protocol packet is transmitted to the first port again, and the connection between the first port and the second port is maintained by receiving the connection detection protocol packet from the second port. The network topology detection method according to claim 7, wherein the network topology detection method is confirmed.
  9. 　相互にパケットの送受信を行う複数のスイッチと、前記複数のスイッチを制御する制御装置とを備えたネットワークトポロジー検出システムに用いられるネットワークトポロジー検出プログラムであって、
    　コンピュータに、
    　前記複数のスイッチのポートのうち接続が検出されていないポートを示す情報を保持する複数の接続未検出グループ、および前記複数のスイッチのポートのうち接続が検出されたポートを示す情報を保持する複数の接続検出グループを記憶する記憶処理と、
    　前記接続未検出グループに属するポートに、当該複数のスイッチ間の接続状態を検出するためのパケットである接続検出プロトコルパケットをグループ毎に送信するトポロジー管理処理とを実行させ、
    　前記トポロジー管理処理で、
    　第１の接続未検出グループに属する第１のポートに前記接続検出プロトコルパケットを送信した後に、第２の接続未検出グループに属する第２のポートから当該接続検出プロトコルパケットを受信した場合に、当該第１のポートを示す情報および当該第２のポートを示す情報を前記接続検出グループに移動させる
    　ためのネットワークトポロジー検出プログラム。     A network topology detection program used in a network topology detection system comprising a plurality of switches that transmit and receive packets to each other and a control device that controls the plurality of switches,
    On the computer,
    A plurality of connection undetected groups that hold information indicating ports in which connection is not detected among the ports of the plurality of switches, and a plurality that hold information indicating ports in which connections are detected among the ports of the plurality of switches. A storage process for storing the connection detection group of
    A port belonging to the connection undetected group, and a topology management process for transmitting a connection detection protocol packet, which is a packet for detecting a connection state between the plurality of switches, for each group;
    In the topology management process,
    When the connection detection protocol packet is received from the second port belonging to the second connection undetected group after transmitting the connection detection protocol packet to the first port belonging to the first connection undetected group, A network topology detection program for moving information indicating a first port and information indicating the second port to the connection detection group.
  10. 　トポロジー管理処理で、
    　第１の接続未検出グループに属する第１のポートに接続検出プロトコルパケットを送信し、第２の接続未検出グループに属する第２のポートから当該接続検出プロトコルパケットを受信した場合に、前記第１のポートを示す情報を第１の接続検出グループに移動させ、前記第２のポートを示す情報を第２の接続検出グループに移動させ、
    　再度、前記第１のポートに接続検出プロトコルパケットを送信し、前記第２のポートから当該接続検出プロトコルパケットを受信することにより前記第１のポートと前記第２のポートとの間の接続が維持されていることを確認させる
    　請求項９記載のネットワークトポロジー検出プログラム。     In topology management processing,
    When the connection detection protocol packet is transmitted to the first port belonging to the first connection non-detection group and the connection detection protocol packet is received from the second port belonging to the second connection non-detection group, the first The information indicating the second port is moved to the first connection detection group, the information indicating the second port is moved to the second connection detection group,
    The connection detection protocol packet is transmitted to the first port again, and the connection between the first port and the second port is maintained by receiving the connection detection protocol packet from the second port. The network topology detection program according to claim 9, wherein the network topology detection program according to claim 9 is confirmed.
PCT/JP2014/006140 2014-02-03 2014-12-09 Network topology detection system, method and program, and control apparatus WO2015114714A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2014018185 2014-02-03
JP2014-018185 2014-02-03

Publications (1)

Publication Number Publication Date
WO2015114714A1 true WO2015114714A1 (en) 2015-08-06

Family

ID=53756337

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2014/006140 WO2015114714A1 (en) 2014-02-03 2014-12-09 Network topology detection system, method and program, and control apparatus

Country Status (1)

Country Link
WO (1) WO2015114714A1 (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2011080870A1 (en) * 2009-12-28 2011-07-07 日本電気株式会社 Communications system and port information collection method
WO2011108205A1 (en) * 2010-03-05 2011-09-09 日本電気株式会社 Communication system, path control apparatus, packet forwarding apparatus and path control method
WO2013115177A1 (en) * 2012-01-30 2013-08-08 日本電気株式会社 Network system and topology management method

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2011080870A1 (en) * 2009-12-28 2011-07-07 日本電気株式会社 Communications system and port information collection method
WO2011108205A1 (en) * 2010-03-05 2011-09-09 日本電気株式会社 Communication system, path control apparatus, packet forwarding apparatus and path control method
WO2013115177A1 (en) * 2012-01-30 2013-08-08 日本電気株式会社 Network system and topology management method

Similar Documents

Publication Publication Date Title
US9059909B2 (en) System and method for configuration of link aggregation groups
US9455916B2 (en) Method and system for changing path and controller thereof
US9876698B2 (en) Interconnect congestion control in a storage grid
EP2983342A2 (en) Internet protocol addressing of industrial control devices employing the network ring topology
US10491504B2 (en) System for support in the event of intermittent connectivity, a corresponding local device and a corresponding cloud computing platform
CN105591817B (en) A kind of negotiation mode processing method and Intelligent Network Element
CN109347674B (en) Data transmission method and device and electronic equipment
KR101576833B1 (en) Redundancy Check Method for Communication Data using Receiving Buffer in Redundancy Apparatus
US10234841B2 (en) Programmable logic controller, slave device, and duplex system
WO2015114714A1 (en) Network topology detection system, method and program, and control apparatus
CN103684888A (en) Method for configuring network equipment in batch
US11258637B2 (en) Method for operating TSN-enabled network coupling elements
CN109067666B (en) Message transmission method and device
JP5212152B2 (en) Communications system
JP6062394B2 (en) Relay device and relay method
KR101544106B1 (en) method for access to SDN using single Ethernet port
JP5922601B2 (en) Network equipment
CN113328917B (en) Method for EPA link diagnosis, EPA device and computer storage medium
KR20120089931A (en) Automatic system comprising apparatus for communication interlock
JP6162996B2 (en) Network control apparatus and method
JP6367630B2 (en) Redundant system, redundant device and switching method thereof
KR101136401B1 (en) Plant multiplexing communication system and method for lossless switching
JP6220916B2 (en) Slave device and control method thereof
JP2013168815A (en) Apparatus and method for identifying transmission source causing congestion
JP6098212B2 (en) How to build a multicast tree

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: 14881148

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: 14881148

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: JP