JP5622687B2 - Interconnection node, communication system and communication method - Google Patents

Description

  The present invention relates to an interconnection node, a communication system, and a communication method.

  In a network where a plurality of rings are conventionally connected, when a failure occurs at a plurality of locations on a single ring, it is impossible to bypass the failure within the ring. For example, when connecting ERP (Ethernet (registered trademark) Ring Protection Switching) rings, if one node is connected between adjacent rings, a failure occurs at one location in a ring. Although a detour is formed by ERP and communication is restored, if a failure occurs at a plurality of locations in a single ring (hereinafter referred to as multiple failures), a detour cannot be formed in the ring.

  In such a case, Non-Patent Document 1 defines a method for configuring a detour route via another ring. In the technique described in Non-Patent Document 1, adjacent rings are connected by two nodes called interconnection nodes.

  On the other hand, the failure notification by ERP shows that a failure has occurred in the ring, but the failure node cannot be specified. Therefore, it cannot be determined that there is a multiple failure. As a method for solving this, there is a method using Non-Patent Document 2. In Non-Patent Document 2, a failure node is specified by using ERP and network topology.

  In addition, even in the case where one node is connected between adjacent rings, in the case of a network configuration in which each ring is connected to two rings (hereinafter referred to as a chain ring network), multiplexing is performed within a single ring. Even if a failure occurs and failure detouring becomes impossible in the ring, failure detouring is possible by relaying between rings. For example, when the chain network includes rings R1 to R4, the ring R1 and the ring R2 are interconnected by the node M1, the ring R2 and the ring R3 are interconnected by the node M2, and the ring R3 and the ring R4 are interconnected. This is a daisy chain network in which a ring M4 is interconnected by a node M3, and a ring R4 and a ring R1 are interconnected by a node M4.

“G.8032 / Y.1344 Ethernet Ring Protection Switching”, ITU-T, Jun 2008. Yuji Ogawa, Katsuyoshi Takahashi, Kazuyuki Kashima, “Study on Faster Layer 3 Fault Switching in Ethernet Ring”, IEICE General Conference 2009 B-8-45, Mar 2010.

  However, according to the technique described in Non-Patent Document 1, in order to configure a detour path even when multiple failures occur, it is necessary to connect adjacent rings with two nodes called interconnection nodes. There is. Therefore, there is a problem that the number of nodes increases.

  In addition, in the case of a chain network configuration, it is possible to configure a detour path even when multiple failures occur with one node connecting adjacent rings, but other than when multiple failures occur There was a problem of forming a loop.

  Moreover, although the specification of a failure node is described in Non-Patent Document 2, Non-Patent Document 2 does not describe connection of a plurality of rings.

  The present invention has been made in view of the above, and in a chain network configuration, it is possible to construct a fault bypass path when multiple faults that cannot be bypassed within a single ring occur without forming a loop. It is an object to obtain an interconnection node, a communication system and a communication method that can be used.

In order to solve the above-described problems and achieve the object, the present invention provides the mutual communication in a communication system in which a plurality of rings are connected in a ring by providing one interconnection node between the rings. In the communication system, one of the interconnection nodes is a two-port blocking node that sets two ports of one ring among the rings to which the interconnection node is connected. When the node is not a two-port blocking node, one ring for a ring I / F that relays a frame received from one of two rings connected to the own node to the other ring and a ring connected to the own node In the case of detecting a multiple failure, a multiple failure detection unit that detects multiple failures in which multiple failures occur, A multiple failure notification unit that transmits a serious failure detection notification to the other interconnection node, and two nodes connected to one of the two rings to which the own node is connected when the own node is a two-port blocking node A port is set as a blocking port, and when the multiple failure is detected or when the multiple failure detection notification is received, a blocking port control unit for releasing blocking of one of the two blocked ports, and A transfer path initialization unit that instructs other nodes in the communication system to initialize transfer information when blocking of one of the two ports blocked by the blocking port control unit is released It is characterized by that.

  According to the present invention, an increase in the number of interconnection nodes is suppressed, and a fault detour path can be constructed when multiple faults that cannot be detoured within a single ring occur without forming a loop.

FIG. 1 is a diagram showing a configuration example of a communication system according to the present invention. FIG. 2 is a diagram illustrating a configuration example of a conventional communication system to which the technique described in Non-Patent Document 1 is applied. FIG. 3 is a diagram illustrating an example of a state before a multiple failure occurs in a conventional chain ring network. FIG. 4 is a diagram illustrating an example of a state after multiple failures occur in the chain ring network illustrated in FIG. FIG. 5 is a diagram illustrating a functional configuration example of the interconnection node. FIG. 6 is a diagram illustrating a functional configuration example of a single ring node. FIG. 7 is a diagram illustrating an example of a state during normal operation. FIG. 8 is a diagram illustrating an example of a procedure for detecting a minimum interconnection node. FIG. 9 is a diagram illustrating an example of operation when multiple failures occur. FIG. 10 is a diagram illustrating an example of an operation when multiple failures are resolved.

  Hereinafter, embodiments of an interconnection node, a communication system, and a communication method according to the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited to the embodiments.

Embodiment.
FIG. 1 is a diagram showing a configuration example of a communication system according to the present invention. In the communication system of this embodiment, as shown in FIG. 1, adjacent rings are connected by one node, all the rings are connected in a daisy chain, and all the rings form one large ring. A network (chain ring network) configuration is assumed.

  1 is composed of Ethernet (registered trademark) rings (rings) 2-1 to 2-4, and nodes (communication devices) 1-1 to 1-4 are connected between adjacent rings. It is an interconnection node. Hereinafter, a node connecting the rings in this way is referred to as an interconnection node. As shown in FIG. 1, the node 1-1 connects the ring 2-1 and the ring 2-2, the node 1-2 connects the ring 2-2 and the ring 2-3, and the node 1-3 connects to the ring 2 -3 and the ring 2-4 are connected, and the node 1-4 connects the ring 2-4 and the ring 2-1. Each ring is always two nodes and is connected to two different rings. For example, in the ring 2-1, the nodes 1-1 and 1-4 are connected to the rings 2-2 and 2-4. Although not shown in FIG. 1, nodes connected to a single ring are connected to the rings 2-1 to 2-4 in addition to the interconnection nodes.

  Similarly, the chain ring network on the right side of FIG. 1 is composed of rings 2-5 to 2-10, and nodes 1-5 to 1-10 are interconnecting nodes that connect adjacent rings. Although FIG. 1 shows two cases of four rings (left side) and five rings (right side), the number of rings constituting the chain ring network is not limited to these.

  Here, failure detection in a conventional communication system configured by connecting a plurality of rings will be described. When each ring applies ERP, when a failure occurs in the ring, a detour path that avoids this failure is configured by ERP. However, when multiple faults occur in a single ring, it is impossible to bypass the fault in that ring. Non-Patent Document 1 describes a method of configuring a detour route that passes between rings when a plurality of rings are connected and failure detouring within a single ring becomes impossible.

  FIG. 2 is a diagram illustrating a configuration example of a conventional communication system to which the technique described in Non-Patent Document 1 is applied. This communication system includes rings 110, 120, 130, and 140, and the ring 110 and the ring 120 are connected to each other by a node 201 and a node 202 that are interconnection nodes. Similarly, the ring 120 and the ring 130 are connected by the node 203 and the node 204, the ring 130 and the ring 140 are connected by the node 205 and the node 206, and the ring 140 and the ring 110 are connected by the node 207 and the node 208, respectively. ing. In addition to the interconnection nodes, nodes 111, 112, 113, 114, 121, 131, 132, and 141, which are single ring nodes connected only to a single ring, are connected to each ring. Here, only the nodes 111, 112, 113, 114, 121, 131, 132, and 141 are illustrated, but a single ring node may be connected to each ring.

  As illustrated in FIG. 2, when two interconnection nodes are provided between each ring and the method described in Non-Patent Document 1 makes it possible to bypass a failure within a single ring due to multiple failures. A detour route passing between the rings can be configured. However, this method has a problem that it is necessary to connect each ring with two interconnection nodes, and the number of nodes increases.

  On the other hand, when each ring is connected by a single interconnection node, if a chain ring network configuration as shown in FIG. 1 is used, a fault detour within a single ring becomes impossible. A detour route that passes between them can be configured.

  FIG. 3 is a diagram illustrating an example of a state before a multiple failure occurs in a conventional chain ring network. FIG. 4 is a diagram illustrating an example of a state after multiple failures occur in the chain ring network illustrated in FIG. 3 and 4, a chain network is configured by the rings 120, 130, and 140, and the ring 110 and the ring 120 are connected by a node 201 that is an interconnection node. Similarly, the ring 120 and the ring 130 are connected by the node 202, the ring 130 and the ring 140 are connected by the node 203, and the ring 140 and the ring 110 are connected by the node 204, respectively. In addition to the interconnection nodes, nodes 111, 112, 113, 114, 121, 131, 132, and 141, which are nodes connected only to a single ring, are connected to each ring. Here, only the nodes 111, 112, 113, 114, 121, 131, 132, and 141 are illustrated, but a single ring node may be connected to each ring.

  As shown in FIG. 3, before the multiple failure occurs, in the ring 110, the node 112 is set as an RPL (Ring Protection Link) owner, and the blocking port 115 is set in the node 112. Similarly, in the ring 120, the node 121 is set as the RPL owner, the blocking port 122 is set in the node 121, the node 132 is set as the RPL owner in the ring 130, the blocking port 133 is set in the node 132, and the ring In 140, the node 141 is set as the RPL owner, and the blocking port 142 is set in the node 141.

  In the network configuration shown in FIG. 3, when a failure occurs at one location in a ring, a detour is formed by ERP and communication is restored, but failure occurs at multiple locations on a single ring (multiple failures). Cannot form a detour in the ring.

  Assume that failures 301 and 302 occur in the ring 110 in the network configuration shown in FIG. Since two faults have occurred in the ring 110, it is not possible to construct a detour path by ERP in the ring 110. That is, due to this multiple failure, communication between the node 111 and the node 112 in the ring 110 and between the node 113 and the node 114 cannot be bypassed in the ring 110.

  At this time, if a route other than the ring 110 (ring 120, ring 130, ring 140) can be routed, a detour can be formed. For example, the interconnect node is a ring relay between the ring 110 and the ring 120, and the interconnect node 202 relays the frame between the ring 120 and the ring 130. Assume that frame relay (hereinafter, ring-to-ring relay) is performed. At this time, communication is possible between the node 111 and the node 112 via the node 201, the rings 120, 130, and 140, and the node 204. In addition, communication can be similarly performed between the node 113 and the node 114.

  However, if all interconnection nodes perform inter-ring relaying as described above in a state where multiple failures have not occurred, there is a problem that a loop passing through all the rings is formed and a broadcast storm occurs. In the example of FIG. 4, a loop occurs in the route of the node 113, the node 201, the node 202, the node 131, the node 203, the node 204, and the node 114.

  Therefore, in the present embodiment, a loop is formed without increasing the number of interconnection nodes by performing the operations described below in the chain ring network shown in FIGS. Instead, a detour path is constructed when multiple failures occur.

  Further, the failure notification by ERP shows that a failure has occurred in the ring, but the failure node cannot be specified. Therefore, it cannot be determined that there is a multiple failure. As a technique for solving this, a technique using Non-Patent Document 2 described above is applied in the present embodiment.

  FIG. 5 is a diagram illustrating a functional configuration example of the interconnection node 3 according to the present embodiment. The interconnection node 3 of the present embodiment includes a MAC (Media Access Control) processing unit (MAC) 11, a physical layer processing unit (PHY) 12, a terminal interface (I / F) 13, and an ERP processing unit (ERP). ) 15, ring I / Fs 14 and 16, network topology management unit 17, multiple failure detection unit 19, multiple failure occurrence / resolution notification unit (multiple failure notification unit) 20, blocking port control unit 21, FDB ( A Forwarding DataBase) flash notification unit (transfer path initialization unit) 22, a minimum interconnection node ID (Identifier) notification unit (minimum identifier notification unit) 23, and an FDB flash notification reception unit 24.

  Each of the ring I / Fs 14 and 16 includes two ports (two ports respectively connected to two adjacent nodes in the ring). The ring I / Fs 14 and 16 perform transfer processing based on FDB (transfer information) held by the own node based on the destinations of the frames received from the rings 4-1 and 4-2. The transfer method using the FDB is the same as that of a conventional ring network node. When a destination frame not registered in the FDB is received, the frame is transferred to all ports and the frame registered in the FDB is received. If it does, the frame is received to the registered transfer destination (port).

  The terminal I / F 13 performs interface processing for communication with a terminal connected to the own node, and the physical layer processing unit 12 transmits / receives data to / from the terminals and between the rings 4-1 and 4-2. Predetermined physical layer processing is performed on transmission / reception data. The MAC processing unit 11 performs predetermined MAC layer processing on transmission / reception data, and passes information necessary for creating a network topology such as adjacent information to the network topology management unit 17. The ERP processing unit 15 performs ERP processing on the transmission / reception data from the rings 4-1 and 4-2, and when receiving a failure notification by ERP, passes the notification to the multiple failure detection unit 19.

  The MAC processing unit 11, the physical layer processing unit 12, the terminal interface 13, the ERP processing unit 15, and the ring I / Fs 14 and 16 are the same as the conventional interconnection node described in Non-Patent Document 1, for example. Therefore, explanation is omitted.

  FIG. 6 is a diagram illustrating a functional configuration example of a node (single ring node) 5 other than the interconnection node according to the present embodiment. The node 5 is a node connected to the single ring 4.

  The single ring node 5 includes a MAC processing unit 11, a physical layer processing unit 12, a terminal interface 13, an ERP processing unit 15, a ring I / F 14, an FDB flash notification receiving unit 24, and an ERP blocking port control unit. 25. Since the MAC processing unit 11, the physical layer processing unit 12, the terminal interface 13, the ERP processing unit 15, and the ring I / F 14 are the same as those of the interconnection node 3, description thereof is omitted. However, since the single ring node 5 does not include the multiple failure detection unit 19 and the network topology management unit 17, the ERP processing unit 15 does not perform an operation of passing the failure notification to the multiple failure detection unit 19, and the MAC processing unit 11 The operation of passing information to the network topology management unit 17 is not performed.

  FIG. 7 is a diagram illustrating an example of a state during normal operation of the communication system according to the present embodiment. As shown in FIG. 7, this communication system forms a chain ring network with rings A, B, C, and D, and includes interconnection nodes 3-1 to 3-4. The interconnection nodes 3-1 to 3-4 have the configuration of the interconnection node 3 shown in FIG. Also, nodes 5-1 to 5-4, which are single ring nodes, are connected to ring A, node 5-5, which is a single ring node, is connected to ring B, and a single ring node is connected to ring C. Nodes 5-6 and 5-7 which are one ring node are connected, and a node 5-8 which is a single ring node is connected to the ring D. The nodes 5-1 to 5-8 have the configuration of the node 5 shown in FIG.

  The network topology management unit 17 holds the network topology as a topology list 18. The topology list 18 is used to determine that a failure occurring in the ring is a multiple failure. The topology list 18 may be created by any method, for example, by the method described in Non-Patent Document 2.

  The multiple failure detection unit 19 identifies a failure occurrence node based on the topology list 18 held by the network topology management unit 17 and the failure notification by ERP, and detects the occurrence of multiple failure. The failure occurrence node is identified by a method described in Non-Patent Document 2, for example.

  The multiple failure occurrence / removal notification unit 20 notifies all interconnection nodes in the communication system of the occurrence or elimination of multiple failures based on the detection result of the multiple failure by the multiple failure detection unit 19 (configuration example in FIG. 7). In this case, for example, the interconnection node 3-1 notifies the interconnection nodes 3-2 to 3-4).

  The blocking port control unit 21 sets or cancels the blocking of a port according to the presence or absence of the occurrence of multiple failures in order to prevent occurrence of loops during normal times and to form detours via multiple rings when multiple failures occur. In addition, the blocking port release of the RPL owner set by ERP is performed in the ring to which the own node is connected.

  Since the blocking setting performed by the blocking port control unit 21 is performed by only one node among a plurality of interconnection nodes, it is necessary to select that node. The minimum interconnection node ID notification unit 23 transmits and receives a minimum interconnection node ID notification necessary for the selection process. The minimum interconnection node ID notification will be described later.

  The FDB flash notification unit 22 instructs all nodes to perform FDB flash (transfer information initialization) when switching to a route through a plurality of rings due to multiple failures.

  When the FDB flash notification receiving unit 24 receives the FDB flash notification transmitted by the FDB flash notification unit 22 of the interconnection node 3, the FDB flash notification receiving unit 24 executes the FDB flash.

  When the own node is operating as the RPL owner, the ERP blocking port control unit 25 receives the blocking port release instruction transmitted by the blocking port control unit 21 of the interconnection node, and sets the own node as the RPL owner. To release the blocking.

  The normal operation (a state in which no failure has occurred) of the present embodiment will be described with reference to FIG. In FIG. 7, examples of blocking ports set by ERP are illustrated as blocking ports 6-1 to 6-4. With ERP, one blocking port is set in each ring.

  In the present embodiment, in order to prevent the occurrence of a loop passing through a plurality of rings, one of the interconnection nodes 3-1 to 3-4 is connected to one ring by the operation of the blocking port control unit 21. The two ports are set to block, and frame relay between rings is not performed. A node that blocks two ports in this way is hereinafter referred to as a two-port blocking node.

  The interconnection nodes 3-1 to 3-4 that are not two-port blocking nodes do not block two ports, so that frames are relayed between rings. That is, the frame received via the ring I / F 14 is relayed within the ring via the ring I / F 14 and also relayed to other rings via the ring I / F 16. Similarly, a frame received via the ring I / F 16 is relayed within the ring via the ring I / F 16 and also relayed to other rings via the ring I / F 14. However, when the 2-port blocking node is not yet determined, the ring I / Fs 14 and 16 perform intra-ring relaying and do not perform inter-ring relaying.

  FIG. 7 shows an example in which the interconnection node 3-2 is a two-port blocking node, and two ports 7-1 and 7-2 connected to the ring B are blocked. Since the other interconnection nodes 3-1, 3-3, 3-4 are not 2-port blocking nodes, 2-port blocking is not performed.

  Since ring B is blocked by ports 7-1 and 7-2 of interconnection node 3-2, if blocking is set for ports other than these ports in ring B, a node that cannot communicate is generated. . Therefore, in ring B, the blocking port control unit 21 of the interconnection node 3-2 instructs to release the blocking port 6-2 of the RPL owner set by ERP. In the other rings A, C, and D, the blocking ports 6-1, 6-3, and 6-4 set by ERP are not released.

  Selection of the 2-port blocking node is performed by communication between the interconnection nodes 3-1 to 3-4. It is assumed that each interconnection node 3-1 to 3-4 has a unique interconnection node ID, and the interconnection node having the smallest interconnection node ID is a 2-port blocking node. Here, the interconnection node having the smallest interconnection node ID is a two-port blocking node. However, the present invention is not limited to this, and the interconnection node having the largest interconnection node ID is a two-port blocking node. For example, the method of selecting a 2-port blocking node is not limited to this.

  The interconnection nodes 3-1 to 3-4 detect a node having the minimum interconnection node ID (minimum interconnection node) in the communication system in order to determine whether or not the own node becomes a 2-port blocking node. To do. In this detection of the minimum interconnection node, each of the interconnection nodes 3-1 to 3-4 transmits a minimum interconnection node ID notification for notifying the minimum interconnection node ID known by the own node. Realize with. The minimum interconnection node ID notification unit 23 transmits and receives the minimum interconnection node ID notification.

  FIG. 8 is a diagram illustrating an example of a procedure for detecting a minimum interconnection node. FIG. 8 shows an example in which the interconnection node 3-3 detects the minimum interconnection node as an example. The interconnection node IDs of the interconnection nodes 3-1, 3-2, 3-3, and 3-4 are 1, 2, 3, and 4, respectively.

  First, in the procedure 1, it is assumed that the selection of the 2-port blocking node is not completed. At this stage, all the interconnection nodes 3-1 to 3-4 stop the inter-ring relay (step S1). That is, the interconnection nodes 3-1 to 3-4 relay the received frame in the ring that received the frame, and do not relay it to other rings.

  In procedure 2, all the interconnection nodes 3-1 to 3-4 transmit the minimum interconnection node ID notification by the operation of the minimum interconnection node ID notification unit 23 (step S2). In FIG. 8, only the interconnection node ID notification from the interconnection node 3-3 is shown for simplification of the drawing, but the same minimum interconnection node ID notification is transmitted from the other interconnection nodes. Is done. It is assumed that the minimum interconnection node ID notification includes information on the source interconnection node ID and the minimum interconnection node ID. When the minimum interconnection node ID notification is transmitted from the own node (the own node is the transmission source), the minimum interconnection node ID stores the interconnection node ID of the own node. Therefore, in the procedure 2 of FIG. 8, the interconnection node ID of the own node is stored in the minimum interconnection node ID and transmitted. This minimum interconnection node ID notification is transmitted only to one of the two connected rings. In the example of FIG. 8, it is assumed that the interconnection node 3-3 transmits a minimum interconnection node ID notification to the ring D among the rings C and D to be connected.

  In FIG. 8, the interconnection node ID and the minimum interconnection node ID stored in the minimum interconnection node ID notification are shown in parentheses, and the notification transmitted in step S2 is the source interconnection node ID itself. And the minimum interconnection node ID is set to 3 which is its own ID (indicated as (3, 3) in FIG. 8).

  In procedure 3 of FIG. 8, the interconnection node 3-4 receives the minimum interconnection node ID notification transmitted by the interconnection node 3-3 in the ring D, and forwards it to the ring A (to the interconnection node 3-1). (Step S3). When performing this transfer, the minimum interconnection node ID notification unit 23 of the interconnection node 3-4 indicates the minimum interconnection node ID in the received minimum interconnection node ID notification, that is, the interconnection node ID of the own node. If the interconnect node ID of the own node is smaller than the minimum interconnect node ID, the minimum interconnect node ID is rewritten to the interconnect node ID of the own node, and if the interconnect node ID of the own node is greater, Transfer without rewriting the connection node ID. In step S3 of FIG. 8, since the interconnection node ID of the own node (ie, 4) is larger than the minimum interconnection node ID (ie, 3) in the minimum interconnection node ID notification, the minimum interconnection node ID is not rewritten. To transfer.

  Similarly, the interconnection node 3-1 receives the minimum interconnection node ID notification transmitted from the interconnection node 3-3 via the interconnection node 3-4, and automatically receives the minimum interconnection node ID in the notification. Since the interconnect node ID of the node is a smaller value, the minimum interconnect node ID is changed to the interconnect node ID of the own node (that is, 1) and transferred to the ring B (step S4). In this way, the minimum interconnection node ID notification is transferred between the interconnection nodes on the chain ring network.

  In the procedure 4 of FIG. 8, the minimum interconnection node ID notification transmitted by the interconnection node 3-3 goes around the chain ring and is received by the transmission source interconnection node ID 3-3 (step S5). Since the transmission source interconnection node ID in the minimum interconnection node ID notification is equal to 3 which is the interconnection node ID of the interconnection node 3-3, the minimum interconnection node ID notification transmitted by itself is chained. It is determined that it has made a round of the ring, and no transfer is made to the other ring. Based on the minimum interconnection node ID in the received minimum interconnection node ID notification, the interconnection node 3-3 indicates that the interconnection node having the minimum interconnection node ID in the communication system is the interconnection node 3-1. Is detected. Similarly, the other interconnection nodes 3-1, 3-2 and 3-4 also detect that the interconnection node having the minimum interconnection node ID in the communication system is the interconnection node 3-1. In this case, the interconnection node 3-1 is selected as 2-port blocking.

  As a result of the minimum interconnection node ID notification being cycled on the chain ring network, the interconnection nodes 3-1 to 3-4 have found that the interconnection node having the minimum interconnection node ID is not its own node. In this case, inter-ring relay is started without performing 2-port blocking (step S6). When the interconnection node having the minimum interconnection node ID is found to be the own node, the blocking port control unit 21 sets 2-port blocking for the own node.

  Further, the interconnection node that sets 2-port blocking transmits a blocking port release instruction to the RPL owner of the ring to which the port for which 2-port blocking is set is connected by the operation of the blocking port control unit 21. The RPL owner (node 5-5 in the example of FIG. 7) that has received the blocking port release instruction releases the blocking set as the RPL owner by the operation of the ERP blocking port control unit 25.

  The interconnection nodes 3-1 to 3-4 that have completed the detection of the minimum interconnection node ID perform flushing of the FDB held by themselves. Then, by the operation of the FDB flash notification unit 22, the FDB flash notification is transmitted to all nodes in the ring to which the own node is connected (step S6). In the nodes 5-1 to 5-8 other than the interconnection node, when the FDB flash notification is received, the FDB flash notification receiving unit 24 performs the FDB flash.

  Next, the operation when multiple failures occur will be described with reference to FIG. FIG. 9 is a diagram illustrating an example of an operation when multiple failures occur according to the present embodiment. It is assumed that multiple failures due to failures 9-1 and 9-2 occur in ring A. In the example of FIG. 9, it is assumed that the interconnection node 3-2 is 2-port blocking, as in the example of FIG.

  The interconnection nodes 3-1 and 3-4 connected to the ring A detect the occurrence of a failure in the ring A by the ERP failure detection function. The ERP function alone cannot detect that a failure has occurred between the node 5-1 and the node 5-2 and between the node 5-3 and the node 5-4. Therefore, in the present embodiment, the failure occurrence node is specified by the method of Non-Patent Document 2.

  The interconnection nodes 3-1 to 3-4 grasp the network topology in the ring to which the interconnection nodes 3-1 to 3-4 are connected in advance by the method of Non-Patent Document 2 and identify the network topology management unit 17 The topology list 18 is stored. When a failure occurs, the multiple failure detection unit 19 identifies a failure occurrence node based on the failure notification received by the ERP failure detection function and the topology list 18, and detects the occurrence of multiple failures.

  In the example of FIG. 9, the multiple failure detection unit 19 of the interconnection node 3-1 detects that a failure has occurred in the node 5-1 and the node 5-2. Similarly, the interconnection node 3-4 detects that a failure has occurred in the nodes 5-3 and 5-4.

  The interconnection nodes 3-1 and 3-4 that have detected multiple failures transmit multiple failure notifications T <b> 1 and T <b> 2 to all the interconnection nodes by the operation of the multiple failure occurrence / resolution notification unit 20. If the own node is 2-port blocking, the multiple failure notifications T1 and T2 may not be transmitted, and the operation may be shifted to the operation of an interconnection node in which 2-port blocking described later is set.

  In the interconnection node 3-2 in which the 2-port blocking is set, when the blocking port control unit 21 receives the multiple failure occurrence notification T <b> 1 or T <b> 2, the blocking of one port out of the two blocked ports is released. . FIG. 9 shows an example in which the blocking of the port 7-2 is canceled among the blocked ports 7-1 and 7-2.

  In the interconnection node 3-2, after the blocking of the port 7-2 is released, the FDB flash notification unit 22 transmits FDB flash notifications T3 and T4 instructing deletion of the FDB entry to all the nodes in the network. Prompt to change the relay route.

  In the nodes (interconnection nodes 3-1, 3-3, 3-4, nodes 5-1 to 5-8) that have received the FDB flash notification T3 or T4, the FDB flash notification receiving unit 24 executes the FDB flash. . Thereby, the nodes in the ring A can communicate using the detour R1 via the rings B, C, and D.

  Next, the operation when multiple failures are resolved will be described with reference to FIG. FIG. 10 is a diagram illustrating an example of an operation when multiple failures are resolved in the present embodiment. FIG. 10 shows an operation when the faults 9-1 and 9-2 described in FIG. 9 are resolved.

  When the multiple failure caused by the failures 9-1 and 9-2 occurring in the ring A is eliminated (at least one of the failures 9-1 and 9-2 is eliminated), the interconnection node 3- connected to the ring A 1 and 3-4 detect the elimination of multiple faults in ring A by the ERP function.

  In the interconnection nodes 3-1 and 3-4 that have detected the multiple fault elimination, the multiple fault occurrence / resolution notification unit 20 transmits multiple fault elimination notices T 5 and T 6 to all the interconnection nodes. The interconnection nodes 3-1 and 3-4 that have detected the elimination of multiple faults and the interconnection nodes 3-2 and 3-3 that have received the multiple fault elimination notification T5 or T6 stop the inter-ring relay and will be described with reference to FIG. The detected minimum interconnection node is detected, and an interconnection node that performs 2-port blocking is selected. The subsequent operation is the same as that before the multiple failure occurs.

  As described above, in this embodiment, adjacent rings are connected to each other by one interconnection node 3, and all interconnection nodes 3 are connected to two rings so that each ring is circular. When multiple failures do not occur, one of the interconnection nodes 3 performs 2-port blocking. When the interconnection node 3 detects a multiple failure, the interconnection node 3 transmits a multiple failure notification and receives the multiple failure notification. The interconnection node 3 that has set the 2-port blocking is one of the two blocked ports. The FDB flush notification is sent to all nodes. Therefore, it is possible to construct a fault detour path even when multiple faults occur without limiting the number of interconnection nodes and forming a loop.

  As described above, the interconnection node, the communication system, and the communication method according to the present invention are useful for a ring network, and are particularly suitable for a communication system that connects a plurality of ring networks.

1-1 to 1-10, 5-1 to 5-8, 111, 112, 113, 114, 121, 131, 132, 141, 201 to 208 Nodes 2-1 to 2-10, 110, 120, 130, 140 ring 3,3-1 to 3-4 interconnection node 7-1,7-2 port 9-1,9-2,301,302 failure 11 MAC processing unit (MAC)
12 Physical layer processing unit (PHY)
13 Terminal interface (I / F)
14,16 Ring I / F
15 ERP processing part (ERP)
17 Network Topology Management Unit 18 Topology List 19 Multiple Failure Detection Unit 20 Multiple Failure Generation / Resolution Notification Unit 21 Blocking Port Control Unit 22 FDB Flash Notification Unit 23 Minimum Interconnection Node ID Notification Unit 24 FDB Flash Notification Reception Unit 25 ERP Blocking Port Control 115, 122, 133, 142, 6-1 to 6-4 Blocking port

Claims (6)

The interconnection node in a communication system in which a plurality of rings are connected in a ring by providing one interconnection node between the rings, each connected to two rings,
In the communication system, one of the interconnection nodes is a two-port blocking node configured to block two ports of one ring among the rings connected to the interconnection node;
A ring I / F that relays a frame received from one of two rings to which the node is connected when the node is not a two-port blocking node;
A multiple failure detection unit that detects multiple failures in which a plurality of failures occur in one ring with respect to the ring to which the node is connected;
When detecting the multiple failure, multiple failure notification unit for transmitting a multiple failure detection notification to the other interconnect node;
When the own node is a two-port blocking node, two ports connected to one of the two rings to which the own node is connected are set as blocking ports and the multiple failure is detected or the multiple failure is detected. A blocking port control unit for releasing blocking of one of the two blocked ports when receiving a detection notification;
A transfer path initialization unit that instructs the other nodes in the communication system to initialize transfer information when blocking of one of the two ports blocked by the blocking port control unit is released;
An interconnection node comprising: An interconnection node identifier is assigned to each of all interconnection nodes in the communication system,
When the two-port blocking node has not been determined, the minimum interconnection in which the interconnection node identifier of the own node is stored as the source node identifier and the interconnection node identifier of the own node is stored as the minimum interconnection node identifier When the node identifier notification is transmitted to one node to which the own node is connected and the minimum interconnection node identifier notification is received from the other interconnection node, the minimum interconnection node identifier included in the notification is the interconnection of the own node. If it is smaller than the node identifier, the notification is transferred to a ring different from the ring of the reception source. If the minimum interconnection node identifier included in the notification is larger than the interconnection node identifier of its own node, the minimum in the notification Rewrite the interconnect node identifier with the interconnect node identifier to make it different from the source ring. When the minimum interconnection node identifier whose source node identifier is the interconnection node identifier of the own node is received, the interconnection node corresponding to the minimum interconnection node identifier of the notification is two ports. A minimum identifier notification unit that recognizes a blocking node;
The interconnect node of claim 1, comprising: When the multiple fault notification unit detects the cancellation of the multiple fault, it transmits a multiple fault cancellation notification to the other interconnection node,
The interconnection node according to claim 2, wherein the minimum identifier notification unit transmits the minimum interconnection node identifier notification when the multiple failure resolution notification is received. An ERP processing unit for detecting a failure by ERP;
A network topology management unit that acquires and holds the network topology;
Further comprising
The interconnection node according to claim 1, wherein the multiple failure detection unit detects the multiple failure based on the network topology and a failure notification by ERP. A communication system in which a plurality of rings are connected in a ring shape,
A single ring node connected to a single said ring;
An interconnection node according to any one of claims 1 to 4 connected to two rings;
The communication system is characterized in that one interconnect node is provided between the rings. A communication method in a communication system in which a plurality of rings are connected in a ring by providing one interconnect node between the rings, each connected to two rings,
In the communication system, one of the interconnection nodes is a two-port blocking node configured to block two ports of one ring among the rings connected to the interconnection node;
A link relay step in which the interconnection node relays a frame received from one of two rings to which the node is connected to the other ring;
A blocking setting step in which the two-port blocking node sets, as blocking ports, two ports connected to one of the two rings to which the node is connected;
A multiple failure detection step in which the interconnection node detects a multiple failure in which a plurality of failures occur in one ring with respect to the ring to which the node is connected;
When the interconnection node detects the multiple failure, a multiple failure notification step of transmitting a multiple failure detection notification to the other interconnection node;
A blocking release step of releasing blocking of one of the two blocked ports when the two-port blocking node detects the multiple failure or receives the multiple failure detection notification;
The 2-port blocking node, when releasing the one blocking of the two ports that are Oite block the deblocking step, instructs the initialization of the transfer information to other nodes in the communication system A transfer path initialization step;
A communication method comprising:

Images