JP2012161027A - Edge relay device, redundant system for edge relay device, wide area network system, and frame transfer method for edge relay device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an edge relay device contributing to ensuring of redundancy, and to provide a redundant system of an edge relay device, a wide area network system, and a frame transfer method for an edge relay device.SOLUTION: An edge relay device (26a) has a relay unit controlling a stream of frames among core relay devices (28a and 28e), an external relay device (24a), and the other edge relay device (26b) via an NNI port (32n), an access port (32a), and an IC port (32i). When the unique MAC address in a network, which is allocated to itself, is set as a "my representative address", and the unique MAC address in the network, which is allocated to other edge relay device, is set as a "mate representative address", the relay unit controls the stream of the frame based on a combination of the my representative address and the mate representative address included in a destination address and a transmission source address of the frame.

Description

  The present invention relates to an edge relay device, a redundant system of edge relay devices, a wide area network system, and a frame transfer method for the edge relay device.

  Some wide area network systems employ a MAC-in-MAC network (core network) (for example, Patent Document 1). The core network is composed of a plurality of switching hubs (relay devices). These switching hubs are edge switches (edge relay devices) arranged at the ends of the core network and core switches (core relays) arranged between the edge switches. Device).

  Each edge switch is assigned a unique MAC address (core network address) in the core network, and each edge switch sends its own core network address to the MAC frame that comes from outside the core network. An encapsulated frame including a header including a core network address of the destination edge switch as a destination address is generated.

  When the edge switch sends the encapsulated frame toward the core switch, the core switch forwards the encapsulated frame based on the destination address included in the header of the encapsulated frame, and the encapsulated frame arrives at the destination edge switch. . The destination edge switch removes the header from the encapsulated frame, returns it to the normal MAC frame, and sends the normal MAC frame outside the core network.

JP 2009-65429 A

  In a MAC-in-MAC network, for example, when an edge switch fails, there is a problem that a local network connected to the failed edge switch cannot communicate through the core network. .

  The present invention has been made in view of the above-described circumstances, and an object of the present invention is to provide an edge relay device, an edge relay device redundancy system, a wide area network system, which contribute to ensuring redundancy in a MAC-in-MAC network, and Another object of the present invention is to provide a frame transfer method for an edge relay device.

  In order to achieve the above object, according to one aspect of the present invention, in an edge relay device disposed at a boundary of a MAC-in-MAC network, a core relay device disposed inside the boundary of the network NNI port used for connection to the network, an access port used for connection to an external relay device arranged outside the network, and a connection to another edge relay device arranged on the boundary. An IC port, and a relay unit that controls the flow of frames between the core relay device, the external relay device, and the other edge relay device through the NNI port, the access port, and the IC port, The unique MAC address in the network assigned to itself is used as the my representative address and assigned to the other edge relay device. When the unique MAC address in the network to be used as the mate representative address, the relay unit, based on the combination of the my representative address and the mate representative address, included in the destination and source of the frame, An edge relay device for controlling the flow is provided.

  In order to achieve the above object, according to one aspect of the present invention, there is provided a redundant system of edge relay devices arranged at the boundary of a MAC-in-MAC network, which includes a first edge relay device and a first edge relay device. 2 edge relay devices, each of the first edge relay device and the second edge relay device is disposed outside the network and an NNI port connected to a core relay device disposed inside the boundary of the network An access port connected to the external relay device, an IC port used for connection between the first edge relay device and the second edge relay device, the NNI port, the access port, and the IC port The flow of frames between the core relay device, the external relay device, the first edge relay device, and the second edge relay device through A unique MAC address in the network assigned to itself as a my representative address, and for the first edge relay device, a unique MAC address in the network assigned to the second edge relay device Is the mate representative address, and, for the second edge relay device, the unique MAC address in the network assigned to the first edge relay device is the mate representative address, the relay unit transmits the destination and source of the frame. The redundant system of the edge relay device is provided that controls the flow of the frame based on the combination of the my representative address and the mate representative address included in the network.

  In order to achieve the above object, according to one aspect of the present invention, a MAC-in-MAC first network including the redundant system of the edge relay device is connected to the first network, and A wide area network system including a plurality of second networks connected to each other via a first network is provided.

  Still further, in order to achieve the above object, according to an aspect of the present invention, in a frame transfer method for an edge relay device arranged at a boundary of a MAC-in-MAC network, the edge relay device includes: An NNI port is connected to a core relay device arranged inside a boundary of the network, an access port of the edge relay device is connected to an external relay device arranged outside the network, and the edge relay device Are connected to other edge relay devices arranged at the boundary, and each of the edge relay devices and the other edge relay devices is assigned a unique MAC address in the network, and is assigned to the edge relay device. The assigned MAC address is my representative address, and the other edge relay device is assigned the When the AC address is a mate representative address, based on the combination of the my representative address and the mate representative address included in the destination and transmission source of the frame, the NNI port, the access port, and the IC port through the IC port are used. There is provided a frame transfer method for an edge relay device that controls a flow of the frame among a core relay device, the external relay device, and the other edge relay device.

  According to the present invention, there are provided an edge relay device, a redundancy system for an edge relay device, a wide area network system, and a frame transfer method for the edge relay device that contribute to ensuring redundancy in a MAC-in-MAC network. The

1 is a diagram schematically illustrating a configuration of a wide area network system according to an embodiment. FIG. It is a figure which shows the format of the B-TAG frame of the IEEE802.1ah standard which propagates the PBB network of the wide area network system of FIG. FIG. 2 is a block diagram schematically showing a configuration of a first edge switch and a second edge switch in the wide area network system of FIG. 1. (A) is a table | surface which shows roughly the content of FDB of a 1st edge switch and a 2nd edge switch, (b) is a table | surface schematically showing the content of the VID table of a 1st edge switch and a 2nd edge switch. It is a table | surface which shows. It is a block diagram for demonstrating the content of port setting DB of a 1st edge switch and a 2nd edge switch, and is a figure for demonstrating the flow of the flame | frame in the 1st pattern of a 1st case. It is a figure for demonstrating the state in which the failure generate | occur | produced in the PB network of the wide area network system of FIG. It is a figure for demonstrating the flow of the frame in the 2nd pattern of a 1st case. It is a figure for demonstrating the flow of the flame | frame at the time of transmission of a 2nd case. It is a figure for demonstrating the flow of the frame at the time of reception of a 2nd case. FIG. 10 is a diagram for describing a flow of a frame when transmitted from an NNI port of a second edge switch in a third case. In the third case, it is a diagram for explaining the flow of the frame when transmitted from the IC port of the second edge switch. It is a figure for demonstrating the flow of the flame | frame in a 4th case. It is a figure for demonstrating the flow of the flame | frame in a 5th case. It is a figure for demonstrating the flow of the flame | frame in a 6th case. It is a figure for demonstrating the flow of the frame in a 7th case. It is a figure for demonstrating the flow of the flame | frame corresponding to a 4th case when an access ring is a normal state.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
FIG. 1 schematically illustrates the configuration of a wide area network system 10 according to an embodiment. Specifically, the wide area network system 10 connects a plurality of user networks 12a, 12b, 12c, 12d and user networks 12a, 12b, 12c, 12d to each other at the layer 2 level using an OSI (Open Systems Interconnection) reference model. And a wide area network 14. Hereinafter, the user networks 12a, 12b, 12c, and 12d are also simply referred to as the user network 12.

[User network]
For example, if the user of the wide area network system 10 has a plurality of branches throughout the country, each user network 12 is provided corresponding to each branch. Each user network 12 includes, for example, user terminals 16a, 16b, 16c, and 16d made of personal computers (PCs), and switching hubs (relay devices) 18a connected to the user terminals 16a, 16b, 16c, and 16d, The switching hubs 18a, 18b, 18c, and 18d are connected to the wide area network 14, including 18b, 18c, and 18d.
The communication system in the user network 12 is based on, for example, the IEEE 802.1Q standard. In the case of this standard, a MAC (Media Access Control) frame includes a C-TAG (customer VLAN tag). Hereinafter, a MAC frame including C-TAG is also referred to as a C-TAG frame.

[Wide area network]
On the other hand, the wide area network 14 is installed by, for example, a communication carrier. The wide area network 14 includes provider bridge networks (PB networks) 20 a, 20 c, 20 d and a provider backbone bridge network (PBB network) 22.
Hereinafter, the PB networks 20a, 20c, and 20d are also simply referred to as the PB network 20.

[PB network]
The user network 12 is connected to the corresponding PB network 20, and the PB network 20 is connected to the PBB network 22. Each of the PB networks 20 is generally configured by a plurality of switching hubs connected to each other. However, in FIG. 1, the PB network 20a is configured by two switching hubs 24a and 24b, and the PB network 20c. , 20d are constituted by one switching hub 24c, 24d, respectively.
Hereinafter, the switching hubs 24a, 24b, 24c, and 24d are also simply referred to as the switching hub 24.

  In FIG. 1, two user networks 12a and 12b are connected to the PB network 20a, and one user network 12c and 12d are connected to the PB networks 20c and 20d, respectively. Note that three or more user networks 12 may be connected to one PB network 20.

  The communication system in the PB network 20 is based on, for example, the IEEE 802.1ad standard. Therefore, the switching hub 24 of the PB network 20 adds an S-TAG (service VLAN tag) different for each user to the C-TAG frame received from the user network 12, thereby providing an S-TAG frame (extended VLAN tag frame). ) And the obtained S-TAG frame is transmitted to the PBB network 22 according to the destination. Conversely, the switching hub 24 removes the S-TAG from the S-TAG frame received from the PBB network 22 and converts it into a C-TAG frame, and converts the obtained C-TAG frame according to the destination. Transmit to the network 12.

[PBB network]
The PBB network (core network) 22 includes a plurality of switching hubs. More specifically, the PBB network (core network) 22 includes a plurality of edge switches (edge relay devices) 26a, 26b, 26c, and 26d, and core switches (core relay devices) 28a, 28b, 28c, 28d, and 28e. Hereinafter, the edge switches 26a, 26b, 26c, and 26d and the core switches 28a, 28b, 28c, 28d, and 28e are also simply referred to as the edge switch 26 and the core switch 28, respectively.

The edge switch 26 is located at the end or boundary of the PBB network 22, that is, at the entrance / exit, and is connected to the switching hub (external relay device) 24 of the PB network 20. The core switch 28 forms a core switch (core SW) network 30 to which the edge switch 26 is connected.
In FIG. 1, the PBB network 22 includes four edge switches 26 and five core switches 28, but the number of edge switches 26 and core switches 28 is generally larger than this. Therefore, the configuration of the core switch 28 in the core SW network 30 is generally more complicated.

The communication system of the PBB network 22 is a MAC-in-MAC system, and conforms to, for example, the IEEE 802.1ah standard. Accordingly, B-TAG frames (backbone TAG frames) are transmitted and received within the PBB network 22.
FIG. 2 shows the format of the B-TAG frame. The B-TAG frame includes a backbone destination address (B-DA), a backbone source address (B-SA), a B-TAG (backbone tag), and an I-TAG (service instance tag) in the S-TAG frame. It has been granted. However, the I-TAG includes a customer destination address (C-DA) and a customer transmission source address (C-SA) that were originally included in the S-TAG frame.
Hereinafter, the backbone destination address (B-DA), the backbone transmission source address (B-SA), and the B-TAG are also referred to as a B-TAG frame header (PBB header). Further, the backbone destination address (B-DA) is simply referred to as a destination (B-DA), and the backbone transmission source address (B-SA) is also simply referred to as a transmission source (B-SA).

As the destination (B-DA), in the case of unicast, a unique MAC address (B-MAC) in the PBB network 22 assigned to each of the edge switches 26 is used. An address is used.
As the transmission source (B-SA), a unique MAC address (B-MAC) in the PBB network 22 assigned to each of the edge switches 26 is used.
A unique VLAN identifier in the PBB network 22 is used as the B-VID (backbone VLAN ID) in the B-TAG.

  Addition (encapsulation) of PBB header, that is, destination (B-DA), transmission source (B-SA) and B-TAG to S-TAG frame, and exclusion of PBB header from B-TAG frame (decapsulation) ) Is performed only by the edge switch 26. The PBB header format of the B-TAG frame is the same as the S-TAG header format. For this reason, the core switch 28 does not need to comply with the IEEE 802.1ah standard, and can relay the B-TAG frame as long as it conforms to the IEEE 802.1Q standard.

[Redundant system of edge switch]
In the present embodiment, among the edge switches 26, the edge switch 26a and the edge switch 26b have a predetermined relationship and are paired or paired to ensure redundancy of the edge switch 26a and the edge switch 26b. . That is, the edge switch 26a and the edge switch 26b constitute a redundant system of edge switches.
Hereinafter, the edge switch 26a is also referred to as a first edge switch 26a, and the edge switch 26b is also referred to as a second edge switch 26b. The edge switch 26c is also referred to as a third edge switch 26c, and the edge switch 26d is also referred to as a fourth edge switch 26d.

FIG. 3 is a block diagram showing a schematic configuration of the first edge switch 26a. Since the configuration of the second edge switch 26b is the same as the configuration of the first edge switch 26a, FIG. 3 is also a block diagram showing a schematic configuration of the second edge switch 26b.
The first edge switch 26 a has a plurality of ports 32. The port 32 here is in charge of the physical layer at the OSI reference level, and includes, for example, the same number of physical connectors as the number of ports 32 and one interface LSI (large scale integrated circuit) to which the connectors are connected. Composed.

  Each port 32 is connected to a relay unit 34 made of, for example, a communication LSI, and receives an S-TAG frame or a B-TAG frame and inputs it to the relay unit 34 according to the connection destination. On the other hand, the port 32 transmits the S-TAG frame or the B-TAG frame output from the relay unit 34 toward the connection destination.

The relay unit 34 is connected to the memory 36, and the relay unit 34 and the memory 36 are in charge of the data link layer at the OSI reference level in cooperation with each other. That is, the relay unit 34 and the memory 36 determine the destination of the S-TAG frame or B-TAG frame input from the port 32, perform encapsulation and decapsulation according to the destination, and then perform the S-TAG. The frame or B-TAG frame is output to the port 32.
Hereinafter, the S-TAG frame and the B-TAG frame are also simply referred to as frames.

  In order to transfer the frame between the ports 32 while appropriately converting the format of the frame, the relay unit 34 has a learning unit 40, a search unit 42, an encapsulation unit 44, and a decapsulation unit when viewed in terms of functions. 46, a forwarding database (FDB) 48, a VID table 49, a my representative address 50, a mate representative address 52, and a port setting DB (database) 54 are registered in the memory 36.

[Learning Department]
The learning unit 40 learns the content of the received frame and updates the content of the FDB 48 as needed. Specifically, when the learning unit 40 receives an S-TAG frame from the PB network 20, the learning unit 40 includes a VLAN ID (S-VID) included in the S-TAG frame and a customer transmission source address (C-SA). Is read. If the read S-VID and C-SA are not registered in the FDB 48, the learning unit 40 registers the S-VID and C-SA in the FDB 48 in association with the received port 32 number.

  On the other hand, when receiving the B-TAG frame, the learning unit 40 reads the S-VID, the transmission source (B-SA), and the C-SA in the B-TAG frame. If the read S-VID, source (B-SA), and C-SA are not registered in the FDB 48, the learning unit 40 associates the received S-VID, source ( B-SA) and C-SA are registered in the FDB 48.

[FDB]
Search key information and destination information (search results) are registered in the FDB 48. That is, as shown in FIG. 4A, in the FDB 48, C-MAC, S-VID, port 32 number, and B-MAC are registered in association with each other.
In the FDB 48 of FIG. 4A, appropriate numerical values and “...” Are described for each item.
[VID table]
In the VID table 49, as shown in FIG. 4B, S-VID, I-SID (service instance ID), and B-VID are registered in association with each other. Registration of the S-VID, I-SID, and B-VID in the VID table 49 is manually performed by a network administrator, for example.
In the VID table 49 of FIG. 4B, appropriate numerical values and “...” Are described for each item.

[Search section]
The search unit 42 determines the port 32 that is the output destination of the frame input from the port 32 while referring to the FDB 48, the VID table 49, the my representative address 50, the mate representative address 52, and the port setting DB 54. Then, the frame is relayed from the port 32 to the port 32 according to the determination of the search unit 42.

Specifically, when transferring a frame from the PB network 20 to the PBB network 22, the search unit 42 associates the search key in the FDB 48 with the C-DA and S-VID of the S-TAG frame as search keys. The registered destination port number and B-MAC are searched.
On the other hand, the search unit 42 searches for the I-SID and B-VID registered in the VID table 49 in association with the search key using the S-VID of the S-TAG frame as a search key.

[Encapsulation part]
When the received S-TAG frame is transmitted to the core switch 28 or the edge switch 26 in the PBB network 22, the encapsulating unit 44 adds a PBB header and converts it into a B-TAG frame.
Specifically, the encapsulating unit 44 assigns the B-MAC, I-SID, and B-VID values searched by the searching unit 42 to the B-SA, I-SID, and B when adding the PBB header. -Store in the VID area.

[Decapsulation part]
The decapsulation unit 46 removes the PBB header and converts it to an S-TAG frame when transmitting the received B-TAG frame toward the switching hub 24 in the PB network 20.

[My representative address]
The my representative address 50 is one unique address in the PBB network 22 assigned to each of the first edge switch 26a and the second edge switch 26b. The unique address is encapsulated by the edge switch 26. Are used as the destination (B-DA) or source (B-SA) of the B-TAG frame. Therefore, for the first edge switch 26a, the unique address assigned to the first edge switch 26a is the my representative address 50, and for the second edge switch 26b, the unique address assigned to the second edge switch 26b. Is my representative address 50.
The my representative address 50 is set by an administrator of the PBB network 22, for example, and the my representative address 50 of the first edge switch 26a is different from the my representative address 50 of the second edge switch 26b.

[Mate representative address]
The mate representative address 52 is a unique address in the PBB network 22 assigned to the paired second edge switch 26b or the first edge switch 26a, and is set by an administrator of the PBB network 22, for example. For the first edge switch 26a, the unique address assigned to the second edge switch 26b is the mate representative address 52, and for the second edge switch 26b, the unique address assigned to the first edge switch 26a is the mate. This is the representative address 52.
In this specification, the word “mate” is used to mean a pair or a pair, and the first edge switch 26a is a mate of the second edge switch 26b, and the second edge switch 26b is the first edge switch. It is a mate of 26a.

[Port setting DB]
The port setting DB 54 is set by an administrator of the PBB network 22, for example. In the port setting DB 54, a rule for the relay unit 34 to relay a frame from one port 32 to another port 32 is defined for each port 32 according to the connection destination of the port 32.
Specifically, as shown in FIG. 5, the port 32 of the first edge switch 26a and the second edge switch 26b includes an access port (AC port) 32a, an interconnection port (IC port) 32i, and a network. -There is a network interface port (NNI port) 32n.

The access port 32a is used for connection with the external switch 24, the IC port 32i is used for connection between the paired first edge switch 26a and second edge switch 26b, and the NNI port 32n is a core SW It is used for connection to the core switches 28a, 28b, 28e in the network 30. The number of access ports 32a, IC ports 32i, and NNI ports 32n is not limited to one or two, and may be three or more.
Hereinafter, rules set on the ports 32 of the first edge switch 26a and the second edge switch 26b based on the port setting DB 54 will be described.

[IC port rule setting]
In the IC port 32i, a mate B-SA output filter (Mate B-SA Egress Filter) 60 is set as an output side relay rule, while a My B-SA input filter (My B) is set as an input side relay rule. -SA Ingress Filter 62 and Mate B-SA Ingress Marker 64 are set.

(1) Mate B-SA Output Filter According to the mate B-SA output filter 60, when the transmission source (B-SA) of the B-TAG frame matches the mate representative address 52, the relay unit 34 is connected to the IC port. 32i operates so as not to transmit the frame.
(2) My B-SA input filter According to the My B-SA input filter 62, when the transmission source (B-SA) of the B-TAG frame received by the IC port 32i matches the My representative address 50, The relay unit 34 operates so as to discard the frame and not relay it to another port 32.

(3) Mate B-SA input marker According to the mate B-SA input marker 64, when the transmission source (B-SA) of the B-TAG frame received by the IC port 32i matches the mate representative address 52, A marker indicating that the frame has been received by the IC port 32i, that is, has come directly from the paired second edge switch 26b, is attached.

[NNI port rule settings]
The NNI port 32n has a mate B-SA output filter (Mate B-SA Egress Filter) 70 and a My / Mate B-DA output filter (My / Mate B-DA Egress Filter) 72 as relay rules on the output side. Is set as a relay rule on the input side, a My B-SA input filter (My B-SA Ingress Filter) 74 is set.
(4) Mate B-SA Output Filter According to the mate B-SA output filter 70, when the transmission source (B-SA) of the B-TAG frame matches the mate representative address 52, the relay unit 34 is connected to the NNI port. 32n operates so as not to transmit the frame.
However, as an exception, for a frame to which a marker is given by the mate B-SA input marker 64, even if the transmission source (B-SA) matches the mate representative address 52, the relay unit 34 has the NNI port 32n in the frame. Is allowed to be sent.

(5) My / Mate B-DA Output Filter According to the My / Mate B-DA output filter 72, the destination (B-DA) of the B-TAG frame is either the My representative address 50 or the Mate representative address 52. The relay unit 34 operates so that the NNI port 32n does not transmit the frame.
However, as an exception, for the B-TAG frame encapsulated by the first edge switch 26a itself, even if the destination (B-DA) matches the mate representative address 52, the relay unit 34 determines that the NNI port 32n Allow to send.

(6) My B-SA input filter According to the My B-SA input filter 74, when the transmission source (B-SA) of the B-TAG frame received by the NNI port 32n matches the My representative address 50, The relay unit 34 operates so as to discard the frame and not relay it to another port 32.

[Access port rule settings]
A My / Mate B-DA output forwarder (My / Mate B-SA Egress Forwarder) 80 is set in the access port 32a as a relay rule on the output side.
(7) My / Mate B-DA Output Forwarder According to the My / Mate B-DA output forwarder 80, the destination (B-DA) of the B-TAG frame received by the NNI port 32n or the IC port is the My representative address 50. And the mate representative address 52, the relay unit 34 operates to transmit the frame from the access port 32a.
That is, according to the My / Mate B-DA output forwarder 80, when the destination (B-DA) of the B-TAG frame matches the My representative address 50, it is natural that the destination also matches the Mate representative address 52. The relay unit 34 processes the B-TAG frame as addressed to itself.

Then, the relay unit 34 transmits the decapsulated frame from the access port 32a even when the MAC address included in the customer destination address (C-DA) of the B-TAG frame addressed to itself is not registered in the FDB 48. Yes (flooding). At this time, if it is not received by the IC port 32i, the B-TAG frame is transmitted as it is from the IC port 32i (flooding).
At this time, since the My / Mate B-DA output filter 72 is set in the NNI port 32n, the B-TAG frame is not transmitted from the NNI port 32n.
The relay unit 34 also transmits the decapsulated frame from the access port 32a (flooding) even when the frame destination (B-DA) is a multicast address or a broadcast address.

  In FIG. 5, the mate B-SA output filter 60, the my B-SA input filter 62, and the mate B-SA input marker 64 are provided only for the IC port 32i of the first edge switch 26a due to space. Although displayed, the mate B-SA output filter 60, the my B-SA input filter 62, and the mate B-SA input marker 64 are also set for the IC port 32i of the second edge switch 26b.

  Similarly, in FIG. 5, only one NNI port 32n of the first edge switch 26a has a mate B-SA output filter 70, a my / mate B-DA output filter 72, and a my B-SA input filter 74. Although shown, the mate B-SA output filter 70, the My / Mate B-DA output filter 72, and the My B-SA input filter are connected to all the NNI ports 32n of the first edge switch 26a and the second edge switch 26b. 74 is set.

  Similarly, in FIG. 5, the My / Mate B-DA output forwarder 80 is displayed only for the access port 32a of the first edge switch 26a, but the access port 32a of the second edge switch 26b also displays the My / Mate B-DA output forwarder 80a. / Mate B-DA output forwarder 80 is set.

[Ring Management Department / Ring Management DB]
In order to utilize the redundant configuration of the first edge switch 26a and the second edge switch 26b, the switching hubs 24a and 24b of the PB network 20a are connected to each other, whereby the first edge switch 26a and the second edge switch are connected. 26b is connected directly or indirectly. That is, the first edge switch 26a, the second edge switch 26b, and the switching hubs 24a and 24b constitute a ring-type transmission path (access ring).
In this ring transmission line, a complete ring is formed by regarding the first edge switch 26a and the second edge switch 26b as one switching hub.

  In the ring type transmission line, for example, a ring protocol is employed to prevent a permanent loop of the frame. According to the ring protocol, when the ring type transmission line is normal, the access port 32a of the first edge switch 26a is set in the forwarding state and allows the frame to pass, while the access port 32a of the second edge switch 26b is in the blocking state. Is set to prohibit the passage of frames. When a failure occurs in the ring type transmission line, the access ports 32a of the first edge switch 26a and the second edge switch 26b are both set to the forwarding state and allow passage of frames.

  According to the ring protocol, in order to detect a failure such as disconnection of the transmission medium in the ring type transmission line, for example, from the access port 32a of both the first edge switch 26a and the second edge switch 26b, the PB network 20a The control frame is transmitted toward the other access port 32a so as to pass through. The first edge switch 26a and the second edge switch 26b detect the occurrence of a failure in the ring transmission line when reception of the control frame is interrupted.

In order to operate based on such a ring protocol, the first edge switch 26a and the second edge switch 26b have a ring management unit 56 and a ring management DB (database) 58, respectively. The ring management unit 56 transmits / receives a control frame and detects whether the ring transmission path is in a normal state or a failure occurrence state.
The ring management DB 58 stores the state of the ring type transmission path detected by the ring management unit 56. The ring management DB 58 also stores the port 32 set as a master port or a slave port.
The setting of the master port and the slave port is performed by, for example, the administrator of the wide area network system 10 or the ring transmission line. The setting of the master port and the slave port can be performed for each service VLAN (S-VLAN), and thus load distribution can be achieved.

  The relay unit 34 refers to the ring management DB 58 and determines whether or not the access port 32a can transmit and receive a frame. That is, in the first edge switch 26a, the access port 32a set as the master port is always set to the forwarding state regardless of the occurrence of the failure. On the other hand, in the second edge switch 26b, the access port 32a set as the slave port is set to the blocking state when it is normal, and is set to the forwarding state when a failure occurs.

When the access port 32a is set to the blocking state by the ring protocol, transmission / reception of a frame other than the control frame (user frame) on the access port 32a is prohibited regardless of the registered content of the FDB 48 and the setting of the port setting DB 54.
In FIG. 1, the access port 32a of the second edge switch 26b in the blocking state is represented by a black square.

[Operation]
Hereinafter, the operation of the above-described wide area network system 10 will be described for a plurality of cases with the first edge switch 26a and the second edge switch 26b as the center.
In any case, as shown in FIG. 6, a failure occurs in the transmission path between the switching hub 24a and the switching hub 24b, and the access port 32a of the second edge switch 26b It is assumed that a transition is made from the blocking state to the forwarding state.

Usually, the MAC address has a 6-byte address, but in order to simplify the explanation, it is assumed that the B-MAC, that is, the my representative address 50 of the first edge switch 26a is "A", and the second edge switch 26b. B-MAC, that is, my representative address 50 is “B”. In this case, the mate representative address 52 of the first edge switch 26a is “B”, and the mate representative address 52 of the second edge switch 26b is “A”.
The B-MACs of the third edge switch 26c and the fourth edge switch 26d are “X” and “Y”, respectively. The multicast address is assumed to be “MC”. Furthermore, it is assumed that the MAC addresses of the user terminals 16a, 16b, 16c, and 16d are “a”, “b”, “c”, and “d”, respectively.

[First case / B-SA: B, B-DA: X]
As a first case, a case will be described in which the user terminal 16b in the user network 12b transmits a frame toward the user terminal 16c in the user network 12c. In the first case, as shown in FIG. 5, the access port 32a of the second edge switch 26b receives the frame. The second edge switch 26b encapsulates the received frame and transmits the frame from one of the IC port 32i and the NNI port n.

Here, the propagation path of the frame differs depending on the registered contents of the FDB 48 of the second edge switch 26b and, in some cases, the registered contents of the FDB 48 of the first edge switch 26a.
When “c”, which is a customer destination address (C-DA), has already been registered in the FDB 48 of the second edge switch 26b, a pattern in which one of the NNI ports 32n is registered corresponding to “c”. (First pattern) and a pattern (second pattern) in which the IC port 32i is registered. In the latter second pattern, one of the NNI ports 32n is registered in the FDB 48 of the first edge switch 26a corresponding to “c”.

<First pattern>
Hereinafter, the first pattern will be described with reference to FIG.
In the FDB 48 of the second edge switch 26b, when one of the NNI ports 32n is registered in association with “c” of the customer destination address (C-DA), the NNI port 32n is selected as a frame output destination candidate. Selected. For this reason, the frame is processed by the mate B-SA output filter 70 and the my / mate B-DA output filter 72, which are output side relay rules set in the NNI port 32n.

  In this case, “B” as the frame transmission source (B-SA) does not match the mate representative address 52 of the second edge switch 26b, and the destination (B-DA) is the my representative of the second edge switch 26b. It does not match either the address 50 or the mate representative address 52. For this reason, the frame is permitted to pass through the mate B-SA output filter 70 and the my / mate B-DA output filter 72, and is transmitted from the NNI port 32n of the second edge switch 26b.

  Then, the frame transmitted from the NNI port 32n is transferred based on the destination (B-DA) in the core SW network 30 constituted by the core switch 28b and the like, and reaches the third edge switch 26c. Then, the frame is decapsulated by the third edge switch 26c, and then reaches the user terminal 16c of the user network 12c via the PB network 20c.

<Second pattern>
Hereinafter, the second pattern will be described with reference to FIG.
In the FDB 48 of the second edge switch 26b, when the IC port 32i is registered in association with “c” of the customer destination address (C-DA), the IC port 32i is selected as a frame output destination candidate. For this reason, the frame becomes a processing target of the mate B-SA output filter 60 that is the relay rule on the output side set in the IC port 32i.

In this case, “B” as the frame transmission source (B-SA) does not match the mate representative address 52 of the second edge switch 26b. For this reason, the frame is permitted to pass through the mate B-SA output filter 60 and transmitted from the IC port 32i of the second edge switch 26b.
The frame transmitted from the IC port 32i is received by the IC port 32i of the first edge switch 26a, and is a relay rule on the input side of the first edge switch 26a, which is the My B-SA input filter 62 and the mate B- It becomes a processing target of the SA input marker 64.

  In this case, the frame transmission source (B-SA) is the address “B” of the second edge switch 26 b and does not match the my representative address 50 of the first edge switch 26 a but matches the mate representative address 52. . For this reason, the frame is allowed to pass through the My B-SA input filter 62 and is marked by the mate B-SA input marker 64.

In the FDB 48 of the first edge switch 26a, when the NNI port 32n is registered in association with the destination (B-DA) “X”, the frame permitted to pass through the My B-SA input filter 62 is It becomes a processing target of the mate B-SA output filter 70 and the my / mate B-DA output filter 72, which are relay rules on the output side of the first edge switch 26a set in the NNI port 32n.
In this case, “B” as the frame transmission source (B-SA) matches the mate representative address 52, but since the marker is attached to the frame, the frame passes through the mate B-SA output filter 70. Allowed. On the other hand, since the destination (B-DA) of the frame is “X” and does not match either the my representative address 50 or the mate representative address 52, the frame is allowed to pass through the My / Mate B-DA output filter 72. Is done.

  Therefore, the frame is transmitted from the NNI port 32n of the first edge switch 26a. The frame transmitted from the NNI port 32n is transferred based on the destination (B-DA) in the core SW network 30 constituted by the core switch 28a and the like, and finally reaches the user terminal 16c of the user network 12c. To do.

Further, a frame permitted to pass through the My B-SA input filter 62 is a relay rule on the output side of the first edge switch 26a set in the access port 32a, which is the My / Mate B-DA output forwarder 80. It becomes a processing target.
In this case, the destination (B-DA) of the frame is “X” and does not match either the my representative address 50 or the mate representative address 52 of the first edge switch 26a. There is no transmission from the access port 32a.

[Second case / B-SA: B, B-DA: MC]
Hereinafter, the second case will be described. The second case is the same as the first case in that the user terminal 16b in the user network 12b transmits a frame toward the user terminal 16c in the user network 12c, but the customer destination address (C-DA ) “C” is not registered in the FDB 48 of the second edge switch 26b and the FDB 48 of the first edge switch 26a. As described above, the case where “c” of the customer destination address (C-DA) is not registered may be the case where the FDB 48 is flushed (erased) for some reason other than when the user terminal 16 c is new.

  In the second case, it is conceivable that a frame transmitted from the second edge switch 26b toward the core switch 28b returns to the first edge switch 26a and is received. Therefore, hereinafter, the operation during transmission will be described first with reference to FIG. 8, and then the operation during reception will be described with reference to FIG.

<When sending>
In this case, as shown in FIG. 8, in the second edge switch 26b, when the frame is encapsulated, the multicast address “MC” is set as the destination (B-DA), and the received access port 32a and All the ports 32 to which the same B-VID is assigned are set as frame output destination candidates.

Therefore, the frame is processed by the mate B-SA output filter 70 and the my / mate B-DA output filter 72, which are relay rules on the output side of the second edge switch 26b set in the NNI port 32n. Become.
In this case, “B” as the frame transmission source (B-SA) does not match the mate representative address 52, and the frame is permitted to pass through the mate B-SA output filter 70. On the other hand, since the frame destination (B-DA) is “MC” and does not coincide with either the my representative address 50 or the mate representative address 52, the frame is allowed to pass through the My / Mate B-DA output filter 72. Is done. Therefore, the frame is transmitted from the NNI port 32n of the second edge switch 26b.

  Further, since the frame destination (B-DA) is “MC”, the frame becomes a processing target of the mate B-SA output filter 60 which is the relay rule on the output side set in the IC port 32i. In this case, “B” as the frame transmission source (B-SA) does not match the mate representative address 52 of the second edge switch 26b. For this reason, the frame is permitted to pass through the mate B-SA output filter 60 and transmitted from the IC port 32i of the second edge switch 26b.

  Furthermore, although not shown, if there is another access port 32a to which the same B-VID as the received access port 32a is assigned, the frame is also transferred from the access port 32a to the PB network 20a without being encapsulated. Sent to.

The frame transmitted from the IC port 32i of the second edge switch 26b is received by the IC port 32i of the first edge switch 26a, and is a relay rule on the input side of the first edge switch 26a. And the mate B-SA input marker 64.
In this case, the frame transmission source (B-SA) is the address “B” of the second edge switch 26 b and does not match the my representative address 50 of the first edge switch 26 a but matches the mate representative address 52. To do. For this reason, the frame is allowed to pass through the My B-SA input filter 62 and is marked by the mate B-SA input marker 64.

  The frame permitted to pass through the My B-SA input filter 62 is decapsulated and transmitted from the access port 32a of the first edge switch 26a because the frame destination (B-DA) is “MC”. The

A frame permitted to pass through the My B-SA input filter 62 is a mate B-SA output filter 70 that is a relay rule on the output side of the first edge switch 26a set in the NNI port 32n. It becomes the processing target of the My / Mate B-DA output filter 72.
In this case, “B” as the frame transmission source (B-SA) matches the mate representative address 52, but since the marker is attached to the frame, the frame passes through the mate B-SA output filter 70. Allowed. On the other hand, since the frame destination (B-DA) is “MC” and does not coincide with either the my representative address 50 or the mate representative address 52, the frame is allowed to pass through the My / Mate B-DA output filter 72. Is done. Therefore, the frame is transmitted from the NNI port 32n of the second edge switch 26b.

  Thus, the frame transmitted from the NNI port 32n of the first edge switch 26a and the second edge switch 26b is transferred based on the destination (B-DA) in the core SW network 30 constituted by the core switches 28a, 28b and the like. Is done. Since the frame destination (B-DA) is “MC”, the frames are transmitted from all the ports to which the same B-VID is assigned also in the core switches 28 a and 28 b in the core SW network 30.

  As a result, any frame transmitted from the NNI port 32n of the first edge switch 26a and the second edge switch 26b reaches the third edge switch 26c and is decapsulated by the third edge switch 26c. The user terminal 16c of the user network 12c is reached via the PB network 20c.

<When receiving>
Hereinafter, as an example in which a frame transmitted from the NNI port 32n of the second edge switch 26b is received by the first edge switch 26a, the frame passes through the core switch 28a of the core SW network 30 to the first edge switch 26a. The case where it has been reached will be described with reference to FIG.
The frame received by the NNI port 32n of the first edge switch 26a becomes a processing target of the My B-SA input filter 74, which is a relay rule on the input side.

  In this case, the transmission source (B-SA) of the frame is the address “B” of the second edge switch 26b and does not match the my representative address 50 of the first edge switch 26a. For this reason, the frame is allowed to pass through the My B-SA input filter 74. Since the frame destination (B-DA) is “MC”, the NNI port 32n, the access port 32a, and the IC port 32i other than the received one are set as output destination candidates.

  Therefore, the frame becomes a processing target of the mate B-SA output filter 70 and the my / mate B-DA output filter 72, which are output side relay rules set in the other NNI port 32n. In this case, since the frame transmission source (B-SA) is “B” and matches the mate representative address 52, the frame is prohibited from passing through the mate B-SA output filter 70. For this reason, the frame is not sent back toward the core SW network 30.

  Further, the frame is a processing target of the mate B-SA output filter 60 that is the relay rule on the output side set in the IC port 32i. In this case, “B” that is the frame transmission source (B-SA) matches the mate representative address 52 of the first edge switch 26 a, and therefore the frame is prohibited from passing through the mate B-SA output filter 60. . For this reason, the frame is not transmitted toward the second edge switch 26b through the IC port 32i.

  For the access port 32a, the My / Mate B-DA output forwarder 80 is set, but the filter is not set. For this reason, the frame whose destination (B-DA) is “MC” is decapsulated and then transmitted toward the PB network 20a.

[Third case / B-SA: B, B-DA: A]
As a third case, the user terminal 16b in the user network 12b transmits a frame toward the user terminal 16a in the user network 12a, and the customer destination address (C-DA) is the second edge switch 26b. A case where it is registered in the FDB 48 will be described.
The third case is different from the first case in the frame destination (B-DA), but is shown in FIGS. 10 and 11 according to the registered contents of the FDB 48 of the second edge switch 26b. As described above, the frame is transmitted from the NNI port 32n or the IC port 32i.

  This means that even if the frame is processed by the mate B-SA output filter 70 and the my / mate B-DA output filter 72 set in the NNI port 32n, the mate B-SA output filter 70 does not pass through. Of course, the frame encapsulated by itself is an exception, even if it matches “A” whose destination (B-DA) is the mate representative address 52, with the exception of My / Mate B-DA. This is because passage of the output filter 72 is permitted.

  As shown in FIG. 10, when the NNI port 32n of the first edge switch 26a receives the frame, the frame passes through the My B-SA input filter 74 set in the NNI port 32n. The frame is transmitted from the access port 32a by the My / Mate B-DA output forwarder 80 set in the access port 32a, and reaches the user terminal 16a via the switching hub 24a.

On the other hand, as shown in FIG. 11, when a frame is transmitted from the IC port 32i of the second edge switch 26b, the frame is the My B-SA input filter 62 set in the IC port 32i of the first edge switch 26a. Is permitted, and a marker is given by the mate B-SA input marker 64.
The frame is transmitted from the access port 32a by the My / Mate B-DA output forwarder 80 set in the access port 32a, and reaches the user terminal 16a via the switching hub 24a.

[Fourth case / B-SA: X, B-DA: A]
As a fourth case, referring to FIG. 12, the user terminal 16c in the user network 12c transmits a frame to the user terminal 16b in the user network 12b, and the NNI port 32n of the first edge switch 26a is a frame. Will be described.
Even in the fourth case, a failure has occurred between the switching hub 24a and the switching hub 24b constituting the access ring. However, the edge switch 26c cannot detect the occurrence of the failure and before the failure occurs. Similarly to the above, it is assumed that the frame is encapsulated using “A” as the destination (B-DA).

  In the fourth case, the frame received by the NNI port 32n of the first edge switch 26a is “X” whose transmission source (B-SA) is different from the my representative address 50. Therefore, the My B-SA input filter 74 is allowed to pass. The frame is transmitted from the access port 32 a by the My / Mate B-DA output forwarder 80 because the destination (B-DA) matches “A” having the My representative address 50.

Here, a failure occurs between the switching hub 24a and the switching hub 24b constituting the access ring, and the contents of the FDB 48 of the first edge switch 26a are once erased when the failure occurs. For this reason, the customer destination address (C-DA) “b” is not registered in the FDB 48 of the first edge switch 26a.
Therefore, the frame becomes a flooding target, and the IC port 32i of the first edge switch 26a is also selected as a candidate for the output destination of the frame. Since the frame transmission source (B-SA) is “X”, the frame is allowed to pass through the mate B-SA output filter 60 set in the IC port 32i. Therefore, the frame is transmitted from the IC port 32i of the first edge switch 26a toward the second edge switch 26b.

  The frame received by the IC port 32i of the second edge switch 26b is allowed to pass through the My B-SA input filter 62 set in the IC port 32i, and the marker is added by the mate B-SA input marker 64. Absent. Since this frame matches “A” whose destination (B-DA) is the mate representative address 52, the access of the second edge switch 26b is performed by the My / Mate B-DA output forwarder 80 set in the access port 32a. Sent from port 32a. Then, the frame reaches the user terminal 16b to which the customer destination address (C-DA) “b” is assigned via the switching hub 24b.

[Fifth Case / B-SA: X, B-DA: B]
As a fifth case, referring to FIG. 13, the user terminal 16c in the user network 12c transmits a frame to the user terminal 16b in the user network 12b, and the NNI port 32n of the first edge switch 26a is a frame. Will be described.
In the fifth case, after the failure between the switching hub 24a and the switching hub 24b constituting the access ring, the edge switch 26c has learned “b” which is the MAC address of the user terminal 16b. Assume that the frame is encapsulated using “B” as the destination (B-DA).

  In the fifth case, since the frame received by the NNI port 32n of the first edge switch 26a is “X” whose transmission source (B-SA) is different from the my representative address 50, the My B-SA input filter 74 is allowed to pass.

  Here, since the edge switch 26c has learned “b” which is the MAC address of the user terminal 16b after the failure occurs, the first edge switch 26a also has “b” which is the MAC address of the user terminal 16b after the failure occurs. Is learning. Therefore, according to the FDB 52, the IC port 32i of the first edge switch 26a is selected as a frame output destination candidate. Since the frame transmission source (B-SA) is “X”, the frame is allowed to pass through the mate B-SA output filter 60 set in the IC port 32i. Therefore, the frame is transmitted from the IC port 32i of the first edge switch 26a toward the second edge switch 26b.

  The frame received by the IC port 32i of the second edge switch 26b is allowed to pass through the My B-SA input filter 62 set in the IC port 32i, and the marker is added by the mate B-SA input marker 64. Absent. Since this frame matches “B” whose destination (B-DA) is the my representative address 50, the access of the second edge switch 26b is performed by the My / Mate B-DA output forwarder 80 set in the access port 32a. Sent from port 32a. Then, the frame reaches the user terminal 16b to which the customer destination address (C-DA) “b” is assigned via the switching hub 24b.

[Sixth case / B-SA: Y, B-DA: X]
As a sixth case, referring to FIG. 14, the user terminal 16d in the user network 12d transmits a frame to the user terminal 16c in the user network 12c, and this frame is transmitted to the NNI port of the first edge switch 26a. The case where 32n received is demonstrated.

  In the sixth case, the frame received by the NNI port 32n of the first edge switch 26a is “Y” whose transmission source (B-SA) is different from the my representative address 50. Therefore, the My B-SA input filter 74 is allowed to pass. The frame is transmitted from the access port 32a of the first edge switch 26a because the destination (B-DA) does not match either “A”, which is the my representative address 50, or “B”, which is the mate representative address 52. There is nothing.

  As a candidate for the output destination of the frame received by the NNI port 32n, for example, another NNI port 32n is selected according to the registered contents of the FDB 48. The frame is permitted to pass through the mate B-SA output filter 70 and the my / mate B-DA output filter 72 set in the other NNI port 32n. Therefore, the frame is sent back from the NNI port 32n of the first edge switch 26a toward the core SW network 30.

  Note that the IC port 32i can also be selected as a frame output destination candidate in the first edge switch 26a in accordance with the registered contents of the FDB 48. In this case, although not shown, the frame is permitted to pass through the mate B-SA output filter 60 set in the IC port 32i, and is transmitted toward the second edge switch 26b.

  The frame received by the IC port 32i of the second edge switch 26b is allowed to pass through the My B-SA input filter 62 set in the IC port 32i, and the mate B-SA input marker 64 gives the marker. Not done. This frame is transmitted from the access port 32a of the second edge switch 26b because the destination (B-DA) does not match either “B” that is the my representative address 50 or “A” that is the mate representative address 52. Never happen.

Also in the second edge switch 26b, the NNI port 32n is selected as a candidate for the output destination of the frame received by the IC port 32i. As in the case of the first edge switch 26a, the frame is permitted to pass through the mate B-SA output filter 70 and the my / mate B-DA output filter 72 set in the NNI port 32n. Therefore, the frame is transmitted from the NNI port 32n of the second edge switch 26b toward the core SW network 30.
[Seventh case / B-SA: X, B-DA: MC]
The seventh case will be described with reference to FIG. In the seventh case, the frame transmitted from the user terminal 16c in the user network 12c to the user terminal 16b in the user network 12b is sent to the third edge switch 26c as a destination (B-DA), which is a multicast address “MC”. ”And the encapsulated frame is received by the NNI port 32n of the first edge switch 26a.

  In the seventh case, the frame received by the NNI port 32n of the first edge switch 26a is “X” whose transmission source (B-SA) is different from the my representative address 50. Therefore, the My B-SA input filter 74 is allowed to pass. Since the destination (B-DA) is “MC” representing multicast, the frame is transmitted from the access port 32a of the first edge switch 26a.

  Further, since the destination (B-DA) is MC as a candidate for the output destination of the frame received by the NNI port 32n, another NNI port 32n is selected. The frame is permitted to pass through the mate B-SA output filter 70 and the my / mate B-DA output filter 72 set in the other NNI port 32n. Therefore, the frame is sent back from the NNI port 32n of the first edge switch 26a toward the core SW network 30.

  On the other hand, since the destination (B-DA) is MC as a frame output destination candidate, the IC port 32i of the first edge switch 26a is also selected. The frame is permitted to pass through the mate B-SA output filter 60 set in the IC port 32i, and is transmitted toward the second edge switch 26b.

  The frame received by the IC port 32i of the second edge switch 26b is allowed to pass through the My B-SA input filter 62 set in the IC port 32i, and the marker is added by the mate B-SA input marker 64. Absent. Since the destination (B-DA) is “MC” representing multicast, this frame is transmitted from the access port 32a of the second edge switch 26b.

  Also in the second edge switch 26b, the NNI port 32n is selected as a candidate for the output destination of the frame received by the IC port 32i. As in the case of the first edge switch 26a, the frame is permitted to pass through the mate B-SA output filter 70 and the my / mate B-DA output filter 72 set in the NNI port 32n. Therefore, the frame is transmitted from the NNI port 32n of the second edge switch 26b toward the core SW network 30.

  As described above, the flow of the frame has been described for the representative first to seventh cases, but there are other cases. Therefore, Table 1 shows the mate B-SA output filters 60 and 70 and the My B-SA input of each of the first edge switch 26a and the second edge switch 26b in the first to seventh cases and these types of cases. The handling of frames by the filters 62 and 74, the mate B-SA input marker 64, the My / Mate B-DA output filter 72, and the My / Mate B-DA output forwarder 80, the transmission source (B-SA) and the destination ( (B-DA).

  In the wide area network system 10 of the above-described embodiment, the wide area network 14 adopts the MAC-in-MAC PBB network 22, and the core switch 28 in the PBB network 22 is connected to the user terminal 16 of the user network 12. There is no need to learn MAC addresses.

In addition, in the wide area network system 10 of the above-described embodiment, the user networks 12a and 12b are connected to two edge switches, that is, the first edge switch 26a and the second edge switch 26b, respectively. Is secured.
That is, even if one of the first edge switch 26a and the second edge switch 26b, for example, the first edge switch 26a fails, if the other second edge switch 26b is operating normally, the user network 12a, Frames transmitted from the user terminals 16a and 16b of the 12b toward the user terminals 16c and 16d of the other user networks 12c and 12d are transferred through the second edge switch 26b.
Thus, in the wide area network system 10, the redundancy of the edge switch is ensured and the stability of communication is ensured.

  In particular, in the wide area network system 10 of the embodiment described above, the my representative address 50 and the mate representative address 52 are set in the first edge switch 26a and the second edge switch 26b, and the mate B-SA output filter 60, My B-SA input filter 62, Mate B-SA input marker 64, Mate B-SA output filter 70, My / Mate B-DA output filter 72, My B-SA input filter 74, and My / Mate B-DA By setting the output forwarder 80, the redundancy of the edge switch is secured with a simple configuration, and the stability of communication is secured.

  In the wide area network system 10 of the above-described embodiment, the first edge switch 26a, the second edge switch 26b, and the switching hubs 24a and 24b form a ring-type transmission line, so that the redundancy of the PB network 20a is achieved. And stability of communication is ensured.

That is, for example, as shown in FIG. 6, even if a failure occurs in the transmission path between the switching hub 24a and the switching hub 24b, or the transmission path between the switching hub 24a and the first edge switch 26a. Even if a failure occurs or a failure occurs in the transmission path between the switching hub 24b and the second edge switch 26b, the user terminals 16a and 16b of the user networks 12a and 12b are not connected to the other user networks 12c. , 12d can communicate with the user terminals 16c, 16d.
It should be noted that the failure in the ring type transmission line includes a failure of a part or the whole of each of the first edge switch 26a and the second edge switch 26b.

On the other hand, in the wide area network system 10 according to the above-described embodiment, problems caused by the occurrence of a failure in the ring transmission path (access ring) are solved.
Specifically, FIG. 16 shows a state of frame transfer from the user terminal 16c to the user terminal 16b when the access ring is in a normal state. If the access ring is in a normal state, a frame addressed to the user terminal 16b from the user terminal 16c sequentially passes through the first edge switch 26a, the switching hub 24a, and the switching hub 24b.
When a failure occurs in the access ring, the access port of the second edge switch 26b is set to the forwarding state by the ring protocol, and the connection destination (accommodation destination) of the user network 12b is changed from the first edge switch 26a to the second edge switch 26b. Is done.
However, the other edge switches 26c and 26d cannot detect the change of the connection destination, and after the change of the connection destination, “A” which is the connection destination before the change is used as the destination (B-DA) as the B-TAG frame. Will be sent for a certain period. In this case, in the prior art, the frame may be discarded in the first edge switch that is the connection destination before the change, and may not reach the destination user terminal 16b.

Therefore, in the present embodiment, as described in the fourth case with reference to FIG. 12, the first edge switch 26a transfers the B-TAG frame addressed to itself to the second edge switch 26b through the IC port 32i. The second edge switch 26b transmits the received B-TAG frame addressed to the mate from the access port 32a toward the switching hub 24b. Accordingly, even immediately after the connection destination of the user network 12b is changed, the user terminal 16b that is the destination is reliably reached without being discarded.
Thus, in the wide area network system 10 of one embodiment mentioned above, the trouble when a connection destination is changed is eliminated with a simple configuration.

  The fixed period is until the other terminal switches 26c and 26d re-learn if the user terminal 16b and the user terminals 16c and 16d communicate in both directions, and other if the communication is performed in one direction. This is until the FDB registration in the edge switches 26c and 26d is deleted as the aging time elapses.

The present invention can be implemented with various modifications without being limited to the above-described embodiments.
For example, in the above-described embodiment, the ring protocol used in the ring-type transmission line is not particularly limited. The STP (Spanning Tree Protocol) method defined in IEEE 802.1D, ITU-T G. The method of Ethernet Ring Protection defined in 8032 may be adopted.

  In the above-described embodiment, PBB defined in IEEE802.1ah is adopted as the MAC-in-MAC method, but EoE (Ethernet over Ethernet (Ethernet is a registered trademark)) is adopted. Also good.

  Finally, the specific configuration and processing procedure of the apparatus used in the above-described embodiment are all preferable and of course not limited thereto.

10 wide area network system 12 user network 14 wide area network 16 user terminal 20 PB network (second network)
22 PBB network (first network)
24 Switching hub (external relay device)
26 Edge Switch (Edge Relay Device)
26a First edge switch (first edge relay device)
26b Second edge switch (second edge relay device)
28 Core switch (core relay device)
30 Core SW network 19 Transmission medium 32a Access port 32i IC port 32n NNI port

Claims (8)

In the edge relay device arranged at the boundary of the MAC-in-MAC network,
An NNI port provided for connection with a core relay device disposed inside the boundary of the network;
An access port provided for connection with an external relay device arranged outside the network;
An IC port provided for connection with another edge relay device arranged at the boundary;
A relay unit that controls the flow of frames between the core relay device, the external relay device, and the other edge relay device through the NNI port, the access port, and the IC port;
When a unique MAC address in the network assigned to itself is my representative address, and a unique MAC address in the network assigned to the other edge relay device is a mate representative address,
The relay unit controls the flow of the frame based on a combination of the my representative address and the mate representative address included in the destination and source of the frame.
Edge relay device. The edge relay device according to claim 1,
The relay unit is
Of the frames, the frame including the my representative address in the transmission source is stopped when received by each of the NNI port and the IC port,
Of the frames, the frame including the mate representative address in the transmission source is stopped so as not to be transmitted from the IC port.
Among the frames, a frame including one of the my representative address and the mate representative address as the destination is not transmitted from the NNI port except for a frame received by the access port. Stop,
Among the frames, the frame including the mate representative address in the transmission source is stopped so as not to be transmitted from the NNI port except for the frame received by the IC port.
Among the frames, for the frame where the destination includes the My representative address and the Mate representative address, a flow is generated so as to be transmitted from the access port.
Edge relay device. A redundant system of edge relay devices arranged at the boundary of a MAC-in-MAC network,
A first edge relay device and a second edge relay device;
Each of the first edge relay device and the second edge relay device is
An NNI port connected to a core relay device arranged inside the boundary of the network;
An access port connected to an external relay device arranged outside the network;
An IC port used for connection between the first edge relay device and the second edge relay device;
Relay unit for controlling the flow of frames among the core relay device, the external relay device, the first edge relay device, and the second edge relay device through the NNI port, the access port, and the IC port And
A unique MAC address in the network assigned to itself is used as a my representative address, and for the first edge relay device, a unique MAC address in the network assigned to the second edge relay device is used as a mate representative address. For an edge relay device, when a unique MAC address in the network assigned to the first edge relay device is a mate representative address,
The relay unit controls the flow of the frame based on a combination of the my representative address and the mate representative address included in the destination and source of the frame.
Redundant system for edge relay equipment. In the redundant system of the edge relay device according to claim 3,
The relay unit is
Of the frames, the frame including the my representative address in the transmission source is stopped when received by each of the NNI port and the IC port,
Of the frames, the frame including the mate representative address in the transmission source is stopped so as not to be transmitted from the IC port.
Among the frames, a frame including one of the my representative address and the mate representative address as the destination is not transmitted from the NNI port except for a frame received by the access port. Stop,
Among the frames, the frame including the mate representative address in the transmission source is stopped so as not to be transmitted from the NNI port except for the frame received by the IC port.
Among the frames, for the frame where the destination includes the My representative address and the Mate representative address, a flow is generated so as to be transmitted from the access port.
Redundant system for edge relay equipment. In the redundant system of the edge relay device according to claim 3 or 4,
The first edge relay device and the second edge relay device together with the external relay device constitute a ring type transmission line.
Redundant system for edge relay equipment. A MAC-in-MAC first network including the edge relay device redundancy system according to any one of claims 3 to 5;
A wide area network system comprising a plurality of second networks connected to the first network and connected to each other via the first network. In the frame transfer method for the edge relay device arranged at the boundary of the MAC-in-MAC network,
The NNI port of the edge relay device is connected to a core relay device arranged on the inner side of the network boundary,
Connecting the access port of the edge relay device with an external relay device arranged outside the network;
The IC port of the edge relay device is connected to another edge relay device arranged at the boundary,
A unique MAC address in the network is assigned to each of the edge relay device and the other edge relay device,
When the MAC address assigned to the edge relay device is my representative address and the MAC address assigned to the other edge relay device is a mate representative address,
Based on the combination of the my representative address and the mate representative address included in the destination and source of the frame, the core relay device, the external relay device, and the core through the NNI port, the access port, and the IC port Controlling the flow of the frame between other edge relay devices;
Frame transfer method for edge relay device. The frame transfer method for an edge relay device according to claim 7,
Of the frames, the frame including the my representative address in the transmission source is stopped when received by each of the NNI port and the IC port,
Of the frames, the frame including the mate representative address in the transmission source is stopped so as not to be transmitted from the IC port.
Among the frames, a frame including one of the my representative address and the mate representative address as the destination is not transmitted from the NNI port except for a frame received by the access port. Stop,
Among the frames, the frame including the mate representative address in the transmission source is stopped so as not to be transmitted from the NNI port except for the frame received by the IC port.
Among the frames, for the frame where the destination includes the My representative address and the Mate representative address, a flow is generated so as to be transmitted from the access port.
Frame transfer method for edge relay device.

Images