Communication system, communication device, and communication control method

Abstract

To achieve MC-LAG without being based on MAC learning information.SOLUTION: A communication system is interposed between a first opposed device and a second opposed device and achieves communication between the first opposed device and the second opposed device. The communication system includes a first communication device and a second communication device that build multi-chassis link aggregation with each of the first opposed device and the second opposed device. The first communication device includes an inter-chassis link port connected to the second communication device, performs alive monitoring on a first path from the inter-chassis link port to the first opposed device, and when a failure is detected on the first path, transmits a frame from the inter-chassis link port to the first opposed device.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present disclosure relates to a communication system, a communication device, and a communication control method.

Background Art

[0002] Link aggregation that constructs a logical single link by aggregating a plurality of communication paths (physical links) is known. For example, IEEE (Institute of Electrical and Electronics Engineers) 802.1AX-2020 (Non-Patent Document 1) defines a Multi-Chassis Link Aggregation (MC-LAG) function that treats a plurality of physical ports of two communication devices (for example, switches) as a single logical port. MC-LAG is used, for example, to expand the communication bandwidth or ensure link redundancy.

[0003] Two communication devices that construct MC-LAG (hereinafter also referred to as parallel devices) are generally connected by a physical link called an Inter-Chassis Link (ICL). ICL is used as a backup path when a failure occurs in the multi-chassis link. When a failure occurs in the physical link between one of the two parallel devices and the opposing device, frames are transferred between the parallel device and the opposing device via the ICL and the physical link between the other parallel device and the opposing device.

[0004] Normally, the frame transfer path is determined based on the destination MAC (Media Access Control) address of the frame and MAC learning information (Non-Patent Document 2).

Prior Art Documents

Non-Patent Documents

[0005]

Non-Patent Document 1

[0006] To implement MC-LAG based on MAC learning information, it is necessary to continuously synchronize and share MAC learning information between two parallel devices. Therefore, each parallel device requires a large-capacity table to store MAC learning information. Furthermore, synchronization processing of learning information between the two parallel devices (hereinafter referred to as "learning synchronization") is required, and learning synchronization tends to be complex. If proper learning synchronization is not performed, a frame that has been learned by one parallel device but not by the other may be forwarded to all ports by the other parallel device. Such forwarding of frames to all ports (flooding) includes unnecessary communication to destinations other than the intended destination, which can strain the communication bandwidth created by MC-LAG.

[0007] The purpose of this disclosure is to solve the problems described above, and its objective is to realize MC-LAG without relying on MAC learning information. [Means for solving the problem]

[0008] The communication system of this disclosure intervenes between a first opposing device and a second opposing device to enable communication between the first opposing device and the second opposing device. The communication system comprises a first communication device and a second communication device. The first communication device and the second communication device establish a multi-chassis link aggregation with each of the first opposing device and the second opposing device. The first communication device includes an inter-chassis link port connected to the second communication device. The first communication device performs liveness monitoring on a first path from the inter-chassis link port to the first opposing device. If the first communication device detects a fault in the first path, it transmits a frame from the inter-chassis link port to the first opposing device.

[0009] The communication device of this disclosure intervenes between a first opposing device and a second opposing device to enable communication between the first opposing device and the second opposing device. The communication device, together with other communication devices, can construct a multi-chassis link aggregation with each of the first opposing device and the second opposing device. The communication device comprises an inter-chassis link port, a reachability determination unit, and a transfer control unit. The inter-chassis link port is a port for communicating with other communication devices. The reachability determination unit determines the reachability of the path from the inter-chassis link port to the first opposing device. The transfer control unit determines the destination of the frame from the inter-chassis link port. If the reachability determination unit detects a fault in the path, the transfer control unit transmits the frame to the first opposing device.

[0010] The communication control method of this disclosure is a communication control method performed in a communication system that intervenes between a first opposing device and a second opposing device to realize communication between the first opposing device and the second opposing device. The communication system comprises a first communication device and a second communication device. The first communication device and the second communication device establish a multi-chassis link aggregation with each of the first opposing device and the second opposing device. The first communication device includes an inter-chassis link port connected to the second communication device. The communication control method includes the steps of the first communication device performing a liveness check on the path from the inter-chassis link port to the first opposing device, and, if the first communication device detects a fault in the path, the first communication device transmits a frame from the inter-chassis link port to the first opposing device. [Effects of the Invention]

[0011] According to this disclosure, MC-LAG can be implemented without relying on MAC learning information. [Brief explanation of the drawing]

[0012] [Figure 1] Figure 1 is a diagram showing an example of the configuration of a communication system according to Embodiment 1. [Figure 2] Figure 2 shows the format of the ETH-CC frame. [Figure 3] Figure 3 is a functional block diagram that schematically shows the configuration of each component of the communication device shown in Figure 1. [Figure 4] Figure 4 shows the frame transfer path when a failure occurs in the link between the port of the communication device and the port of the opposing device shown in Figure 1. [Figure 5] Figure 5 shows the frame transfer path when a failure occurs in the link between the port of the communication device and the port of the opposing device shown in Figure 1. [Figure 6] Figure 6 shows the frame transfer path when a failure occurs in the link between the port of the communication device and the port of the opposing device shown in Figure 1. [Figure 7]FIG. 7 is a diagram showing a frame transfer path when a failure occurs in the link between the port of the communication device in FIG. 1 and the port of the counterpart device. [Figure 8] FIG. 8 is a flowchart showing the flow of MEP setting performed in each of the counterpart device and the communication device in FIG. 1. [Figure 9] FIG. 9 is a flowchart showing the flow of the alive / dead monitoring process performed by the reachability determination unit in FIG. 3. [Figure 10] FIG. 10 is a flowchart showing the flow of the process of determining the transfer destination of a frame received from the ICL port, which is performed by the switch control unit in FIG. 3. [Figure 11] FIG. 11 is a flowchart showing the flow of the process of determining the transfer destination of a frame received from any of the counterpart devices, which is performed by the switch control unit in FIG. 3. [Figure 12] FIG. 12 is a flowchart showing the flow of the process of determining the transfer destination of a frame received from a counterpart device that is not the target of alive / dead monitoring, which is performed by the switch control unit of the communication system according to Embodiment 2. [Figure 13] FIG. 13 is a diagram showing a configuration example of a communication system according to Embodiment 3. [Figure 14] FIG. 14 is a diagram showing a frame transfer path when a failure occurs in the link between the port of the communication device in FIG. 13 and the port of the counterpart device. [Figure 15] FIG. 15 is a diagram showing a frame transfer path when a failure occurs in the link between the port of the communication device in FIG. 13 and the port of the counterpart device. [Figure 16] FIG. 16 is a diagram showing a frame transfer path when a failure occurs in the link between the port of the communication device in FIG. 13 and the port of the counterpart device. [Figure 17] FIG. 17 is a diagram showing a frame transfer path when a failure occurs in the link between the port of the communication device in FIG. 13 and the port of the counterpart device. [Figure 18]FIG. 18 is a flowchart showing the flow of a process for determining the transfer destination of a frame received from an ICL port, which is performed by the switch control unit of the communication system in FIG. 13. [Figure 19] FIG. 19 is a flowchart showing the flow of a process for determining the transfer destination of a frame received from an ICL port, which is performed by the switch control unit of the communication system in FIG. 13.

Embodiments for Carrying Out the Invention

[0013] [Description of Embodiments of the Present Disclosure] First, embodiments of the present disclosure will be listed and described.

[0014] (1) A communication system according to an embodiment of the present disclosure is interposed between a first opposing device and a second opposing device to realize communication between the first opposing device and the second opposing device. The communication system includes a first communication device and a second communication device. The first communication device and the second communication device construct a multi-chassis link aggregation with each of the first opposing device and the second opposing device. The first communication device includes an inter-chassis link port connected to the second communication device. The first communication device performs a liveness monitoring on a first path from the inter-chassis link port to the first opposing device. When the first communication device detects a failure in the first path, the first communication device transmits a frame from the inter-chassis link port to the first opposing device.

[0015] According to the above configuration, the frame forwarding path in the event of a failure is determined based on the results of a health check of the first path from the inter-chassis link port to the first opposing device. A failure in the first path means that frame forwarding between the second communication device and the first opposing device is not possible. When the first communication device receives a frame via the inter-chassis link port, it can determine that the frame was sent from the second opposing device to the second communication device. Therefore, the first communication device transmits the frame to the first opposing device. Since the frame forwarding destination can be determined based on the results of the health check of the first path, MC-LAG can be implemented without relying on MAC learning information. Because it is not necessary to constantly synchronize and share MAC learning information between the first and second communication devices, a large-capacity table for storing MAC learning information and learning synchronization between the first and second communication devices are not required.

[0016] (2) In the configuration described in (1) above, the second communication device includes a port connected to the first opposing device. The first communication device transmits frames from the inter-chassis link port to the second opposing device if it has not detected a fault in the first path. If the second communication device detects a fault in communication with the first opposing device via the port, it transmits frames from the second opposing device to the first communication device after a reference time has elapsed since the occurrence of the fault.

[0017] With the above configuration, if a failure occurs in the port of the second communication device connected to the first peer device, it is possible to avoid the frame transmitted from the second peer device to the second communication device returning to the second peer device via the first communication device (frame circulation). If, for example, the first communication device receives a frame via the inter-chassis link port between the time a failure occurs in the port of the second communication device and the time the failure is detected by the first communication device, the first communication device will transmit the frame from the inter-chassis link port to the second peer device because it has not detected a failure in the first path. Therefore, the problem of frame circulation occurs. With the above configuration, the possibility of a frame being forwarded to the inter-chassis link port before the detection of a failure is reflected in the reachability determination of the health check can be reduced, thus avoiding frame circulation.

[0018] (3) In the configuration described in (2) above, an inter-chassis link port is configured with a managed termination point as one end of the first path. The first opposing device is configured with a opposing managed termination point corresponding to the managed termination point as the other end of the first path. The first communication device performs a liveness check based on the reception status of ETH-CC frames transmitted from the opposing managed termination point to the managed termination point. The reference time is 3.5 times or more the transmission period of the ETH-CC frame.

[0019] With the above configuration, the reference time is 3.5 times or more the transmission cycle of the ETH-CC frame, which ensures that failures are reliably reflected in the reachability determination of the health check. As a result, frame cycles can be prevented.

[0020] (4) In the configuration described in (1) above, the first communication device performs a liveness check on both the first path and the second path from the inter-chassis link port to the second opposing device. If no fault is detected in the first path and a fault is detected in the second path, the first communication device transmits a frame from the inter-chassis link port to the second opposing device.

[0021] According to the above configuration, for a frame from the inter-chassis link port to be transmitted from the first communication device to the second opposing device, it is necessary that no fault is detected in the first path, and that a fault is detected in the second path. Since the occurrence of a fault in the second path is reliably reflected in the reachability determination of the liveness monitoring, frame circulation can be prevented.

[0022] (5) A communication device according to one embodiment of the present disclosure intervenes between a first opposing device and a second opposing device to realize communication between the first opposing device and the second opposing device. The communication device, together with other communication devices, can construct a multi-chassis link aggregation with each of the first opposing device and the second opposing device. The communication device comprises an inter-chassis link port, a reachability determination unit, and a transfer control unit. The inter-chassis link port is a port for communicating with other communication devices. The reachability determination unit performs a liveness check on the path from the inter-chassis link port to the first opposing device. The transfer control unit determines the destination of the frame from the inter-chassis link port. If the reachability determination unit detects a fault in the path, the transfer control unit transmits the frame to the first opposing device.

[0023] According to the above configuration, the frame transfer path in the event of a failure is determined based on the results of a health check of the path from the inter-chassis link port to the first opposing device. As a result, MC-LAG can be implemented without relying on MAC learning information. Furthermore, since it is not necessary to constantly synchronize and share MAC learning information between the communication device and other communication devices, a large-capacity table for storing MAC learning information and learning synchronization between the communication device and other communication devices are unnecessary.

[0024] (6) A communication control method according to one embodiment of the present disclosure is a communication control method performed in a communication system that intervenes between a first opposing device and a second opposing device to realize communication between the first opposing device and the second opposing device. The communication system comprises a first communication device and a second communication device. The first communication device and the second communication device establish a multi-chassis link aggregation with each of the first opposing device and the second opposing device. The first communication device includes an inter-chassis link port connected to the second communication device. The communication control method includes the steps of the first communication device performing a liveness check on the path from the inter-chassis link port to the first opposing device, and, if the first communication device detects a fault in the path, the first communication device transmits a frame from the inter-chassis link port to the first opposing device.

[0025] According to the above configuration, the frame transfer path in the event of a failure is determined based on the results of a health check of the path from the inter-chassis link port to the first opposing device. Therefore, MC-LAG can be implemented without relying on MAC learning information. Furthermore, since it is not necessary to constantly synchronize and share MAC learning information between the first and second communication devices, a large-capacity table for storing MAC learning information and learning synchronization between the first and second communication devices are unnecessary.

[0026] [Details of the embodiments of this disclosure] Embodiments of this disclosure will be described below with reference to the drawings. In the drawings, identical or corresponding parts are denoted by the same reference numerals, and their descriptions will not be repeated.

[0027] (Embodiment 1) Figure 1 is a diagram showing an example configuration of the communication system 100 according to Embodiment 1. The communication system 100 intervenes between opposing devices 1 and 2 to enable communication between them. The communication system 100 includes communication devices 101A and 101B.

[0028] The communication device 101A includes port P1A, port P2A, and inter-chassis link (ICL) port PicA. Port P1A is connected to port P11 of the opposing device 1. Port P2A is connected to port P21 of the opposing device 2.

[0029] The communication device 101B includes port P1B, port P2B, and ICL port PicB. Port P1B is connected to port P12 of the opposing device 1. Port P2B is connected to port P22 of the opposing device 2. ICL port PicB is connected to ICL port PicA.

[0030] Communication devices 101A and 101B establish multi-chassis link aggregation with opposing devices 1 and 2. In the multi-chassis link aggregation with opposing device 1, the physical links connecting ports P1A and P11 and the physical links connecting ports P1B and P12 are combined to form a single logical link. Similarly, in the multi-chassis link aggregation with opposing device 2, the physical links connecting ports P2A and P21 and the physical links connecting ports P2B and P22 are combined to form a single logical link.

[0031] If there are no problems in the communication path between the communication system 100 and each of the opposing devices 1 and 2, a frame transmitted from port P11 of opposing device 1 to the communication system 100 is received at port P1A of communication device 101A and forwarded to opposing device 2 from port P2A. A frame transmitted from port P12 of opposing device 1 to the communication system 100 is received at port P1B of communication device 101B and forwarded to opposing device 2 from port P2B.

[0032] Similarly, if there are no problems in the communication path between the communication system 100 and each of the opposing devices 1 and 2, a frame transmitted from port P21 of opposing device 2 to the communication system 100 is received at port P2A of communication device 101A and forwarded to opposing device 1 from port P1A. A frame transmitted from P22 of opposing device 2 to the communication system 100 is received at port P2B of communication device 101B and forwarded to opposing device 1 from port P1B.

[0033] Ports PicA and P12 are configured with corresponding Maintenance Association End Points (MEPs) 10A and 12, respectively. Ports PicB and P11 are configured with corresponding MEPs 10B and 11, respectively. Each of MEPs 10A, 10B, 11, and 12 may conform to ITU-T Y.1731. Each of MEPs 10A, 10B, 11, and 12 verifies whether frame transmission is functioning correctly.

[0034] Furthermore, ITU-T Y.1731 specifies a procedure for measuring frame delay between MEPs (Managed Event Providers) by exchanging OAM (Operations, Administration and Maintenance) frames for frame delay measurement between a pair of devices connected to each other via a transmission path. Although not shown in the diagram, MEPs are also configured on ports P1A and P2A, and their counterpart MEPs are also configured on ports P11 and P21. Communication device 101A performs health checks on both the path between ports P1A and P11 and the path between ports P2A and P21. Similarly, MEPs are also configured on ports P1B and P2B, and their counterpart MEPs are also configured on ports P12 and P22. Communication device 101B performs health checks on both the path between ports P1B and P12 and the path between ports P2B and P22.

[0035] In Embodiment 1, communication device 101A performs a liveness check on the path from ICL port PicA through communication device 101B to port P12 of peer device 1, based on the reception status of ETH-CC frames transmitted from MEP12 to MEP10A. Specifically, communication device 101A transmits an ETH-CC (Ethernet Continuity Check) frame from MEP10A to MEP12. If LOC (Loss of Continuity) or RDI (Remote Defect Indication) is detected in the ETH-CC frame transmitted from MEP12 to MEP10A, communication device 101A determines that a failure has occurred in the path. If neither RDI nor LOC is detected in MEP10A, communication device 101A determines that there is no failure in the path and that the path is reachable (or connectable).

[0036] Similarly, communication device 101B performs a health check on the path from ICL port PicB through communication device 101A to port P11 of peer device 1, based on the reception status of ETH-CC frames transmitted from MEP11 to MEP10B. Specifically, communication device 101B transmits ETH-CC frames from MEP10B to MEP11. If LOC or RDI is detected in the ETH-CC frame transmitted from MEP11 to MEP10B, communication device 101B determines that a failure has occurred in that path. If neither RDI nor LOC is detected in MEP10B, communication device 101B determines that there is no failure in that path and that the path is reachable. In Embodiment 1, the path from ICL port PicA to port P12 of the opposing device 1 via communication device 101B, and the path from ICL port PicB to port P11 of the opposing device 1 via communication device 101A, each correspond to a "first path" from the ICL port to the first opposing device. That is, MEP10A is set on ICL port PicA as one end of the path from ICL port PicA to port P12 of the opposing device 1 via communication device 101B, and MEP12 is set on the opposing device 1 as the other end of the said path. Similarly, MEP10B is set on ICL port PicB as one end of the path from ICL port PicB to port P11 of the opposing device 1 via communication device 101A, and MEP11 is set on the opposing device 1 as the other end of the said path.

[0037] Figure 2 shows the format of an ETH-CC frame. The "Flags" field shown in Figure 2 stores RDI flags and period information.

[0038] RDI is a feature supported by ITU-T Y.1731. By setting the RDI flag of an ETH-CC frame to "1", a failure in an MEP is notified to the corresponding MEP (opposite MEP). If the RDI flag of an ETH-CC received from the opposite MEP is "1", the MEP detects RDI (RDI state). Also, if the MEP cannot receive ETH-CC frames from the opposite MEP for 3.5 cycles or more, the MEP detects LOC (LOC state). If the MEP is in the LOC state, it sets the RDI flag of the ETH-CC frame to "1", and if it is not in the LOC state, it sets the RDI flag to "0" and sends the ETH-CC frame to the opposite MEP at predetermined intervals. Therefore, two opposing MEPs can determine that bidirectional frame transmission is working normally when neither the LOC state nor the RDI state is present.

[0039] The "Period" item in the "Flags" field sets the length of the period during which ETH-CC frames are exchanged. A default transfer period is specified depending on the use of the ETH-CC frames. For example, for fault management, the length of one period is 1 second, and for performance monitoring, the default transfer period is 100 milliseconds.

[0040] The "MEG ID" field stores the MEG ID used to identify the MEG (Maintenance Entity Group) to which the MEP belongs.

[0041] Figure 3 is a functional block diagram schematically showing the configurations of the communication devices 101A and 101B in Figure 1. The configuration shown in Figure 3 is common to both communication devices 101A and 101B. In Figure 3, each of the communication devices 101A and 101B is referred to collectively as communication device 101. ICL ports PicA and PicB are referred to collectively as ICL port Pic. Ports P1A and P1B are referred to collectively as port P1. Ports P2A and P2B are referred to collectively as port P2.

[0042] In addition to ICL port Pic and ports P1 and P2, the communication device 101 further includes a switch unit 10, a switch control unit 12, an LACP control unit 14, an operation command receiving unit 16, a reachability determination unit 20, an ETH-CC frame transmission unit 21, an ETH-CC frame reception unit 22, an MEP management table 24, an MEP setting table 25, and a forwarding destination setting table 26.

[0043] The communication device 101 may include multiple ICL ports. Each of ports P1 and P2 may be a physical port or a logical port composed of multiple physical ports. If a port is a logical port, a frame received from any of the multiple physical ports that make up that logical port is treated as if it was received by that logical port. Also, a frame transmitted from a logical port is actually transmitted from any of the multiple physical ports that make up that logical port.

[0044] The MEP management table 24 manages information regarding the transfer availability status (results of liveness monitoring) detected by the MEP installed on the communication device 101. Table 1 below shows an example of the records included in the MEP management table 24.

[0045] [Table 1]

[0046] The record shown in Table 1 indicates that MEP number 1001 has VLAN ID 1001. Furthermore, this record indicates that neither LOC nor RDI is detected at the MEP, and that reachability is established between the MEP and the opposing MEP. The VLAN ID of the MEP is identical to the VLAN ID of the opposing MEP and is unique to the path between the two MEPs. In other words, the MEP and the opposing MEP belong to the same and unique VLAN. In Figure 1, MEPs 10A and 1012 belong to the same VLAN, while MEPs 10B and 101 belong to different VLANs.

[0047] Referring again to Figure 3, the MEP configuration table 25 stores an ICL management number representing an ICL port and an MEP number representing the MEP corresponding to the route to be monitored for health status. Table 2 below shows an example of the records included in the MEP configuration table 25.

[0048] [Table 2]

[0049] The record shown in Table 2 indicates that MEP number 1001 is configured on the ICL port with ICL management number 1, corresponding to liveness monitoring of the path to the opposing device 1. In Figure 3, the ICL port with ICL management number 1 is Pic.

[0050] The forwarding destination setting table 26 is configured with the ports that can forward frames from the ICL port Pic. Table 3 below shows an example of the records included in the MEP setting table 25.

[0051] [Table 3]

[0052] The records shown in Table 3 indicate that frames received from the ICL port with ICL management number 1 are forwarded to port P1 or P2.

[0053] The ETH-CC frame transmission unit 21 generates an ETH-CC frame (check frame) based on the information registered in the MEP management table 24. The ETH-CC frame transmission unit 21 forwards the check frame from the ICL port Pic via the switch unit 10. Specifically, if the MEP is in the LOC state, the ETH-CC frame transmission unit 21 sets the RDI flag of the check frame to "1". If the MEP is not in the LOC state, the ETH-CC frame transmission unit 21 sets the RDI flag of the check frame to "0". The ETH-CC frame transmission unit 21 transmits the check frame to the opposing MEP at predetermined intervals.

[0054] The ETH-CC frame receiver 22 receives inspection frames from the opposing MEP through the ICL port Pic and the switch unit 10, and writes information about the MEP status contained in the inspection frame to the MEP management table 24. This updates the information registered in the MEP management table 24 to the latest information. Specifically, the ETH-CC frame receiver 22 detects RDI (RDI state) if the RDI flag of the inspection frame received from the opposing MEP is "1". Also, if the ETH-CC frame receiver 22 cannot receive inspection frames from the opposing MEP for 3.5 cycles or more, it detects LOC (LOC state). If the RDI state or LOC state of any MEP registered in the MEP management table 24 changes, the ETH-CC frame receiver 22 outputs an LOC / RDI change notification to the reachability determination unit 20.

[0055] The reachability determination unit 20 performs a liveness check on each MEP registered in the MEP management table 24 in response to receiving an LOC / RDI change notification from the ETH-CC frame receiving unit 22. The reachability determination unit 20 writes the result of the liveness check (reachability) to the MEP management table 24 and outputs an MEP management table update notification to the switch control unit 12.

[0056] In response to receiving an MEP management table update notification from the reachability determination unit 20, the switch control unit 12 controls the switch unit 10, which functions as a frame forwarding switch, by referring to the MEP management table 24, the MEP setting table 25, and the forwarding destination setting table 26. This determines the direction of frame forwarding by the communication device 101 (in other words, the frame forwarding path). Therefore, the switch control unit 12 functions as a forwarding control unit that controls the forwarding of frames. Specifically, the switch control unit 12 controls the switch unit 10 to forward frames received from port P1 to port P2 or ICL port Pic depending on whether there is a fault at port P2. The switch control unit 12 controls the switch unit 10 to forward frames received from port P2 to port P1 or ICL port Pic depending on whether there is a fault at port P1. The switch control unit 12 controls the switch unit 10 to forward frames received from port Pic to port P1 or P2 based on the results of the liveness check.

[0057] The LACP (Link Aggregation Control Protocol) control unit 14 controls the switch unit 10 to realize MC-LAG together with other communication devices.

[0058] The operation command receiving unit 16 can receive various commands from outside the communication device 101 and write the information to the MEP management table 24, the MEP setting table 25, and the transfer destination setting table 26. For example, the operation command receiving unit 16 is connected to a management device (not shown) via a network separate from the network through which the main signal passes. The management device is connected to, for example, the operator's personal computer (PC) and transfers commands entered through the personal computer to the operation command receiving unit 16.

[0059] The operation command receiving unit 16 and the switch control unit 12 may be implemented by a processor reading and executing a program stored in memory (not shown). Alternatively, one or both of the operation command receiving unit 16 and the switch control unit 12 may be implemented by dedicated hardware circuits (for example, an ASIC (Application Specific Integrated Circuit) or an FPGA (Field-Programmable Gate Array)).

[0060] Figure 4 shows the frame transfer path when a failure occurs in the link between port P1A of communication device 101A and port P11 of opposing device 1 in Figure 1. In the explanation referring to Figure 4, Tables 1 to 3 are assumed to be tables relating to communication device 101B.

[0061] As shown in Figure 4, when communication device 101A detects a fault in port P1A, it forwards the frame received from the opposing device 2 from ICL port PicA to communication device 101B.

[0062] The communication device 101B detects a failure in the path from MEP10B to MEP11 during the health check. As a result, as shown in Table 4 below, the LOC of the record included in the MEP management table 24 of the communication device 101B is updated from "Not Detected" to "Detected," and the reachability of the record is updated from "Yes" to "No."

[0063] [Table 4]

[0064] Since the path from MEP10B to MEP11 is unreachable, the frame forwarded from communication device 101A is a frame transmitted from the opposing device 2. Therefore, communication device 101B forwards the frame from communication device 101A to the opposing device 1.

[0065] Figure 5 shows the frame transfer path when a failure occurs in the link between port P1B of communication device 101B and port P12 of opposing device 1 in Figure 1. In the explanation referring to Figure 5, Tables 1 to 3 are assumed to be tables relating to communication device 101A.

[0066] As shown in Figure 5, when the communication device 101B detects a fault in port P1B, it forwards the frame received from the opposing device 2 from ICL port PicB to the communication device 101A.

[0067] The communication device 101A detects a failure in the path from MEP10A to MEP12 during its health monitoring. As a result, the records in the communication device 101A's MEP management table 24 are updated in the same way as the records shown in Table 4.

[0068] If the path from MEP10A to MEP12 is unreachable, the frame forwarded from communication device 101B is a frame transmitted from the opposing device 2. Therefore, communication device 101A forwards the frame from communication device 101B to the opposing device 1.

[0069] Figure 6 shows the frame forwarding path when a failure occurs in the link between port P2A of communication device 101A and port P21 of peer device 2 in Figure 1. As shown in Figure 6, when communication device 101A detects a failure in port P2A, it forwards the frame received from peer device 1 from ICL port PicA to communication device 101B.

[0070] Communication device 101B confirms reachability of the path from MEP10B to MEP11 during the liveness check. If reachability of the path from MEP10B to 11 is confirmed, the frame forwarded from communication device 101A is a frame transmitted from the opposing device 1. Therefore, communication device 101B forwards the frame from communication device 101A to the opposing device 2.

[0071] Figure 7 shows the frame forwarding path when a failure occurs in the link between port P2B of communication device 101B and port P22 of peer device 2 in Figure 1. As shown in Figure 7, when communication device 101B detects a failure in port P2B, it forwards the frame received from peer device 1 from ICL port PicB to communication device 101A.

[0072] Communication device 101A confirms reachability of the path from MEP10A to MEP12 during the liveness check. If reachability of the path from MEP10A to MEP12 is confirmed, the frame forwarded from communication device 101B is a frame transmitted from the opposing device 1. Therefore, communication device 101A forwards the frame from communication device 101B to the opposing device 2.

[0073] In the communication system 100, the frame transfer path in the event of a failure is determined based on the results of the health check performed by the MEP configured on the respective ICL ports of the communication devices 101A and 101B. Since it is not necessary to constantly synchronize and share MAC learning information between the communication devices 101A and 101B, the communication system 100 does not require a large-capacity table to store MAC learning information or synchronization of learning information between the communication devices 101A and 101B. Therefore, the communication system 100 can realize MC-LAG without relying on MAC learning information.

[0074] Figure 8 is a flowchart showing the MEP setting process performed in each of the opposing device 1 and communication devices 101A and 101B shown in Figure 1. Hereafter, steps will simply be referred to as S.

[0075] As shown in Figure 8, in S10, the user configures a VLAN and MEP for health monitoring on the opposing device 1. Specifically, the user registers port P11 of the opposing device 1 connected to communication device 101A with VLAN_A and sets MEP11 belonging to VLAN_A to port P11. The user also registers port P12 of the opposing device 1 connected to communication device 101B with VLAN_B and sets MEP12 belonging to VLAN_B to port P12, and proceeds to S11.

[0076] In S11, the user configures a VLAN and MEP for health monitoring on the communication device 101A. Specifically, the user registers port P1A and ICL port PicA, which are connected to the opposing device 1, to VLAN_A. The user also registers ICL port PicA to VLAN_B and configures MEP 10A, which belongs to VLAN_B, to ICL port PicA, and then proceeds to S12.

[0077] In S12, the user configures a VLAN and MEP for health monitoring on the communication device 101B. Specifically, the user registers port P1B and ICL port PicB, which are connected to the opposing device 1, to VLAN_B. The user also registers ICL port PicB to VLAN_A and configures MEP10B, which belongs to VLAN_A, on ICL port PicB, and then terminates the process. Note that steps S11 and S12 shown in Figure 8 are performed by the user (or system administrator) via, for example, the operation command reception unit shown in Figure 3.

[0078] Figure 9 is a flowchart showing the flow of the health monitoring process performed by the reachability determination unit 20 in Figure 3. The process shown in Figure 9 is executed for each MEP registered in the MEP management table 24.

[0079] As shown in Figure 9, in S20, the reachability determination unit 20 determines whether or not it has received an LOC / RDI change notification from the ETH-CC frame receiving unit 22. If the LOC / RDI change notification has not been received (NO in S20), the reachability determination unit 20 returns to processing in S20. If the LOC / RDI change notification has been received (YES in S20), in S21, the reachability determination unit 20 determines whether or not an LOC has been detected in the MEP with MEP number N (where N is a natural number). If an LOC has been detected (YES in S21), the reachability determination unit 20 proceeds to processing in S22. If an LOC has not been detected (NO in S21), in S23, the reachability determination unit 20 determines whether or not an RDI has not been detected in the MEP with MEP number N.

[0080] If an LOC is detected (YES in S21) or an RDI is detected (YES in S23), in S22, the reachability determination unit 20 updates the reachability of the MEP with MEP management number N in the MEP management table 24 to "None" and proceeds to S25. If an RDI is not detected (NO in S23), in S24, the reachability determination unit 20 updates the reachability of the MEP with MEP management number N in the MEP management table 24 to "Yes" and proceeds to S25. In S25, the reachability determination unit 20 outputs an MEP management table update notification to the switch control unit 12 and terminates processing. The reachability determination unit 20 may be included in the switch control unit 12.

[0081] Figure 10 is a flowchart showing the process of determining the forwarding destination of frames received from the ICL port, as performed by the switch control unit 12 in Figure 3.

[0082] As shown in Figure 10, in S31, the switch control unit 12 determines whether or not it has received an MEP management table update notification from the reachability determination unit 20. If it has not received an MEP management table update notification (NO in S31), the switch control unit 12 returns to processing S31. If it has received an MEP management table update notification (YES in S31), in S32, the switch control unit 12 refers to the MEP setting table 25, obtains the MEP number for the opposing device 1, and proceeds to processing S33. In S33, the switch control unit 12 obtains the reachability of the MEP number obtained from the MEP management table 24 and proceeds to processing S34.

[0083] In S34, the switch control unit 12 determines whether the port is reachable, as determined in S33. If it is reachable (YES in S34), in S35, the switch control unit 12 refers to the forwarding destination setting table 26 to obtain the port connected to the opposing device 2 and proceeds to S37. If it is not reachable (NO in S34), in S36, the switch control unit 12 refers to the forwarding destination setting table 26 to obtain the port connected to the opposing device 1 and proceeds to S37. In S37, the switch control unit 12 sets the port to which the frame received from the ICL port will be forwarded to the port obtained from the forwarding destination setting table 26 in S35 or S36, and terminates the process.

[0084] Figure 11 is a flowchart showing the process performed by the switch control unit 12 in Figure 3 to determine the forwarding destination of a frame received from either the opposing device 1 or 2. The process shown in Figure 11 is performed for each of the opposing devices 1 and 2. In the following explanation, if "one opposing device" is opposing device 1, then "the other opposing device" is opposing device 2. If "one opposing device" is opposing device 2, then "the other opposing device" is opposing device 1.

[0085] As shown in Figure 11, in S41, the switch control unit 12 determines whether the state of the port connected to one of the opposing devices has changed. If the state of the port connected to one of the opposing devices has not changed (NO in S41), the switch control unit 12 returns to processing in S41. If the state of the port connected to one of the opposing devices has changed (YES in S41), in S42, the switch control unit 12 determines whether a failure has occurred in the port connected to one of the opposing devices.

[0086] If there is no failure in the port connected to one of the opposing devices (NO in S42), in S44, the switch control unit 12 sets the port for forwarding frames received from the other opposing device to the port connected to one of the opposing devices and terminates processing. If there is a failure in the port connected to one of the opposing devices (YES in S42), in S43, the switch control unit 12 sets the port for forwarding frames received from the other opposing device to ICL port Pic and terminates processing.

[0087] As described above, according to the communication system of Embodiment 1, each communication device is configured with an MEP for monitoring the availability of the frame path. This makes it possible to implement MC-LAG without relying on MAC learning information.

[0088] In the above explanation, the "first route" is the route from the ICL port of one communication device through the other communication device to the opposing device 1. However, the "first route" may also be the route from the ICL port of one communication device through the other communication device to the opposing device 2.

[0089] (Embodiment 2) In Embodiment 1, one communication device monitors the availability of the path from its inter-chassis link port to one of the two opposing devices via the other communication device. A certain amount of time (for example, about 3.5 times the transmission cycle of an ETH-CC frame) may be required for the reachability in the availability monitoring to change from available to unavailable after a failure occurs in the path. Therefore, even if the failure occurs, the reachability in the availability monitoring will be determined to be available until that time has elapsed. On the other hand, during the time between the occurrence of the failure and the change in reachability in the availability monitoring, the communication device may forward a frame received from the opposing device to the ICL port. Therefore, the communication device may recognize a frame received via the ICL port as a frame sent through a path that was determined to be reachable. In this case, the communication device forwards the frame according to the determination that it is reachable, and transmits the frame to the other opposing device.

[0090] Thus, between the time a failure occurs and the time it is reflected in the reachability determination in the liveness monitoring, there is a possibility that frames transmitted from a peer device that is not subject to liveness monitoring may be returned to that peer device. Therefore, in Embodiment 2, after a standard time has elapsed since a failure occurred in the port of the communication device connected to the peer device that is subject to liveness monitoring, frames received from a communication device that is not subject to liveness monitoring are forwarded to the ICL port. According to the communication system of Embodiment 2, the possibility of frames being forwarded to the ICL port before the detection of the failure is reflected in the reachability determination of the liveness monitoring can be reduced, thereby reducing frame circulation.

[0091] Since the configuration of the communication system according to Embodiment 2 is the same as that of the communication system according to Embodiment 1, the explanation will not be repeated.

[0092] Figure 12 is a flowchart showing the process flow for determining the forwarding destination of frames received from a peer device 2 that is not subject to liveness monitoring, as performed by the switch control unit 12 of the communication system according to Embodiment 2. The flowchart shown in Figure 12 is the flowchart shown in Figure 11 with S50 and S51 added.

[0093] As shown in Figure 12, the switch control unit 12 executes S41 and S42 in the same manner as in Embodiment 1. If the answer to S42 is NO, it executes S44 in the same manner as in Embodiment 1 and terminates the process.

[0094] If the answer in S42 is YES, in S50 the switch control unit 12 determines whether a reference time has elapsed since the time the fault occurred. If the reference time has not elapsed since the time the fault occurred (NO in S50), in S51 the switch control unit 12 sets the switch unit 10 to discard the frame received from the opposing device 2 and returns the process to S42. If the reference time has elapsed since the time the fault occurred (YES in S50), in S43 the switch control unit 12 sets the port to which the frame received from the opposing device 2 is forwarded to the ICL port and terminates the process.

[0095] From the standpoint of preventing frame cycling, the length of the reference time for S50 is preferably such that it can reliably reflect the occurrence of a failure in the reachability determination of the liveness monitoring. In Embodiment 2, the liveness of the communication path is monitored by the transmission and reception of ETH-CC frames. If the MEP cannot receive ETH-CC frames from the opposing MEP for 3.5 cycles or more, the MEP detects a LOC. Therefore, it is desirable that the reference time be 3.5 times or more the transmission cycle of the ETH-CC frame.

[0096] As described above, the communication system according to Embodiment 2 can achieve MC-LAG without relying on MAC learning information, just like the communication system according to Embodiment 1. Furthermore, Embodiment 2 can suppress the frame cycle in which a frame is returned to the opposing device that sent it, thereby suppressing the occurrence of unnecessary communication.

[0097] Furthermore, similar to Embodiment 1, in Embodiment 2, the "first path" is not limited to a path from the ICL port of one communication device through the other communication device to the opposing device 1, but may also be a path from the ICL port of one communication device through the other communication device to the opposing device 2. Also, it is not necessary for both communication devices to perform health checks on paths leading to the same opposing device; the two communication devices may perform health checks on paths leading to different opposing devices.

[0098] (Embodiment 3) In Embodiment 2, in order to suppress frame recurrence, the system waits for a specified time after the failure occurs before transferring the frame to the ICL port. In Embodiment 3, in order to suppress frame recurrence, a health check is performed on each of the two opposing devices.

[0099] Figure 13 is a diagram showing an example configuration of the communication system 300 according to Embodiment 3. Figure 13 corresponds to Figure 1 of Embodiment 1. The communication system 300 has a configuration in which MEP20A, 20B, 21, and 22 are added to the communication system 100 shown in Figure 1. The configuration of the communication system 300 other than these is the same as the configuration of the communication system 100, so the explanation of the same configuration will not be repeated.

[0100] As shown in Figure 13, ports PicA and P22 are configured with corresponding MEP20A and MEP2022, respectively. Ports PicB and P21 are configured with corresponding MEP20B and MEP2021, respectively. Similar to Embodiment 1, each of MEP20A, 20B, 21, and 22 can conform to ITU-T Y.1731. Each of MEP20A, 20B, 21, and 22 verifies whether frame transmission is functioning correctly.

[0101] In the communication system 300, MEPs 20A and 20B have been added to the communication devices 101A and 101B, respectively, for monitoring the status of the opposing device 2. As a result, the records included in the MEP management table 24 and the MEP setting table 25 differ from those in Embodiment 1. Table 5 below shows an example of the records included in the MEP management table 24 of the communication device in the communication system 300.

[0102] [Table 5]

[0103] As shown in Table 5, records corresponding to the MEP added for monitoring the status of the opposing device 2 have been added to Table 1. The added records indicate that neither LOC nor RDI were detected in the MEP for monitoring the status of the opposing device 2, and that reachability was confirmed between the MEP and the opposing MEP.

[0104] Table 6 below shows an example of records included in the MEP setting table 25 of the communication device in the communication system 300.

[0105] [Table 6]

[0106] As shown in Table 6, MEP numbers corresponding to the MEPs added for monitoring the status of the opposing device 2 have been added. The records shown in Table 6 indicate that, in addition to MEP 1001, which corresponds to monitoring the status of the path to opposing device 1, is set on the ICL port of ICL management number 1, MEP 1002, which corresponds to monitoring the status of the path to opposing device 2.

[0107] Figure 14 shows the frame transfer path when a failure occurs in the link between port P1A of communication device 101A and port P11 of opposing device 1 shown in Figure 13. Figure 14 corresponds to Figure 4 of Embodiment 1. In the explanation referring to Figures 14 and 16, Tables 5, 6, and 3 are assumed to be tables relating to communication device 101B.

[0108] As shown in Figure 14, when communication device 101A detects a fault in port P1A, it forwards the frame received from the opposing device 2 from ICL port PicA to communication device 101B.

[0109] In its health check, the communication device 101B detects a failure in the path from MEP10B to MEP11 (corresponding to MEP number 1001 in Table 5) and confirms reachability in the path from MEP20B to 21 (corresponding to MEP number 1002 in Table 5). As a result, as shown in Table 7 below, the LOC of the record corresponding to MEP number 1001 in the MEP management table 24 of the communication device 101B is updated from "Not Detected" to "Detected," and the reachability of the record is updated from "Yes" to "No."

[0110] [Table 7]

[0111] If the path from MEP10B to 11 is unreachable, but the path from MEP20B to 21 is reachable, then the frame forwarded from communication device 101A is a frame transmitted from the opposing device 2. Therefore, communication device 101B forwards the frame from communication device 101A to the opposing device 1.

[0112] Figure 15 shows the frame transfer path when a failure occurs in the link between port P1B of communication device 101B and port P12 of opposing device 1 in Figure 13. Figure 15 corresponds to Figure 5 of Embodiment 1. In the explanation referring to Figures 15 and 17, Tables 5, 6, and 3 are assumed to be tables relating to communication device 101A.

[0113] As shown in Figure 15, when communication device 101B detects a fault in port P1B, it forwards the frame received from the opposing device 2 from ICL port PicB to communication device 101A.

[0114] The communication device 101A detects a failure in the path from MEP10A to MEP12 (corresponding to MEP number 1001 in Table 5) during its health check, and confirms reachability in the path from MEP20A to 22 (corresponding to MEP number 1002 in Table 5). As a result, the record corresponding to MEP number 1001 in the MEP management table 24 of the communication device 101A is updated in the same way as the record corresponding to MEP number 1001 shown in Table 7.

[0115] If the path from MEP10A to MEP12 is unreachable, but the path from MEP20A to MEP22 is reachable, then the frame forwarded from communication device 101B is a frame transmitted from the opposing device 2. Therefore, communication device 101A forwards the frame from communication device 101B to the opposing device 1.

[0116] Figure 16 shows the frame transfer path when a failure occurs in the link between port P2A of communication device 101A and port P21 of the opposing device 2, as shown in Figure 13. Figure 16 corresponds to Figure 6 of Embodiment 1. As shown in Figure 16, when communication device 101A detects a failure in the physical link between port P2A and port P12, it transfers the frame received from opposing device 1 from ICL port PicA to communication device 101B.

[0117] Communication device 101B, in its health check, confirms reachability to the path from MEP10B to MEP11 and detects a failure in the path from MEP20B to 21. As a result, as shown in Table 8 below, the LOC of the record corresponding to MEP number 1002 in the MEP management table 24 of communication device 101B is updated from "Not detected" to "Detected," and the reachability of the record is updated from "Yes" to "No."

[0118] [Table 8]

[0119] If the path from MEP10B to MEP11 is reachable, but the path from MEP20B to MEP21 is not reachable, then the frame forwarded from communication device 101A is a frame transmitted from the opposing device 1. Therefore, communication device 101B forwards the frame from communication device 101A to the opposing device 2.

[0120] Figure 17 shows the frame transfer path when a failure occurs in the link between port P2B of communication device 101B and port P22 of peer device 2 in Figure 13. Figure 17 corresponds to Figure 7 of Embodiment 1. As shown in Figure 17, when communication device 101B detects a failure in port P2B, it transfers the frame received from peer device 1 from ICL port PicB to communication device 101A.

[0121] The communication device 101A, in its status monitoring, confirms the reachability of the path from MEP10A to MEP12 and detects a failure in the path from MEP20A to 22. As a result, among the records included in the communication device 101A's MEP management table 24, the record corresponding to MEP number 1002 is updated in the same way as the record corresponding to MEP number 1002 shown in Table 8.

[0122] If the path from MEP10A to MEP12 is reachable, but the path from MEP20A to MEP22 is not reachable, then the frame forwarded from communication device 101B is a frame transmitted from the opposing device 1. Therefore, communication device 101A forwards the frame from communication device 101B to the opposing device 2.

[0123] In the communication system 300, for a frame from the ICL port Pic to be transmitted from the communication device 101 to one of the opposing devices 1 or 2, it is necessary that no faults are detected in the path to one of the opposing devices, and that a fault is detected in the path to the other opposing device. Since the reachability determination of the liveness monitoring reliably reflects the occurrence of a fault in the path to the other opposing device, frame circulation can be prevented.

[0124] Figures 18 and 19 are flowcharts showing the process flow for determining the forwarding destination of frames received from the ICL port, performed by the switch control unit 12 of the communication system 300 in Figure 13. Figures 18 and 19 correspond to Figure 1 of Embodiment 1. The flowcharts shown in Figures 18 and 19 are connected to each other at nodes A1 and A2.

[0125] As shown in Figure 18, in S61, the switch control unit 12 determines whether or not it has received an MEP management table update notification from the reachability determination unit 20. If it has not received an MEP management table update notification (NO in S61), the switch control unit 12 returns to processing S61. If it has received an MEP management table update notification (YES in S61), the switch control unit 12 proceeds to processing S62 via node A1. In S62, the switch control unit 12 refers to the MEP setting table 25, obtains the MEP numbers for opposing devices 1 and 2, and proceeds to processing S63. In S63, the switch control unit 12 obtains the reachability of each of the MEP numbers for opposing devices 1 and 2 and proceeds to processing S64 in Figure 19 via node A2.

[0126] As shown in Figure 19, in S64, the switch control unit 12 determines whether or not the opposing device 1 is reachable. If the opposing device 1 is reachable (YES in S64), in S65, the switch control unit 12 determines whether or not the opposing device 2 is reachable. If the opposing device 2 is reachable (YES in S65), in S66, the switch control unit 12 sets the switch unit 10 to discard the frame received from the ICL port and returns processing to node A1. If both opposing devices 1 and 2 are reachable in the liveness check from one communication device, there is no need to transfer a frame from the other communication device to the one communication device via the ICL port. Therefore, it is not normal for the switch unit 10 of one communication device to receive a frame via the ICL port. For this reason, the switch control unit 12 of one communication device sets the switch unit 10 to discard the frame received from the ICL port.

[0127] If the opposing device 2 is not reachable (NO in S65), in S67, the switch control unit 12 refers to the forwarding destination setting table 26 to obtain the port connected to the opposing device 2 and proceeds to S68.

[0128] If the opposing device 1 is not reachable (NO in S64), in S69, the switch control unit 12 determines whether the opposing device 2 is reachable. If the opposing device 2 is reachable (YES in S69), in S70, the switch control unit 12 refers to the forwarding destination setting table 26 to obtain the port connected to the opposing device 1 and proceeds to S68. In S68, the switch control unit 12 sets the port to which the frame received from the ICL port will be forwarded to the port obtained from the forwarding destination setting table 26 in S67 or S70, and terminates the process.

[0129] If the opposing device 2 is unreachable (NO in S69), in S71, the switch control unit 12 configures the switch unit 10 to discard the frame received from the ICL port and returns processing to node A1. If both opposing devices 1 and 2 are unreachable during the health check from one communication device, there are no frames to be transferred from the other communication device to the first communication device via the ICL port. Therefore, it is not normal for the switch unit 10 of one communication device to receive a frame via the ICL port. For this reason, the switch control unit 12 of one communication device configures the switch unit 10 to discard the frame received from the ICL port.

[0130] As described above, the communication system according to Embodiment 3 makes it possible to implement MC-LAG without relying on MAC learning information. Furthermore, this communication system can suppress the occurrence of unnecessary communication such as frame looping.

[0131] The embodiments disclosed herein should be considered in all respects to be illustrative and not restrictive. The scope of the present invention is indicated by the claims rather than by the embodiments described above, and all modifications within the meaning and scope of the claims are intended to be included. [Explanation of Symbols]

[0132] 1,2 Opposing device 10 Switch section 12 Switch control unit 14 Control Unit 16 Operation command reception unit 20 Reachability determination section 21 ETH-CC Frame Transmission Unit 22 ETH-CC Frame Receiver 24 Management Tables 25 Configuration Table 26 Transfer destination setting table 100,300 communication systems 101,101A,101B Communication equipment A1, A2 nodes Ports P1, P1B, P1A, P2, P2B, P2A, P11, P12, P21, P22 Pic, PicA, PicB ICL port

Claims

1. A communication system interposed between a first opposing device and a second opposing device to realize communication between the first opposing device and the second opposing device, The system includes a first communication device and a second communication device that establish a multi-chassis link aggregation between each of the first opposing device and the second opposing device, The first communication device is Includes an inter-chassis link port connected to the second communication device, The system performs a liveness check on the first path from the inter-chassis link port through the second communication device to the first opposing device, and if a failure is detected in the first path, it transmits a frame from the inter-chassis link port to the first opposing device. The second communication device includes a port connected to the first opposing device, If the first communication device does not detect a fault in the first path, it transmits a frame from the inter-chassis link port to the second opposing device. If the second communication device detects a failure in communication with the first opposing device via the port, it transmits a frame from the second opposing device to the first communication device after a standard time has elapsed since the occurrence of the failure. A management termination point is set as one end of the first path in the inter-chassis link port. The first opposing device is configured with an opposing management termination point as the other end of the first path. The first communication device performs the liveness check based on the reception status of the ETH-CC frame transmitted from the opposing management termination point to the management termination point. A communication system in which the reference time is 3.5 times or more the transmission period of the ETH-CC frame.

2. A communication system interposed between a first opposing device and a second opposing device to realize communication between the first opposing device and the second opposing device, The system includes a first communication device and a second communication device that establish a multi-chassis link aggregation between each of the first opposing device and the second opposing device, The first communication device is Includes an inter-chassis link port connected to the second communication device, A communication system that performs a liveness check on each of the following paths: a first path from the inter-chassis link port to the first opposing device via the second communication device, and a second path from the inter-chassis link port to the second opposing device via the second communication device; if a failure is detected in the first path, a frame from the inter-chassis link port is transmitted to the first opposing device; and if no failure is detected in the first path and a failure is detected in the second path, a frame from the inter-chassis link port is transmitted to the second opposing device.

3. A communication device interposed between a first opposing device and a second opposing device to realize communication between the first opposing device and the second opposing device, The communication device, together with other communication devices, can establish a multi-chassis link aggregation with each of the first opposing device and the second opposing device. The aforementioned communication device is An inter-chassis link port for communicating with the aforementioned other communication devices, A reachability determination unit that determines the reachability of the path from the inter-chassis link port to the first opposing device via the other communication device, The system includes a transfer control unit that determines the destination of the frame transmitted from the inter-chassis link port, If the reachability determination unit detects an obstacle in the path, the transfer control unit transmits the frame to the first opposing device. The other communication device includes a port connected to the first opposing device. If the reachability determination unit does not detect any obstruction in the path, the transfer control unit transmits the frame from the inter-chassis link port to the second opposing device. If the other communication device detects a failure in communication with the first opposing device via the port, it transmits a frame from the second opposing device to the communication device after a standard time has elapsed since the occurrence of the failure. A management termination point is set as one end of the path in the aforementioned inter-chassis link port. The first opposing device is configured with an opposing management termination point as the other end of the path. The reachability determination unit determines the reachability based on the reception status of the ETH-CC frame transmitted from the opposing management termination point to the management termination point. A communication device in which the reference time is 3.5 times or more the transmission period of the ETH-CC frame.

4. A communication device interposed between a first opposing device and a second opposing device to realize communication between the first opposing device and the second opposing device, The communication device, together with other communication devices, can establish a multi-chassis link aggregation with each of the first opposing device and the second opposing device. The aforementioned communication device is An inter-chassis link port for communicating with the aforementioned other communication devices, A reachability determination unit that determines the reachability of each of the following paths: a first path from the inter-chassis link port to the first opposing device via the other communication device, and a second path from the inter-chassis link port to the second opposing device via the other communication device. The system includes a transfer control unit that determines the destination of the frame transmitted from the inter-chassis link port, A communication device wherein the transfer control unit transmits the frame to the first opposing device when the reachability determination unit detects a fault in the first path, and transmits the frame from the inter-chassis link port to the second opposing device when the reachability determination unit does not detect a fault in the first path and detects a fault in the second path.

5. A communication control method performed in a communication system that interposes between a first opposing device and a second opposing device to realize communication between the first opposing device and the second opposing device, The communication system includes a first communication device and a second communication device that establish a multi-chassis link aggregation with each of the first opposing device and the second opposing device, The first communication device includes an inter-chassis link port connected to the second communication device. The aforementioned communication control method is: The first communication device performs a liveness check on the path from the inter-chassis link port through the second communication device to the first opposing device, The first communication device includes the step of, if the first communication device detects a fault in the path, transmitting a frame from the inter-chassis link port to the first opposing device, The second communication device includes a port connected to the first opposing device, The aforementioned communication control method is: If the first communication device does not detect any fault in the path, the first communication device transmits a frame from the inter-chassis link port to the second opposing device. If the second communication device detects a failure in communication with the first opposing device via the port, the second communication device further includes the step of transmitting a frame from the second opposing device to the first communication device after a reference time has elapsed since the occurrence of the failure. A management termination point is set as one end of the path in the aforementioned inter-chassis link port. The first opposing device is configured with an opposing management termination point as the other end of the path. The aforementioned communication control method is: The first communication device further includes the step of performing the liveness check based on the reception status of the ETH-CC frame transmitted from the opposing management termination point to the management termination point, A communication control method wherein the reference time is 3.5 times or more the transmission period of the ETH-CC frame.

6. A communication control method performed in a communication system that interposes between a first opposing device and a second opposing device to realize communication between the first opposing device and the second opposing device, The communication system includes a first communication device and a second communication device that establish a multi-chassis link aggregation with each of the first opposing device and the second opposing device, The first communication device includes an inter-chassis link port connected to the second communication device. The aforementioned communication control method is: The first communication device performs a liveness check on each of the first path from the inter-chassis link port through the second communication device to the first opposing device, and the second path from the inter-chassis link port through the second communication device to the second opposing device. If the first communication device detects a fault in the first path, the first communication device transmits a frame from the inter-chassis link port to the first opposing device. A communication control method comprising the step of: if the first communication device does not detect a fault in the first path and detects a fault in the second path, the first communication device transmits a frame from the inter-chassis link port to the second opposing device.

Images