JP6499624B2 - Network device and frame transfer method - Google Patents

Description

  The present disclosure relates to a network device and a frame transfer method that realize both device redundancy and path redundancy.

  Network systems using Layer 2 technology in an OSI (Open Systems Interconnection) reference model, represented by Ethernet (registered trademark) technology, are widely used. In the network system, there is a Link Aggregation Group (LAG) technique as a technique for improving reliability and expanding capacity. In this method, apparatuses are connected by a plurality of physical links, and frame transfer is realized by using a plurality of physical links as one logical link.

  When a plurality of physical links are bundled and used as one logical link, ideally, the bandwidth available for communication can be expanded by the number of physical links. For this reason, it is possible to expand the transfer capacity by applying LAG. Furthermore, communication can be continued if there are one or more physical links in which a plurality of physical link failures have not occurred. For this reason, it is possible to provide a path redundancy function by applying LAG.

  However, the number of devices that terminate the LAG must be one for both the transmission side device and the reception side device. For this reason, if a failure occurs in either the transmission side device or the reception side device, communication cannot be continued, and the device redundancy function cannot be provided.

  As a related technique for solving such a problem, there is a network system to which a technique called multi-chassis LAG (MC-LAG) is applied (for example, Non-Patent Document 1).

  As shown in FIG. 1, this is composed of a switch (hereinafter referred to as SW) 1 and SW2, which are network devices that realize device redundancy with a set of a plurality of devices, and physically connects between SW1 and SW2. It is a connected network system. The physically connected interface is called a network device connection port. In the example of FIG. 1, the port P4 of SW1 and the port P4 of SW2 correspond to network device connection ports, respectively.

  The opposing device 3 to the opposing device 5 can be connected to each of SW1 and SW2, or one of them. A port that connects SW1 and SW2 to the opposing device is called an opposing device connection port. In the example of FIG. 1, the ports P1 to P3 of SW1 and the ports P2 and P3 of SW2 correspond to the counter device connection ports, respectively.

  The two devices SW1 and SW2 behave virtually as one device with respect to the opposing devices 3 to 5. Even if one of the network devices fails, the opposing device physically connected to both network devices, that is, the opposing device 3 and the opposing device 4 in the example of FIG. Can be used to continue communication.

  By applying the multi-chassis LAG technology, it is possible to construct a network that simultaneously realizes path redundancy and device redundancy. However, in a configuration in which the multi-chassis LAG is applied, in order to suppress double distribution, one of the ports of the network device that configures the multi-chassis LAG is blocked, for example, from a part of the switches that configure the multi-chassis LAG. It was necessary to prohibit sending frames.

  The reason why the port must be blocked will be described with reference to FIG. FIG. 2 is a network system comprising SW1 and SW2 to which a multi-chassis LAG function is applied and opposing network devices 3 to 5 capable of directly communicating with a network device group having a multi-chassis LAG configuration. The port P4 of SW1 and the port P4 of SW2 connect two network devices to realize the multi-chassis LAG function, and the port P1 and port P2 of the opposite device 3 are the port P3 of SW1 and the port P3 of SW2 The port P1 and the port P2 of the opposite device 4 are physically connected to the port P2 of the SW1, and the port P2 of the SW2, and the port P1 of the opposite device 5 is connected to the port P1 of the SW1.

  The logical connection configuration of the network system is virtually regarded as one port (LAG1) by applying the LAG function to the port P1 and the port P2 of the opposite device 3, and the port P1 and the port P2 of the opposite device 4 are connected to each other. By applying the LAG function, it is virtually regarded as one port (LAG2), and by applying the multi-chassis LAG function to the port P3 of SW1 and the port P3 of SW2, one port of one device is virtually In this configuration (MC-LAG1), the multichassis LAG function is applied to the port P2 of SW1 and the port P2 of SW2 to virtually consider it as one port of one device (MC-LAG2).

Further, in the network device group in the network system, SW1 is set to transmit / receive a frame to / from the opposite device, and SW2 can receive a frame from the opposite device, but SW1 is used to transmit the frame. Suppose that you need to go through.

  Here, consider a case where the opposite apparatus 3 transmits a frame. FIG. 2 shows an example of transmitting a frame in the network system of FIG. Here, it is assumed that the frame is transferred from the port P2 of the opposite device 3 to the port P3 of the SW2 in accordance with the transmission rule in the LAG setting of the opposite device 3. Since SW2 cannot directly transfer a frame to the opposite device, all frames received by SW2 are transferred to SW1 that can be directly transmitted to the opposite device.

  If a frame is transmitted from the port P2 of SW2 to the port P2 of the opposite device 4 here, a frame having the same information as a frame from SW1 described later is transferred to the opposite device 4. Since the opposite device 4 virtually considers it as one port by applying the LAG function to the ports P1 and P2, it receives two frames having the same information from the same virtual port twice, Become.

  SW1 is a setting capable of transmitting a frame to the opposite device. Therefore, SW1 transfers the frame received from SW2 to the opposite device. When a frame is transmitted to the port P3 of SW1, the frame transmitted from the opposite device 3 returns to the opposite device 3 via SW2 and SW1, and a loop failure may occur. For this reason, when a frame is transferred from the port P4 of SW2 to the port P4 of SW1, it is necessary to add “port received at SW2” information. In the case of the example of FIG. 2, information such as “received at the port P3 of SW2” is given. Alternatively, information such as “received by MC-LAG2” may be added as logical port information instead of physical port information. Thus, SW1 can be set not to transmit a frame to SW1 port P3 forming a pair of SW2 port P3 and multichassis LAG, and a broadcast storm due to loop formation can be avoided.

  In this way, when the multi-chassis LAG technology is applied, a configuration in which only one network device can transmit a frame to the opposite device is referred to as an “Act / Stb configuration”. The related art has realized a network system that realizes both path redundancy and device redundancy by applying a multi-chassis LAG of Act / Stb configuration.

  Since the related technology corresponds to the connection methods of all opposing devices, only the Act / Stb configuration can be applied. That is, the Act / Stb configuration is applied in order to share information of all frames transferred to the opposite device and to suppress double delivery among all devices constituting the multi-chassis LAG. On the other hand, with the multi-chassis LAG technology of Act / Stb configuration, it is not possible to transmit a frame from a part of network devices of the network device group constituting the multi-chassis LAG to the opposite device, and the bandwidth utilization efficiency of the network system There is a problem that decreases.

  The problem that only the Act / Stb configuration can be applied is solved by, for example, using a high-layer device (router) in the OSI reference model and exchanging transfer information between network devices that realize device redundancy. It is possible. However, the network device corresponding to the high layer has a problem that the device configuration becomes complicated.

“100G Ethernet Switch for Telecommunications Carriers”, Hitachi Metals Technical Report Vol. 31, p. 62 (2015) http: // www. hitachi-metals. co. jp / rad / pdf / 2015 / vol31_s15. pdf

  In view of this, the present disclosure provides an “ACT / ACT configuration” in which a layer 2 network does not perform dual distribution to an opposing device while enabling frame transmission from all network devices constituting the multichassis LAG to the opposing device. It is intended to be realized only with the device.

  In order to solve the above-described problem, the present disclosure does not transmit a frame in a transmission-side device between network devices that constitute a multichassis LAG based on a failure occurrence state of the network devices that constitute the multichassis LAG. Alternatively, the present disclosure discards a frame received from another network device in a receiving side device between the network devices constituting the multi-chassis LAG based on a failure occurrence state of the network device constituting the multi-chassis LAG. This makes it possible for all devices constituting the multi-chassis LAG to transmit frames to the opposing device, and when all devices constituting the multi-chassis LAG transmit frames to the opposing device. Even if it exists, it becomes possible to realize a network system in which double distribution does not occur.

Specifically, the network device according to the present disclosure is:
It has a link aggregation function that uses multiple physical links as one logical link, is connected to other network devices, and applies the link aggregation function to ports connected to its own device and ports connected to the other network devices. An opposite device connection port that is connected only to the network opposite device and transmits / receives a frame to / from the network opposite device;
A network device connection port that is connected to the other network device, obtains a state of a physical link in the other network device, and bypasses a frame received from the network facing device to the other network device;
A state monitoring unit that monitors the occurrence of a physical link failure between the own device and the other network device with the network facing device;
When the state monitoring unit detects a physical link failure in the own device, the frame received from the network facing device is transmitted from the network device connection port to the other network device, and the state monitoring unit in the own device When a physical link failure is not detected, the frame received from the first network facing device is transferred to a second network facing device different from the first network facing device, and the first network facing A frame transfer unit that does not transfer a frame received from a device to the second network-facing device via the other network device;
It is characterized by having.

Specifically, the frame transfer method according to the present disclosure is:
It has a link aggregation function that uses multiple physical links as one logical link, is connected to other network devices, and applies the link aggregation function to ports connected to its own device and ports connected to the other network devices. An opposite device connection port that is connected only to the network opposite device and transmits / receives a frame to / from the network opposite device;
A network device connection port that is connected to the other network device, obtains a state of a physical link in the other network device, and bypasses a frame received from the network facing device to the other network device;
A frame transfer method for a network device comprising:
A state monitoring unit monitoring the occurrence of a physical link failure between the own device and the other network device with the network facing device;
When the frame transfer unit detects a failure of the physical link in its own device in the state monitoring unit, it transmits the frame received from the network facing device to the other network device from the network device connection port, and the state monitoring When the unit does not detect a physical link failure in its own device, the frame received from the first network opposite device is transferred to a second network opposite device different from the first network opposite device, A frame transfer step that does not transfer a frame received from the first network facing device to the second network facing device via the other network device;
Have

  According to the present disclosure, an “Act / Act configuration” that enables frame transmission from all network devices constituting the multi-chassis LAG to the opposite device but does not perform double distribution to the opposite device is a layer 2 network. It can be realized only by the apparatus.

The figure showing the network system to which the multi-chassis LAG of Act / Stb structure of related technology is applied is shown. 2 shows an example of a frame transfer path in the network system of FIG. The figure showing the network system to which the multi-chassis LAG of Act / Act structure is applied is shown. 4 shows an example of a frame transfer path in the network system of FIG. A function unit of a device that realizes a multi-chassis LAG configuration of Act / Act configuration by discarding a frame to be transferred to a receiving-side network device in a transmitting-side network device based on a failure occurrence state of a network device constituting the multi-chassis LAG. Show. The functional unit of the apparatus which implement | achieves the multi-chassis LAG structure of Act / Act structure by transferring a flame | frame only to the opposing apparatus in a transmission side network apparatus based on the failure occurrence condition of the network apparatus which comprises multi-chassis LAG is shown. . 4 shows an example of a frame transfer path when a failure occurs between a network device and a counter device in the network system of FIG. 4 shows an example of a frame transfer path when a failure occurs between a network device and the network device in the network system of FIG. FIG. 4 shows an example of a frame transfer path when a network device fails in the network system of FIG. The functional part of the apparatus which comprises the multi-chassis LAG which implement | achieves Act / Act structure by discarding a frame in the receiving side apparatus based on the fault occurrence condition of the network apparatus which comprises the multi-chassis LAG is shown.

  Hereinafter, embodiments of the present disclosure will be described in detail with reference to the drawings. In addition, this indication is not limited to embodiment shown below. These embodiments are merely examples, and the present disclosure can be implemented in various modifications and improvements based on the knowledge of those skilled in the art. In the present specification and drawings, the same reference numerals denote the same components.

The present embodiment includes a network device that is virtually treated as a single device, and a state monitoring unit that monitors whether there is an abnormality in the connection with the network facing device,
When the state monitoring unit determines that there is no abnormality in the connection, the network connection port does not transmit a frame to another network device forming the device redundancy configuration, and discards the frame in the network device, or
When the state monitoring unit determines that there is no abnormality in the connection, a frame is transmitted from the opposite device connection port. When the state monitoring unit determines that there is an abnormality in the connection with the opposite device, a frame is transmitted from the network device connection port. A transmission port determining unit that selectively determines a frame transmission port; or
When the state monitoring unit determines that there is no abnormality in the connection, the frame is transmitted from the network connection port to another network device forming the device redundancy configuration, and the frame is discarded in the network device that has received the frame. To
By
All of the opposing device connection ports directly transmit / receive frames to / from the opposing device without blocking a part of the port connecting the network device group that is virtually treated as one device and the opposing device. A characteristic network system and a communication apparatus for realizing the network system are provided.

(Embodiment 1)
FIG. 3 shows a schematic diagram of a system configuration according to the present embodiment. The network system according to the present embodiment realizes both path redundancy and device redundancy simultaneously by applying the multi-chassis LAG technology, and from all network devices that realize the multi-chassis LAG configuration to the opposite device. It has an Act / Act configuration that enables frame transmission.

  The physical connection configuration of the network system of FIG. 3 is the same as the physical connection configuration of the network system of FIG. 1 except for the opposing device 5 and the physical connection between SW1 and the opposing device 5. Also, the logical connection configuration of the network system of FIG. 3 is the same as the logical connection configuration of the network system of FIG. 1 except for the opposing device 5 and the logical connection between SW1 and the opposing device 5.

  Also, SW1 and SW2, which are network devices, are set to be capable of directly transmitting and receiving frames to and from the opposite device. In addition, although the figure which connected between the one network apparatus and one opposing apparatus and between the network apparatus and the network apparatus was made into the figure with a single physical line, it is good also as a structure which connected the some physical line. . For example, SW1 and the opposing device may be connected by a LAG consisting of a plurality of physical connections, and SW2 and the opposing device may also be connected by a LAG consisting of a plurality of physical connections, and both may be collectively set as MC-LAG. . Similarly, a LAG configuration including a plurality of physical connections may be adopted between SW1 and SW2.

  Further, the opposing device is not physically limited to one device. For example, a multi-chassis LAG configuration may be applied to two physical devices, and a device group virtually regarded as one device may be used as a counter device.

  FIG. 4 shows an example of a frame transfer path in a state where no failure has occurred in the network system of FIG. The frame transmitted from the port P1 of the opposite apparatus 3 is received at the port P3 of SW1. While the frame is directly transferred to the opposite device 4, it is not transferred to the port P2 of the opposite device via the port P4 and the port P2 of SW2.

  Similarly, the frame received at SW2 is also transferred directly to the opposite device 4. If no failure has occurred between the opposing devices 3 and 4, the frame received by SW2 is not transferred to the opposing device 4 via the port P4 and port P2 of SW1. Here, the failure includes a failure of a transfer path such as a physical link and a port, and a device failure of SW1 and SW2.

  Here, in the present embodiment, in order to realize a multi-chassis LAG configuration, a path connecting a plurality of network devices (between the port P4 of SW1 and the port P4 of SW2 in the example of FIG. 4). In the state where no failure has occurred between the opposing devices 3 and 4, no frame transmission is performed, or even if a frame is transmitted, the receiving side device transfers the frame to the opposing device. Discard without. This makes it possible to configure a network system that does not perform double distribution of frames to the opposite device.

  In the first embodiment, frame transfer is not performed between SW1 and SW2 in a state where no failure has occurred between the opposing devices 3 and 4. Therefore, when there is a counter device connected to only one of the network devices, that is, a device such as the counter device 5 in FIG. 1, a frame is not applied to the counter device connected to only one of the network devices. May not be transferred. Therefore, in the first embodiment, a configuration in which a counter device connected to only one of the network devices is not allowed.

  Note that the management frame for exchanging the mutual device status information between SW1 and SW2 is not subject to frame transfer prohibition. Hereinafter, when simply described as “frame”, it refers to a user data frame exchanged between the opposite device and the opposite device, and the frame for exchanging the device status information is distinguished and described as a “management frame”.

  FIG. 5 shows an example of functional units of SW1 and SW2. SW1 and SW2 are the multi-chassis having the Act / Act configuration shown in FIG. 4 in the network device group constituting the multi-chassis LAG, by discarding the frame in the transmission-side network device when performing frame transmission / reception between the network device groups. Realize LAG configuration.

  A network system that realizes a device redundancy configuration with an Act / Act configuration by discarding a frame in the transmission side network device is composed of SW1, SW2 constituting the multi-chassis LAG, the counter device 3 as a counter device of the multi-chassis LAG, and the counter device 4 is provided. The SWx includes opposing device connection ports x1, x2, an inter-SW connection port x3, an MC-LAG state monitoring unit x4, a transmission port determination unit x5, a transmission frame discard determination unit x6, and a frame discard unit x7. The transmission port determination unit x5, the transmission frame discard determination unit x6, and the frame discard unit x7 function as a frame transfer unit. The opposing device y includes SW connection ports y1 and y2. The SW connection ports y1 and y2 are respectively connected to a plurality of network devices having a multi-chassis LAG configuration by the LAG function.

  A frame addressed to the opposite device 4 from the opposite device 3 is transmitted to either SW1 or SW2 based on the LAG distribution function of the opposite device 3. Here, a case where a frame is transmitted from the SW connection port 32 of the opposite device 3 to the opposite device connection port 21 of SW2 will be described as an example.

  The transmission port determination unit 25 transfers the frame addressed to the opposing device 4 received at the opposing device connection port 21 of SW2 to the opposing device connection port 22 according to the frame destination information. This function may use a transmission port determination rule based on MAC learning in a single chassis configuration. In the multi-chassis LAG configuration of Act / Act configuration, the frame is transferred to the transmission frame discard determination unit 26 at the same time as the frame is transferred to the opposite device connection port. This is for determining whether a frame is transmitted from the inter-SW connection port 23 to the SW1.

  The transmission frame discard determination unit 26 transmits a frame to SW1 via the inter-SW connection port 23 according to the MC-LAG state held in the MC-LAG state monitoring unit 24, or the frame discard unit 27 To decide whether to discard the frame.

  The MC-LAG state monitoring units 14 and 24 monitor the occurrence of a failure between the opposing devices 3 and 4 in SW1 and SW2. For example, the MC-LAG state monitoring units 14 and 24 confirm the communication with the opposite device and the communication with the device that is a set of the multi-chassis LAG. It may be determined whether a failure occurs with the network facing device based on the link state of the port, that is, whether the link is up or down. Alternatively, when communication with a network facing device is confirmed using the Ethernet (registered trademark) OAM function, and the monitoring result by the Ethernet OAM is not communicated, between the device configuring the multi-chassis LAG and the facing device It may be determined that an abnormality has occurred.

  The MC-LAG state monitoring unit 14 of SW1 and the MC-LAG state monitoring unit 24 of SW2 monitor the failure occurrence status of each other device and exchange failure occurrence information. As a result, it is possible to share the failure occurrence status of each counter device connection port of SW1 and SW2 and change the frame transfer path according to the failure occurrence status. For the exchange of failure occurrence information, the management frame may be exchanged using the connection path between the inter-SW connection port 13 and the inter-SW connection port 23, or the management connection for managing the state of the multi-chassis LAG configuration A management path may be replaced by providing a separate route.

  If no failure has occurred in any of the connection lines connecting SW1 and SW2 and the opposing device 3 and opposing device 4 constituting the multichassis LAG, no frame is transmitted from the inter-SW connection port. For this reason, the transmission frame discard determination unit 26 sends a frame to the frame discard unit 27 based on information from the MC-LAG state monitoring unit 24 such that “no failure is detected in the opposite device connection port of SW1 and SW2”. Set to transfer.

  On the other hand, when a frame cannot be directly transmitted / received to / from the opposite device, such as when a failure is detected in the opposite device connection port 22, it is necessary to transfer the frame to SW 1 via the inter-SW connection port 23. For this reason, the transmission frame discard determination unit 26 transmits the information from the counter device connection port 22 to the counter device based on the information from the MC-LAG state monitoring unit 24 such that “a failure is detected in the counter device connection port 22”. The frame transfer from the inter-SW connection port 23 to the SW1 is permitted without discarding the frame only for the frame to be performed.

  Thereby, when the multi-chassis LAG is operating normally, the frame is not transmitted from SW2 to SW1, and the frame can be transmitted from SW2 to SW1 only when a failure occurs. As a result, the dual distribution from the SW1 and SW2 network devices to the opposite device is eliminated, and the Act / Act configuration of the multichassis LAG can be realized.

  When a failure is detected in the opposing device connection port of another device that constitutes the multichassis LAG, there is no need to transfer the frame to another network device via the inter-SW connection port. For example, consider a case where a frame is received by SW2 in a state where a failure is detected in the counter device connection port 12. In this case, since it is possible to directly transmit the frame from SW2 to the opposite device, it is not necessary to transfer the frame to SW1 via the inter-SW connection port 23. As shown in FIG. Discard the frame.

  Alternatively, as shown in FIG. 6, the MC-LAG state monitoring unit 24 may notify the transmission port determining unit 25 of the MC-LAG state. In this case, the operation of the transmission port determination unit 25 is different from the operation of the transmission port determination unit in the single chassis configuration. However, there is no need to always transfer frames to the transmission frame discard determination unit 26, and the internal bandwidth of SW2 can be saved. When no failure is detected in the opposite device connection port 22 of SW2, the transmission port determination unit 25 transfers the frame only to the opposite device connection port 22 based on the information from the MC-LAG state monitoring unit 24.

  On the other hand, when a failure is detected in the opposing device connection port 22, the transmission port determination unit 25 transfers the frame to the inter-SW connection port 23 based on information from the MC-LAG state monitoring unit 24. The frame transferred to the inter-SW connection port 23 is limited to the frame to be transmitted from the counter device connection port 22 to the counter device.

  The detour frame transferred from the inter-SW connection port 23 to the inter-SW connection port 13 and to be transmitted from the counter device connection port 22 is transferred to the counter device connection port 12 of SW1 and transmitted to the counter device 4. At this time, if a detour frame is transmitted to the opposite device connection port 11, a loop failure may occur. For this reason, the detour frame transferred from the inter-SW connection port 23 is “a frame received by the counter device connection port 21 and not transferred from the counter device connection port having the multi-chassis LAG configuration with the counter device connection port 21. Or “a frame that should be transmitted from the opposing device connection port 22 and transferred only from the opposing device connection port having the multi-chassis LAG configuration with the opposing device connection port 22” and the inter-SW connection port 13 Need to be sent to.

  7 to 9 show the operation of each network device when a path failure between the network device and the opposite device, a path failure between the network devices, and a device failure of the network device occur. Here, as shown in FIG. 5, the multi-chassis LAG configuration in which the transmission frame discard determination unit x6 is configured to transmit only the frame to be transmitted to the opposite device connection port where the failure has occurred based on the information from the MC-LAG state monitoring unit x4. The operation of the device configuration in the configuration in which the frame device is permitted to transfer the frame to the network device and the other frame is discarded in the frame discard unit x7. As shown in FIG. 6, the frame transfer path outside the apparatus is the same even in the method of determining the transfer destination by the transmission port determining unit x5 based on the information of the MC-LAG state monitoring unit x4.

  FIG. 7 shows an example of frame transfer when a failure occurs between a network device group having a multi-chassis LAG configuration and a counter device in the network system of FIG. Here, a case where a failure occurs between the counter device connection port 22 and the SW connection port 42 will be described.

  First, consider the operation of a frame transmitted from the SW connection port 32 of the opposite device 3 to the opposite device connection port 21 of SW2. As shown in FIG. 7, when a failure is detected at the opposing device connection port 22, the MC-LAG state monitoring unit 24 notifies the transmission frame discarding determination unit 26 of information that a failure has occurred. As a result, the frame to be transmitted from the opposite device connection port 22 transferred to the transmission frame discard determination unit 26 is not transferred to the frame discard unit 27 but transferred to the SW 1 via the inter-SW connection port 23. In SW1, the transmission port is determined based on the destination information of the received frame. If the transmission port cannot be determined from the frame destination information because the transmission destination is not learned, the destination information is learned by transmitting a broadcast frame as in the case of the single chassis configuration. When a frame is transferred from SW2 to SW1, it is possible to avoid a loop failure associated with the transfer to the opposite device 3 by adding the reception port information of SW2.

  Next, consider the operation of a frame transmitted from the SW connection port 31 of the opposite device 3 to the opposite device connection port 11 of SW1. As shown in FIG. 7, even when a failure is detected in the opposing device connection port 22, it is possible to transfer a frame directly from SW <b> 1 to the opposing device 4. For this reason, the frame received by SW 1 is determined to be discarded by the transmission frame discard determination unit 16 and discarded by the frame discard unit 17.

  FIG. 8 shows an example of frame transfer when a failure occurs between network devices having a multichassis LAG configuration in the network system of FIG. In the example of FIG. 8, the MC-LAG state monitoring unit x4 of the network devices SW1 and SW2 notifies the transmission frame discard determination unit x6 that a failure has occurred between SW1 and SW2. However, all frames received from the opposite device 3 can be directly transferred to the opposite device 4 as in the case where no failure has occurred. For this reason, the frame transferred to the transmission frame discard determination unit x6 is discarded by the frame discard unit x7.

  FIG. 9 shows an example of frame transfer when a device failure occurs in any one of the network devices having the multi-chassis LAG configuration in the network system of FIG. Here, a case where a failure occurs in SW2 will be described as an example.

  As shown in FIG. 9, when a device failure occurs in SW2, it is possible to determine that a failure has occurred in the adjacent device in the opposing device 3 and the opposing device 4. In other words, the SW connection port 32 and the SW connection port 42 are expected to stop frame transmission from the relevant port upon failure detection and switch to frame transmission from the SW connection port 31 and SW connection port 42. The

  In SW1, the received frame is transferred from the transmission port determination unit 15 to the opposite device connection port 12 and the transmission frame discard determination unit 16. The MC-LAG state monitoring unit 14 indicates that another network device constituting the multi-chassis LAG has failed, more precisely, another network device constituting the multi-chassis LAG, or the network device. It is possible to determine that a failure has occurred in any of the routes leading to. The MC-LAG state monitoring unit 14 defines the operation of the transmission frame discard determination unit based on the above information, but a failure occurs because the frame cannot be transferred to another network device constituting the multi-chassis LAG. As in the case of not doing so, the frame discarding unit 17 discards the frame.

(Embodiment 2)
FIG. 10 shows an example of a system configuration according to the present embodiment. In the network device group constituting the multi-chassis LAG, the multi-chassis LAG configuration of Act / Act configuration shown in FIG. 4 is realized by discarding the frame in the receiving-side network device when performing frame transmission / reception between the network device groups. The functional part of an apparatus is shown.

  The network system that realizes the device redundancy configuration in the Act / Act configuration by discarding the frame in the receiving-side network device includes the network devices SW1 and SW2 that constitute the multi-chassis LAG, the opposing device 3 that is the opposing device of the multi-chassis LAG, A counter device 4 is provided. SWx includes opposing device connection ports x1 and x2, an inter-SW connection port x3, an MC-LAG state monitoring unit x4, a transmission port determination unit x5, a received frame discard determination unit x6, and a frame discard unit x7. The transmission port determination unit x5, the transmission frame discard determination unit x6, and the frame discard unit x7 function as a frame transfer unit. The opposing device y includes SW connection ports x1 and x2. The SW connection ports x1 and x2 are connected to a plurality of devices having a multi-chassis LAG configuration by the LAG function.

  A frame addressed to the opposite device 4 from the opposite device 3 is transmitted to either SW1 or SW2 based on the LAG distribution function of the opposite device 3. Here, a case where a frame is transmitted from the SW connection port 32 of the opposite device 3 to the opposite device connection port 21 of SW2 will be described as an example.

  The transmission port determination unit 25 transmits the frame addressed to the opposing device 4 received at the opposing device connection port 21 of SW2 to the opposing device connection port 22 according to the frame destination information. At this time, in order to share the received frame information between SW1 and SW2, the frame received at the opposing device connection port 21 is transferred from the inter-SW connection port 23 to SW1.

  The inter-SW connection port 13 of SW1 receives the frame transferred from the SW2. Thereafter, the received frame is sent to the received frame discard determination unit 16. Here, when a failure has not occurred in the network device having the multi-chassis LAG configuration and the path connected to the network device, it is necessary to discard the frame in SW1 in order to avoid double transmission to the opposite device. . For this reason, the received frame discard determination unit 16 transfers the frame to the frame discard unit 17 based on the information of the MC-LAG state monitoring unit 14.

  It should be noted that immediately after receiving the frame at the inter-SW connection port 13, it is desirable that the determination process by the received frame discard determination unit 16 and the frame discard by the frame discard unit 17 are performed. For example, when SW1 and SW2 are chassis type SWs composed of a plurality of line cards, the frames transferred from SW2 to SW1 are discarded on the Ingress side, that is, received frame disposed on the line card having the inter-SW connection port 13 It is desirable that the determination unit 16 and the frame discard unit 17 discard the frame. When the frame is discarded on the Egress side, that is, in the line card having the opposite device connection port 12, the frame received by SW1 via the inter-SW connection port is a route from the inter-SW connection port 13 to the opposite device connection port 12. This increases the possibility of contention with other traffic in the apparatus. Since the frame received at the inter-SW connection port 13 is a frame to be discarded in a state where no failure has occurred, it is necessary to promptly determine whether to transfer or discard after receiving the SW1. .

  In this way, in a state where no failure has occurred, the frame received at the inter-SW connection port 13 of SW1 is discarded by the frame discarding unit of the port. By discarding the frame at the received port, it is possible to realize a multi-chassis LAG configuration without using the bandwidth inside the SW1 device. Although there is a restriction such as “limit the connection with the opposite device to the multi-chassis LAG”, it becomes possible to avoid the device internal band exhaustion due to the frame to be discarded, and transfer from the opposite device connection port 11 to the opposite device connection port 12 It is possible to transmit and receive frames between opposing devices without affecting the frames to be transmitted.

  On the other hand, when the MC-LAG state monitoring unit 14 recognizes that a failure has occurred in the opposite device connection port 22 of SW2, the received frame is expected to be transferred to the opposite device. For this reason, among the frames from SW2 transferred to the received frame discard determination unit 16, the frame transferred to the counter device connection port 22 in the failure non-occurrence state is sent to the counter device connection port 12 constituting the set of multi-chassis LAGs. Forward.

  Thus, when the multichassis LAG is operating normally, the frame transmitted from SW2 to SW1 is discarded at SW1, and when a failure occurs, the frame transmitted from SW2 to SW1 is discarded. Transfer to the opposite device connection port. As a result, the dual distribution from the SW1 and SW2 network devices to the opposite device is eliminated, and the Act / Act configuration of the multichassis LAG can be realized.

  The present disclosure can be applied to the information communication industry.

1, 2: SW
11, 12, 21, 22: Counter apparatus connection port 13, 23: SW connection port 14, 24: MC-LAG state monitoring unit 15, 25: Transmission port determination unit 16, 26: Transmission frame discard determination unit 17, 27 : Frame discard unit 3, 4: Counter device 31, 32, 41, 42: SW connection port

Claims (6)

It has a link aggregation function that uses multiple physical links as one logical link, is connected to other network devices, and applies the link aggregation function to ports connected to its own device and ports connected to the other network devices. An opposite device connection port that is connected only to the network opposite device and transmits / receives a frame to / from the network opposite device;
A network device connection port that is connected to the other network device, obtains a state of a physical link in the other network device, and bypasses a frame received from the network facing device to the other network device;
A state monitoring unit that monitors the occurrence of a physical link failure between the own device and the other network device with the network facing device;
When the state monitoring unit detects a physical link failure in the own device, the frame received from the network facing device is transmitted from the network device connection port to the other network device, and the state monitoring unit in the own device When a physical link failure is not detected, the frame received from the first network facing device is transferred to a second network facing device different from the first network facing device, and the first network facing A frame transfer unit that does not transfer a frame received from a device to the second network-facing device via the other network device;
A network device characterized by comprising: The frame transfer unit
If the state monitoring unit does not detect a physical link failure in its own device, discard the frame to be transmitted from the network device connection port to the other network device;
The network device according to claim 1, wherein: The frame transfer unit
When the state monitoring unit does not detect a physical link failure in its own device, the frame received from the first network facing device is not transferred to the network device connection port.
The network device according to claim 1, wherein: The frame transfer unit
If the state monitoring unit has not detected a physical link failure in the other network device, discard the frame received from the network device connection port;
The network device according to claim 1, wherein: The state monitoring unit monitors connectivity with the network facing device by using the Ethernet OAM function.
The network device according to claim 1, wherein the network device is a device. It has a link aggregation function that uses multiple physical links as one logical link, is connected to other network devices, and applies the link aggregation function to ports connected to its own device and ports connected to the other network devices. An opposite device connection port that is connected only to the network opposite device and transmits / receives a frame to / from the network opposite device;
A network device connection port that is connected to the other network device, obtains a state of a physical link in the other network device, and bypasses a frame received from the network facing device to the other network device;
A frame transfer method for a network device comprising:
A state monitoring unit monitoring the occurrence of a physical link failure between the own device and the other network device with the network facing device;
When the frame transfer unit detects a failure of the physical link in its own device in the state monitoring unit, it transmits the frame received from the network facing device to the other network device from the network device connection port, and the state monitoring When the unit does not detect a physical link failure in its own device, the frame received from the first network opposite device is transferred to a second network opposite device different from the first network opposite device, A frame transfer step that does not transfer a frame received from the first network facing device to the second network facing device via the other network device;
A frame transfer method characterized by comprising:

Images