JP4636247B2 - Packet network and layer 2 switch - Google Patents

Description

  The present invention relates to redundancy of a packet network by layer 2 switches.

FIG. 5 is a system configuration diagram showing a configuration example of a packet network using a conventional layer 2 switch. Referring to FIG. 5, a conventional packet network 90 includes a hub 91, rear layer 2 switches (L2SW) 92 ₁ and 92 ₂ , and front layer 2 switches 93 ₁ and 93 ₂ , and terminals 95 _{1 to} 95 ₄ A server 94 is connected.

The hub 91 connects the server 94 and the subsequent layer 2 switches 92 ₁ and 92 ₂ .

The rear layer 2 switch 92 ₁ connects the hub 91 and the front layer 2 switches 93 ₁ and 93 ₂ . Similarly, the post-stage layer 2 switch 92 ₂ also connects the hub 91 and the pre-stage layer 2 switches 93 ₁ and 93 ₂ .

The upstream layer 2 switch 93 ₁ connects the downstream layer 2 switches 92 ₁ and 92 ₂ to the terminals 95 ₁ and 95 ₂ . The upstream layer 2 switch 93 ₂ connects the downstream layer 2 switches 92 ₁ and 92 ₂ to the terminals 95 ₃ and 95 ₄ .

In the packet network 90 by the layer 2 switch, if the redundant connections as shown in FIG. 5 are made, a broadcast packet loop occurs between the subsequent layer 2 switches 92 ₁ and 92 ₂ and the previous layer 2 switches 93 ₁ and 93 _2. As a result, there is a problem that the band of the switch is compressed and communication cannot be normally performed.

In order to solve this problem, when the redundancy as shown in FIG. 5 is configured in the packet network by the layer 2 switch, the redundancy protocol such as the spanning tree protocol is conventionally used as each layer 2 switch 92 ₁ , 92 ₂ , 93 ₁ , 93. ₂ was implemented. Based on Redundancy Protocol, Layer 2 switches 92 _1, 92 _2, 93 _1, 93 ₂ by exchanging routing information with each other to determine the operational path, it had prevented the occurrence of a loop by blocking the other path . As a result, the packet network 90 is maintained in a state where no loop occurs and communication can be performed normally.

  However, in the method based on the spanning tree protocol, it is necessary to implement a spanning tree on all layer 2 switches. Therefore, when a network with layer 2 switches has already been introduced, it is necessary to replace all the layer 2 switches with devices corresponding to the spanning tree protocol and implement a spanning tree in order to configure redundancy as shown in FIG. It was uneconomical.

  In addition, there is a possibility that the operation of the device corresponding to the spanning tree protocol is slightly different for each vendor. For this reason, even if a device corresponding to the spanning tree protocol is introduced, there is no guarantee that the route determination operation is normally performed if devices of different vendors are employed in the aggregation layers in the former stage and the latter stage.

Further, the Spanning Tree Protocol, was determined operational path by converging to exchange routing information between the layer 2 switches 92 _1, 92 _2, 93 _1, 93 _2. When a failure occurs somewhere in the operational path, was determined detour path by exchanging routing information again between the layer 2 switches 92 _1, 92 _2, 93 _1, 93 _2. Therefore, it takes time to switch the route, and many packets are lost when a failure occurs. Further, since all the ports of the layer 2 switch are blocked from the occurrence of the failure until the detour route is determined, the devices connected to the ports not related to the failure route are also affected.

  In contrast, two routes, the shortest route and the next shortest route, are set in advance between the edges, each of the devices at the edge monitors the route for abnormalities, and if an abnormality is detected, the route is switched. The technique to perform is disclosed (refer patent document 1).

According to this, when a device at the edge detects an abnormality, it switches to a predetermined detour path, so that the path can be switched in a short time. Further, since not all routes are reconfigured, communication on routes that are not related to a failure is continued.
JP 2003-258829 A

  However, in the method described in Patent Document 1, since the devices at both edges switch to each other, as shown in FIG. 5, when configuring path redundancy between the preceding stage and the succeeding stage, both the preceding stage and the succeeding stage are used. It is necessary to use a device having a switching function, and there remains a problem that all devices need to be replaced when a redundant configuration is introduced.

  Further, since the routing protocol of layer 3 called OSPF (Open Shortest Path First) is used to set the shortest route and the next shortest route, it is necessary to replace the device with an OSPF.

  An object of the present invention is to provide a packet network using a layer 2 switch that can be constructed with little replacement from existing equipment and can switch path redundancy in a short time.

In order to achieve the above object, the packet network of the present invention provides:
At least two subsequent layer 2 switches connected to each of the operation path and the backup path constituting redundancy;
Packets can be transmitted / received to / from both the operation path and the backup path, and can be transmitted / received to / from the latter layer 2 switch connected to the operation path and the latter layer 2 switch connected to the backup path. A first layer 2 switch that transmits and receives the packet on the operating path and monitors the operating path, and detects a failure of the operating path and switches transmission and reception of the packet to the backup path ;
The preceding layer 2 switch generates a monitoring packet that simulates a packet from a transmission source to a transmission destination using an address acquired from a packet to be switched, transmits the monitoring packet to the operation route, and monitors a response to the monitoring packet To monitor the operation route .

  According to the present invention, the upstream layer 2 switch normally forwards the packet along the operational route and monitors the operational route, and switches the packet forwarding to the backup route when a failure is detected. Thus, it is possible to provide a packet network capable of switching a route in a short time by simple switching control without providing a special function.

  Embodiments for carrying out the present invention will be described in detail with reference to the drawings.

FIG. 1 is a system configuration diagram illustrating a configuration example of a packet network using a layer 2 switch according to the present embodiment. Referring to FIG. 1, a packet network 10 includes a hub 11, rear layer 2 switches (L2SW) 12 ₁ and 12 ₂ , and front layer 2 switches 13 ₁ and 13 ₂ , and terminals 15 ₁ to 15 ₄ and a server 14. Connect.

The hub 11 connects the server 14 and the subsequent layer 2 switches 12 ₁ and 12 ₂ .

In the redundant configuration, the rear layer 2 switch 12 ₁ operates as an operation device, and the rear layer 2 switch 12 ₂ operates as a spare device.

The rear layer 2 switch 12 ₁ connects the hub 11 and the front layer 2 switches 13 ₁ and 13 ₂ . Similarly, the post-stage layer 2 switch 12 ₂ also connects the hub 11 and the pre-stage layer 2 switches 13 ₁ and 13 ₂ . With this configuration, the subsequent layer 2 switch 12 ₁ normally switches packets between the server 14 and the terminals 15 ₁ , 15 ₂ , 15 ₃ , 15 ₄ . When there is a failure in the active system path, the subsequent layer 2 switch 12 ₂ performs the switching.

The upstream layer 2 switch 13 ₁ has an operation port 131 ₁ and a spare port 132 ₁ . The operation port 131 ₁ is connected to the rear layer 2 switch 12 ₁ that is the operation device, and the spare port 132 ₁ is connected to the rear layer 2 switch 12 ₂ that is the spare device. Further, terminals 15 ₁ and 15 ₂ are connected to the preceding layer 2 switch 13 ₁ .

Front layer 2 switch 13 ₁ is normal, by using the operational port 131 _1, and receive packets between the server 14 and the terminal 15 _1, 15 _2, also monitors the operational path in an operational port 131 _1. When the upstream layer 2 switch 13 ₁ detects a failure in the operation path, the upstream layer 2 switch 13 ₁ switches the packet transmission / reception to the spare port 132 ₁ .

The upstream layer 2 switch 13 ₂ has an operation port 131 ₂ and a spare port 132 ₂ . The operation port 131 ₂ is connected to the rear layer 2 switch 12 ₁ that is the operation device, and the spare port 132 ₂ is connected to the rear layer 2 switch 12 ₂ that is the spare device. Further, terminals 15 ₃ and 15 ₄ are connected to the preceding layer 2 switch 13 ₂ .

Front layer 2 switch 13 ₂ is normal, using the operational port 131 _2, and receive packets between the server 14 and the terminal 15 _3, 15 _4, also monitors the operational path in an operational port 131 _2. When the upstream layer 2 switch 13 ₂ detects a failure in the operation path, the upstream layer 2 switch 13 ₂ switches packet transmission / reception to the spare port 132 ₂ .

FIG. 2 is a block diagram showing the configuration of the previous layer 2 switch. The preceding layer 2 switches 13 ₁ and 13 ₂ in FIG. 1 have the same configuration, and in FIG. 2, this is shown as the preceding layer 2 switch 13. Referring to FIG. 2, the layer 2 switch 13 includes an operation port 131, a spare port 132, a port 133, a path failure detection circuit 134, and a MAC table control circuit 135. Operation port 131 is connected to the subsequent stage the layer 2 switch 12 ₁ is operated device. The spare port 132 is connected to the subsequent layer 2 switch 12 ₂ which is a spare device. A port 133 is connected to the terminal 15.

  The MAC table control unit 135 manages a MAC table used for packet switching in the upstream layer 2 switch 13. The MAC table is a table showing a correspondence between a MAC address and a transfer destination port of a packet having the MAC address.

  As an example, the MAC table has a region of the operation port 131 and the spare port 132, and has a configuration in which the MAC address of the packet switched by the preceding layer 2 switch 13 is recorded. The MAC address of the packet switched by the upstream layer 2 switch 13 is obtained by learning.

  An enable flag is provided for each MAC address in the area of each port, and a transfer destination port is indicated by turning on one of the enable flags of the operation port 131 and the spare port 132 for each MAC address. An IP address corresponding to the MAC address is also recorded in the MAC table.

  The MAC table control circuit 135 performs path switching by switching on and off the enable flags of the operation port 131 and the spare port 132 under the control of the path failure detection circuit 134.

  The path failure detection circuit 134 periodically transmits a path state monitoring packet to the server 14 through the operation port 131 and monitors the response thereto to confirm the normality of the operation path. The route state monitoring packet is, for example, a ping (Packet Internet Grouper) packet.

  At that time, the path failure detection circuit 134 obtains a MAC address and an IP address from the packet switched by the previous layer 2 switch 13, and uses the MAC address and the IP address to simulate a packet from the terminal 15 to the server 14. A route state monitoring packet is created and periodically transmitted from the operation port 131. Then, when there is no response from the server 14, the path failure detection circuit 134 determines that a failure has occurred in the operation path, and instructs the MAC table control circuit 135 to switch to the backup path.

As an example here, the terminal 15, from the state management port 131 of the front layer 2 switch 13, and through the subsequent Layer 2 switch 12 ₁ is operated device is carried out server 14 and packet communication, failure in the operation path A description will be given of the operation when a switch occurs to the backup route when a problem occurs.

  The path failure detection circuit 134 of the upstream layer 2 switch 13 holds the MAC address and IP address of the terminal 15 and the MAC address and IP address of the server 14 acquired at the time of switching. At this time, since the operation route is used, the operation port 131 is enabled for all the MAC addresses in the MAC table as shown in FIG.

  The path failure detection circuit 134 of the upstream layer 2 switch 13 transmits a path state monitoring packet using the MAC address and IP address of the terminal 15 to the server 14, thereby causing a path failure between the upstream layer 2 switch 13 and the server 14. To monitor.

  When the path fault detection circuit 134 detects a path fault, the path fault detection circuit 134 instructs the MAC table control circuit 135 to switch paths. Here, it is assumed that the MAC address of the terminal 15 is indicated as “AAAA” in FIG.

Upon receiving the path switching instruction, the MAC table control circuit 135 synchronizes the MAC table of the operational port and the MAC table of the spare port to turn on / off the enable flag of the entry of the MAC address “AAAA”, as shown in FIG. Swap as follows. Thus, the packet from the terminal 15 to the server 14 is to be transmitted from the spare port 132 at the front stage Layer 2 switch 13 at the subsequent stage the layer 2 switch 12 _2. Since the post-stage layer 2 switch 12 ₂ transmits the packet to the server 14 and learns the MAC address and IP address of the terminal 15, the packet from the server 14 to the terminal 15 also communicates with the backup path.

  As described above, according to the present embodiment, the operation route and the protection route are set in advance, and the upstream layer 2 switch 13 normally forwards the packet along the operation route and monitors the operation route to generate a failure. When the packet transfer is detected, the packet transfer is switched to the backup path. Therefore, it is possible to provide a packet network capable of switching the path in a short time by simple switching control without providing a special function in the subsequent layer 2 switch 12. In addition, since the previous layer 2 switch 13 that has detected the failure switches only the route in which the failure is detected to the backup route, there is little influence on the terminals connected to other ports.

  In addition, according to the present embodiment, the upstream layer 2 switch 13 is provided with a MAC table for each port connected to the operation route and the backup route. Can be switched.

  Further, according to the present embodiment, since the upstream layer 2 switch 13 switches the operation path and the backup path in synchronization, packet loss due to switching can be reduced.

  In the present embodiment, the configuration having one backup route for one operation route is exemplified, but the present invention is not limited to this. For example, a redundant configuration with higher reliability may be provided by providing two standby paths for one operational path. In that case, for example, two subsequent layer 2 switches used as spare devices may be provided for one subsequent layer 2 switch used as the operation device.

It is a system configuration diagram showing an example of the configuration of a packet network by the layer 2 switch of the present embodiment. It is a block diagram which shows the structure of a front | former stage layer 2 switch. It is a figure which shows an example of the MAC table at the normal time. It is a figure which shows an example of the MAC table at the time of path | route switching. It is a system configuration | structure figure which shows the structural example of the packet network by the conventional layer 2 switch.

Explanation of symbols

10 packet network 11 the hub 12 _1, 12 ₂ subsequent Layer 2 switch 13 _1, 13 ₂ front layer 2 switch 14 server 15 _{1-15 4} terminal 131 operating ports 132 spare port 133 port 134 path failure detecting circuit 135 MAC table control circuit

Claims (2)

At least two subsequent layer 2 switches connected to each of the operation path and the backup path constituting redundancy;
Packets can be transmitted / received to / from both the operation path and the backup path, and can be transmitted / received to / from the latter layer 2 switch connected to the operation path and the latter layer 2 switch connected to the backup path. in operation path monitors the該運for path transmits and receives the packet, detects a failure of該運for route, possess the front layer 2 switch for switching transmission and reception of the packet to the spare path,
The preceding layer 2 switch generates a monitoring packet that simulates a packet from a transmission source to a transmission destination using an address acquired from a packet to be switched, transmits the monitoring packet to the operation route, and monitors a response to the monitoring packet The operational route is monitored by
Packet network. A layer 2 switch that is connected to a redundant operation path and a backup path and is capable of transmitting and receiving packets through the operation path and the backup path;
An operational port connected to the operational path and transmitting and receiving packets at normal times;
A spare port connected to the spare path and transmitting and receiving packets when the operational path fails; and
A path failure detection unit that monitors the operating path at the operating port;
When a failure of the operational path is detected by said path failure detecting unit, have a switching control unit for switching transmission and reception of packets from the operating port to the spare port,
The path failure detection unit generates a monitoring packet that simulates a packet from a transmission source to a transmission destination using an address acquired from a packet to be switched, transmits the monitoring packet to the operation path, and monitors a response to the monitoring packet To monitor the operational route,
Layer 2 switch.

Images