JP5360257B2 - Frame relay apparatus, route learning program, and route learning method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To discriminate between communication in a subnet and communication between subnets while suppressing increase in cost required for configuring a network. <P>SOLUTION: A flooding processing part 131 adds a flag '0' to a frame to be flooded in the same subnet as a transmission source of the frame, and adds a flag '1' to a frame to be flooded beyond the subnet. A flag determination part 132 outputs the frame with the flag '0' to a transmission source registration part 126-1, and outputs the frame with the flag '1' to a band controller 124-2. The transmission source registration part 126-1 makes a learning table 127-1 learn the correspondence between a transmission source address of the frame and an input port together with the flag '0'. A transmission source registration part 126-2 makes a learning table 127-2 learn the correspondence between a transmission source address of the frame after band control by the band controller 124-2 and an input port together with the flag '1'. <P>COPYRIGHT: (C)2012,JPO&amp;INPIT

Description

  The present invention relates to a frame relay device, a route learning program, and a route learning method, and in particular, a frame capable of distinguishing between intra-subnet communication and inter-subnet communication while suppressing an increase in cost required for network configuration. The present invention relates to a relay device, a route learning program, and a route learning method.

  In recent years, for example, a plurality of communication terminal apparatuses installed in a company or the like are generally connected to each other so as to be communicable to form a LAN (Local Area Network). Further, communication terminal devices constituting the LAN may be logically grouped regardless of physical connection status, and a plurality of VLANs (Virtual LAN) may be formed in the LAN. By forming a VLAN, communication terminal devices belonging to the same VLAN can communicate with each other like communication terminal devices belonging to one LAN, and communication terminal devices belonging to different VLANs can be physically They can be distinguished from each other regardless of the connection status.

  When configuring such a VLAN, for example, in Patent Document 1, an address conversion device that converts a private IP address uniquely used in each VLAN and a global IP address commonly used outside the VLAN is used. By providing, it is disclosed to efficiently route frames within and between VLANs. In particular, in Patent Document 1, by switching a table used for address conversion for each VLAN, address conversion corresponding to a plurality of VLANs can be performed by one address conversion device.

  By the way, in the VLAN, communication terminal devices can be further logically grouped to form a plurality of subnets. These subnets may be connected via an L2 switch (layer 2 switch). That is, for example, as shown in FIG. 6, when the VLAN is divided into subnet # 1 and subnet # 2, the L2 switch 10 is arranged in the subnet # 1, and the L2 switch 20 is arranged in the subnet # 2, the L2 switch By connecting 10 and L2 switch 20 to each other, subnet # 1 and subnet # 2 are connected.

  In the VLAN configuration shown in FIG. 6, when a communication terminal device belonging to subnet # 1 transmits a frame to a communication terminal device belonging to subnet # 2, this frame is relayed by L2 switch 10 and L2 switch 20. . As described above, since a frame across the subnet is always transmitted through the route between the L2 switch 10 and the L2 switch 20, for example, different fee systems are applied to communication within the subnet and communication between the subnets. In some cases, it is possible to distinguish between intra-subnet communication and inter-subnet communication depending on whether or not the frame is transmitted through the path between the L2 switch 10 and the L2 switch 20.

JP 2005-20170 A

  However, when L2 switches are arranged in each of a plurality of subnets, the ports of the respective L2 switches are consumed to connect the L2 switches arranged in each subnet to each other, and a limited number of ports are wasted. There is a problem of end up.

  Specifically, in the VLAN configuration shown in FIG. 6, since the L2 switch 10 and the L2 switch 20 are connected to each other, one L2 switch secures one port for connection to the other L2 switch. Must. For this reason, one port that can be connected to the communication terminal device in the subnet is wasted, and in order to secure ports to be connected to many communication terminal devices, it is necessary to arrange more L2 switches in one subnet. Arise. As a result, the cost for configuring the VLAN increases.

  In the VLAN configuration shown in FIG. 6, since L2 switches are arranged for each subnet, if the number of subnets increases, the number of required L2 switches also increases, and the cost for configuring the VLAN increases. Further increase.

  On the other hand, a VLAN configuration in which a plurality of subnets are connected via a single L2 switch can suppress an increase in cost due to an increase in the number of L2 switch ports and an increase in the number of L2 switches. Communication within a subnet and communication between subnets cannot be distinguished by a route. That is, in the VLAN configuration shown in FIG. 6, it is possible to determine the type of communication based on the presence / absence of frame transmission on the path between the L2 switch 10 and the L2 switch 20, but 1 between two subnets. In a configuration in which two L2 switches are arranged, a frame is relayed by the L2 switch in both intra-subnet communication and inter-subnet communication, and the two cannot be distinguished.

  The present invention has been made in view of such a point, and a frame relay apparatus and a path learning program capable of distinguishing between intra-subnet communication and inter-subnet communication while suppressing an increase in cost required for network configuration An object is to provide a route learning method.

  In order to solve the above problems, a frame relay device according to the present invention includes an acquisition unit that acquires a frame from an input port connected to a first subnet, and an output port that corresponds to the destination of the frame acquired by the acquisition unit A first transmission means for transmitting a frame to the first subnet via the output port when the output port of the frame set by the setting means is connected to the first subnet; Second transmission means for transmitting a frame to the second subnet via the output port when the output port of the frame set by the setting means is connected to a second subnet different from the first subnet; The structure which has is taken.

  According to this configuration, since transmission of frames within the same subnet and transmission of frames across subnets are performed via different paths in the frame relay device, two subnets are connected by one frame relay device. However, communication within a subnet and communication between subnets can be distinguished. In other words, it is possible to distinguish between communication within a subnet and communication between subnets while suppressing an increase in cost required for the configuration of the network.

  The frame relay apparatus according to the present invention includes an acquisition unit that acquires a frame from an input port connected to the first subnet, and a setting unit that sets an output port corresponding to the destination of the frame acquired by the acquisition unit. Adding means for adding to the frame a flag indicating whether or not the output port of the frame set by the setting means is connected to a second subnet different from the first subnet; and the frame acquired by the acquiring means And a registration means for registering a flag added to the frame by the adding means together with a transmission source address and an input port in a learning table for storing a correspondence relationship between the address and the port.

  According to this configuration, the flag for distinguishing between the transmission path of the frame within the same subnet and the transmission path of the frame across the subnet is registered in the learning table together with the correspondence relationship between the address and the port. When newly determining the output port of the frame, it can be determined from the flag registered corresponding to the output port whether or not the frame is transmitted across the subnet. Therefore, even if two subnets are connected by one frame relay device, communication within the subnet and communication between subnets can be distinguished. In other words, it is possible to distinguish between communication within a subnet and communication between subnets while suppressing an increase in cost required for the configuration of the network.

  In the frame relay device according to the present invention, in the configuration described above, the setting unit includes a port corresponding to the destination address of the frame in the learning table when the destination address of the frame has been registered in the learning table by the registration unit. Is set as the output port of the frame.

  According to this configuration, when the destination address of the frame has already been learned, the learned port is used as the output port, so that the frame is sent to the desired destination address without performing flooding to output the frame from all ports. Can be transmitted.

  The frame relay apparatus according to the present invention employs a configuration in which, in the above configuration, the adding unit adds a flag stored in the learning table to the frame corresponding to the output port set by the setting unit.

  According to this configuration, the flag stored at the time of learning the correspondence relationship between the address and the port is added to the frame, so whether or not the frame is transmitted across the subnet is easily distinguished from the flag added to the flag. be able to.

  In the frame relay device according to the present invention, in the above configuration, the setting unit outputs all ports other than the input port of the frame as output ports when the destination address of the frame is not registered in the learning table by the registration unit. The configuration to be set as

  According to this configuration, when the destination address of the frame is unlearned, the flooding that outputs the frame from all ports is performed, so that the frame can be reliably transmitted to the destination address.

  In the frame relay device according to the present invention, in the configuration described above, the adding unit replicates a frame in accordance with the output port set by the setting unit, and the first of the obtained plurality of replicated frames. A flag indicating connection to the first subnet is added to the duplicate frame in which the output port corresponding to the subnet is set, and the second subnet is assigned to the duplicate frame in which the output port corresponding to the second subnet is set. A configuration is adopted in which a flag indicating that the connection is made is added.

  According to this configuration, since the flag corresponding to the subnet corresponding to the output port at the time of flooding is added to the frame, whether or not the frame is transmitted across the subnet is easily distinguished from the flag added to the flag. be able to. As a result, it is possible to learn whether or not the transmission path toward the frame source address crosses the subnet.

  The frame relay apparatus according to the present invention is suitable for the second subnet when the flag added to the frame by the adding means indicates that the output port of the frame is connected to the second subnet. The bandwidth control means for controlling the received bandwidth and the output means for outputting the frame whose bandwidth is controlled by the bandwidth control means from the output port are employed.

  According to this configuration, for frames transmitted to different subnets, band control is performed before output, so when frames are transmitted across subnets, they are suitable for the subnet connected to the output port. Band frames can be output.

  The route learning program according to the present invention is a route learning program executed by a computer, wherein the computer obtains a frame from an input port connected to a first subnet, and the acquisition step includes: A setting step of setting an output port corresponding to the destination of the acquired frame, and whether or not the output port of the frame set in the setting step is connected to a second subnet different from the first subnet An addition step of adding a flag to the frame, and learning of storing the correspondence relationship between the address and port of the flag added to the frame in the addition step together with the transmission source address and input port of the frame acquired in the acquisition step And a registration step of registering in the table.

  The route learning method according to the present invention includes an acquisition step of acquiring a frame from an input port connected to the first subnet, and a setting step of setting an output port corresponding to the destination of the frame acquired in the acquisition step An addition step of adding to the frame a flag indicating whether or not the output port of the frame set in the setting step is connected to a second subnet different from the first subnet, and acquiring in the acquisition step And a registration step of registering the flag added to the frame in the addition step together with the transmission source address and input port of the received frame in a learning table that stores the correspondence relationship between the address and the port.

  According to these, in order to register the flag for distinguishing between the transmission path of the frame within the same subnet and the transmission path of the frame across the subnet, together with the correspondence relationship between the address and the port, the learning table is referred to hereinafter. When newly determining an output port of a frame, whether or not this frame is transmitted across a subnet can be determined from a flag registered corresponding to the output port. Therefore, even if two subnets are connected by one frame relay device, communication within the subnet and communication between subnets can be distinguished. In other words, it is possible to distinguish between communication within a subnet and communication between subnets while suppressing an increase in cost required for the configuration of the network.

  According to the frame relay device, the route learning program, and the route learning method disclosed in the present specification, it is possible to distinguish between communication within a subnet and communication between subnets while suppressing an increase in cost required for network configuration. Can do.

It is a figure which shows an example of the VLAN structure which concerns on one embodiment. It is a block diagram which shows the principal part structure of the L2 switch which concerns on one Embodiment. It is a figure which shows an example of the port table which concerns on one embodiment. It is a figure which shows an example of the learning table which concerns on one embodiment. It is a flowchart which shows operation | movement of L2 switch which concerns on one Embodiment. It is a figure which shows an example of a VLAN structure.

  In an embodiment of the present invention, a frame relay device that relays a frame, when a frame corresponding to a destination address has not yet been learned, and when the frame is flooded, the frame output from each port is a subnet. A flag indicating whether or not the frame is flooded across the frame is added to the frame, and the port corresponding to the destination address is learned including the flag added to the frame. Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. In the following description, an L2 switch is taken as an example of the frame relay device.

  FIG. 1 is a diagram showing an example of a VLAN configuration according to an embodiment of the present invention. As shown in the figure, in this embodiment, the VLAN is divided into subnet # 1 and subnet # 2, and these two subnets are connected only via the L2 switch 100. Therefore, a frame transmitted from a communication terminal device belonging to subnet # 1 always passes through L2 switch 100 regardless of whether the destination address is an address in subnet # 1 or an address in subnet # 2. Will be transferred. Similarly, a frame transmitted from a communication terminal device belonging to subnet # 2 is always transferred via the L2 switch 100.

  FIG. 2 is a block diagram showing a main configuration of the L2 switch 100 according to the present embodiment. The L2 switch 100 shown in the figure is roughly divided into input / output port units 110-1 and 110-2, domain processing units 120-1 and 120-2, and transfer units corresponding to the subnet # 1 and the subnet # 2, respectively. 130-1 and 130-2. The input / output port unit 110-2, the domain processing unit 120-2, and the transfer unit 130-2 are the input / output port unit 110-1 and the domain processing unit 120-, respectively, except that the corresponding subnet is the subnet # 2. 1 and the transfer unit 130-1, only the input / output port unit 110-1, the domain processing unit 120-1, and the transfer unit 130-1 corresponding to the subnet # 1 will be described below.

  The input / output port unit 110-1 includes, for example, a plurality of ports, and is connected to a communication terminal device belonging to the subnet # 1 via each port. The input / output port unit 110-1 inputs a frame transmitted from a communication terminal apparatus belonging to the subnet # 1 to the L2 switch 100 from the port, or a frame relayed by the L2 switch 100 from the port to the subnet # 1. Or output to a communication terminal device. Each communication terminal device belonging to subnet # 1 is assigned a unique MAC (Medium Access Control) address, and each port of the input / output port unit 110-1 is assigned to a connection destination communication terminal device. It corresponds to the assigned MAC address. The frame input / output at each port includes a MAC address of a source communication terminal device (hereinafter simply referred to as “source address”) and a MAC address of a destination communication terminal device (hereinafter simply “destination address”). Is included).

  The domain processing unit 120-1 executes processing at the time of input for a frame input from the subnet # 1, and processing at the time of output for a frame output to the subnet # 1. Specifically, the domain processing unit 120-1 includes a domain ID assigning unit 121-1, a destination determining unit 122-1, a port table 123-1, a bandwidth control unit 124-1, an ID converting unit 125-1, and a transmission source. It has a registration unit 126-1 and a learning table 127-1.

  The domain ID assigning unit 121-1 assigns a unique domain ID corresponding to the subnet # 1 to the frame input from the input / output port unit 110-1. That is, the domain ID assigning unit 121-1 assigns a domain ID unique to the subnet to which the frame transmission source belongs to the frame. The domain ID is used in the L2 switch 100 to identify the subnet of the frame transmission source.

  The destination determination unit 122-1 refers to the learning table 127-1 and searches for an output port corresponding to the domain ID given to the frame and the destination address included in the frame. As a result, if the output port corresponding to the domain ID and the destination address has already been learned in the learning table 127-1, the destination determination unit 122-1 uses the flag stored in the learning table 127-1 together with the output port as a frame. In addition, the frame is set to be output from the searched output port, and is output to the transfer unit 130-1. If the output port corresponding to the domain ID and the destination address is not learned in the learning table 127-1, the destination determination unit 122-1 outputs this frame to the port table 123-1.

  The port table 123-1 stores the correspondence between subnets and ports in the input / output port units 110-1 and 110-2. That is, the port table 123-1, for example, as shown in FIG. 3, stores a unique domain ID and a port number of each port in association with each subnet. In the example shown in FIG. 3, the 1st to 7th ports are associated with the domain ID “sub1” corresponding to the subnet # 1, and the 8th to 14th ports are associated with the domain ID “sub2” corresponding to the subnet # 2. Are associated. This means that the input / output port unit 110-1 includes ports 1 to 7, and the input / output port unit 110-2 includes ports 8 to 14.

  Then, when a frame whose output port has not been learned is input from the destination determination unit 122-1, the port table 123-1 outputs the frame to the transfer unit 130-1, and all the ports other than the input port of this frame. The port is set as a flooding port to be flooded in the frame, and the port number of the flooding port is notified to the transfer unit 130-1. At this time, the port table 123-1 also notifies the transfer unit 130-1 of the domain ID corresponding to each flooding port.

  The bandwidth control unit 124-1 performs bandwidth control on the frame output from the transfer unit 130-2 corresponding to the subnet # 2, and converts the frame into a frame corresponding to the subnet # 1. That is, since subnet # 1 and subnet # 2 generally differ in the required quality of service (QoS: Quality of Service), the bandwidth control unit 124-1 is transmitted from subnet # 2 to subnet # 1. Bandwidth control is executed for the frame to be the service quality corresponding to the subnet # 1.

  The ID conversion unit 125-1 converts the domain ID assigned to the frame whose bandwidth is controlled by the bandwidth control unit 124-1 into a domain ID unique to the subnet # 1. Then, ID conversion section 125-1 outputs the frame after domain ID conversion to transmission source registration section 126-1.

  When a frame from the subnet # 2 via the ID conversion unit 125-1 or a frame from the subnet # 1 via the transfer unit 130-1 is input, the transmission source registration unit 126-1 sets the transmission source address of the frame. To acquire the correspondence relationship between the new destination address and the output port in the learning table 127-1. That is, the transmission source registration unit 126-1 acquires the transmission source address of the frame, detects the input port to which this frame is input, and uses the learning table 127 as the transmission source address and the input port as the destination address and the output port, respectively. -1.

  Here, since the communication terminal device indicated by the transmission source address of the frame is connected to the input port of the frame, if the frame is output from this input port, the frame reaches the communication terminal device. Therefore, the transmission source registration unit 126-1 registers the transmission source address and input port of the input frame in the learning table 127-1 as the destination address and output port, respectively. At this time, the transmission source registration unit 126-1 registers the flag and the domain ID added to the frame in the learning table 127-1 at the same time. The flag registered here is a flag indicating whether or not a frame is transmitted across subnets, as will be described later.

  The learning table 127-1 stores the correspondence between the destination address of the frame and the output port. That is, the learning table 127-1 learns the correspondence between the transmission source address and the input port of the frame input to the L2 switch 100 as the correspondence between the destination address and the output port. In addition, the learning table 127-1 stores, as a flag, identification information indicating whether or not the frame referred to when learning the correspondence between the destination address and the output port is transmitted across the subnet. To do.

  Specifically, the learning table 127-1 stores, for example, a destination address, a transmission source address, a domain ID, an output port, and a flag as illustrated in FIG. The destination address matches the source address of the frame already relayed by the L2 switch 100, and the source address matches the destination address of this frame.

  Further, since the domain ID assigned to the domain ID assigned by the domain ID assigning unit 121-1 corresponding to the subnet # 1 is registered as it is for the frame transferred in the subnet # 1, the domain ID is unique to the subnet # 1. The domain ID is “sub1”. On the other hand, for a frame transferred from subnet # 2 to subnet # 1, the domain ID assigned by domain ID assigning unit 121-2 corresponding to subnet # 2 is converted by ID converting unit 125-1 and obtained. Since the domain ID is registered, the domain ID is also the domain ID “sub1” unique to the subnet # 1.

  Furthermore, the output port is the port number of the port provided in the input / output port unit 110-1 because the input port is included in the input / output port unit 110-1 for the frame transferred in the subnet # 1. On the other hand, for the frame transferred from the subnet # 2 to the subnet # 1, the input port is included in the input / output port unit 110-2, so the port number of the port provided in the input / output port unit 110-2 Become. The flag “0” is stored in the entry corresponding to the frame transferred in the subnet # 1, and the flag “1” is stored in the entry corresponding to the frame transferred across the subnet.

  The forwarding unit 130-1 floods the frame input from the port table 123-1, while the domain processing unit 120-1 corresponding to the output port in which the frame directly input from the destination determination unit 122-1 is set or Output to the domain processing unit 120-2. Specifically, the transfer unit 130-1 includes a flooding processing unit 131 and a flag determination unit 132.

  The flooding processing unit 131 performs a flooding process in which the output ports of the frames output from the port table 123-1 are all the flooding ports notified from the port table 123-1. That is, since the frame whose output port corresponding to the destination address has not been learned is output from the port table 123-1, the flooding processing unit 131 performs flooding so that this frame is output from all ports other than the input port of the frame. Execute the process.

  At this time, the flooding processing unit 131 generates a frame output from each flooding port by duplicating the corresponding frame, and sets the output port of each frame as a flooding port. Then, the flooding processing unit 131 refers to the correspondence relationship between the flooding port and the domain ID notified from the port table 123-1, and if the output port is a port connected to the same subnet as the frame transmission source, When the flag “0” is added and the output port is a port connected to a subnet different from the frame transmission source, the flag “1” is added to the frame. In other words, the flooding processing unit 131 adds a flag “0” to a frame that is flooded in the same subnet as the frame transmission source, and adds a flag “1” to a frame that is flooded across the subnet. .

  The flag determination unit 132 determines whether the flag added to the frame is “0” or “1”, and the frame added with the flag “0” is the transmission source registration unit of the domain processing unit 120-1. The frame is output to 126-1 and the frame with the flag “1” added is output to the bandwidth control unit 124-2 of the domain processing unit 120-2. Specifically, when a frame whose output port corresponding to the destination address has been learned is input from the destination determination unit 122-1, the flag determination unit 132 displays the flag added to the frame by the destination determination unit 122-1. If the flag “0” is added, the frame is output to the transmission source registration unit 126-1. If the flag “1” is added, the frame is output to the bandwidth control unit 124- of the domain processing unit 120-2. Output to 2. In addition, when a copy of a frame whose output port corresponding to the destination address has not been learned is input from the flooding processing unit 131, the flag determination unit 132 sets the copy of the frame to which the flag “0” is added to the transmission source registration unit 126. -1 and a copy of the frame with the flag “1” added is output to the bandwidth control unit 124-2 of the domain processing unit 120-2.

  Next, the operation of the L2 switch 100 configured as described above will be described with reference to the flowchart shown in FIG. In the following, the operation when the frame transmitted from the subnet # 1 is input to the L2 switch 100 will be described in detail. However, the frame transmitted from the subnet # 2 is input to the L2 switch 100. The operation is the same except for the reference numerals of the processing units that execute the respective processes.

  First, when a frame transmitted from the subnet # 1 is input from any of the ports provided in the input / output port unit 110-1 (step S101), the domain ID assigning unit 121-1 uniquely identifies the subnet # 1. Domain ID (for example, “sub1”) is assigned to the frame (step S102). Then, the destination determination unit 122-1 determines whether or not the learning table 127-1 has learned the output port corresponding to the destination address included in the frame and the domain ID assigned to the frame (step S103). ).

  If the result of this determination is that the output port has been learned (step S103 Yes), the destination determination unit 122-1 acquires the output port and flag corresponding to the destination address and domain ID from the learning table 127-1 (step S104). ) After the flag is added to the frame, the output port of the frame is set to the output port acquired from the learning table 127-1 and is output to the flag determination unit 132 of the transfer unit 130-1.

  Here, by adding a flag to the frame, it is possible to determine whether or not this frame is transferred across subnets for frames whose output ports have been learned. Specifically, if the flag “0” is added to the frame, this means that the frame is transferred within the subnet # 1, and if the flag “1” is added to the frame, this frame is This means that the data is transferred from subnet # 1 to subnet # 2. This is because the flag “0” is stored together with the correspondence between the destination address and the output port if the frame is transferred within the subnet during learning in the learning table 127-1, and the frame is transferred across the subnet. This is because the flag “1” is stored together with the correspondence between the destination address and the output port.

  On the other hand, if the output port corresponding to the destination address and the domain ID has not been learned in the learning table 127-1 (No in step S103), the frame is sent to the flooding processing unit of the transfer unit 130-1 via the port table 123-1. It is output to 131. At the same time, all the ports (flooding ports) other than the frame input port and the domain ID corresponding to each flooding port are notified from the port table 123-1 to the flooding processing unit 131, and the flooding processing unit 131 floods the frame. The target flooding port is acquired (step S105).

  Then, the flooding processing unit 131 duplicates the frame, sets the output port for duplication of each frame as a flooding port, and sets the output port to the port of the input / output port unit 110-1 corresponding to the subnet # 1. If the output port is the port of the input / output port unit 110-2 corresponding to the subnet # 2, the flag “1” is added (step S106). The copy of the frame in which the output port and the flag are set in this way is output to the flag determination unit 132.

  Here, by adding a flag to the copy of the frame, it is possible to determine whether or not each copy of this frame is a duplicate that is flooded across subnets for a frame whose output port has not been learned. It becomes. Specifically, if the flag “0” is added to the copy of the frame, this means that the copy is flooded in the subnet # 1, and if the flag “1” is added to the copy of the frame, This duplication means that the subnet # 1 is flooded to the subnet # 2. Although the copy of the frame is different in output port and flag, the stored information itself is the same including the destination address and the source address. Therefore, in the following, without distinguishing between the frame and the copy of the frame. Both are collectively referred to as a “frame”.

  When the frame is output from the destination determination unit 122-1 or the flooding processing unit 131 to the flag determination unit 132, the flag determination unit 132 determines whether or not the flag added to the frame is “1” ( Step S107). In other words, the flag determination unit 132 determines whether or not the frame is transferred across subnets.

  As a result of this determination, if the flag “1” is added to the frame (Yes in step S107), the flag determination unit 132 outputs the frame to the bandwidth control unit 124-2 of the domain processing unit 120-2. Then, the bandwidth control unit 124-2 controls the bandwidth of the frame to a bandwidth suitable for the subnet # 2 (step S108), and the ID conversion unit 125-2 adds the domain ID (for example, “sub1”) to the frame. Is converted into a domain ID (eg, “sub2”) unique to subnet # 2 (step S109).

  Then, the correspondence between the transmission source address of the frame and the input port is registered in the learning table 127-2 as the correspondence between the destination address and the output port by the transmission source registration unit 126-2 (step S110). At this time, the domain ID and the flag added to the frame are simultaneously stored in the learning table 127-2 by the transmission source registration unit 126-2. Therefore, since the domain ID has been converted here, the domain ID (for example, “sub2”) unique to the subnet # 2 and the flag “1” are stored in the learning table 127-2.

  In this way, since the flag added to the frame when the correspondence relationship between the destination address and the output port is learned is simultaneously stored in the learning table 127-2, in the relay of the frame for this destination address, the frame has a subnet. Whether or not the data is transferred across the frame is also stored at the same time. That is, when the flag “1” is stored as described above, it indicates that the relay of the frame for the destination address learned at the same time crosses the subnet. This is because the learned destination address matches the transmission source address of the frame transmitted from the subnet # 1, and the frame transferred from the subnet # 2 to this destination address matches the crossing of the subnet. Yes.

  After registration in the learning table 127-2 by the transmission source registration unit 126-2, the frame is output from the output port set in the frame among the ports provided in the input / output port unit 110-2 (step S111). . If the output port corresponding to the destination address of the frame has already been learned by outputting the frame in this way, the communication terminal device corresponding to the destination address is connected to the output port of the frame, so that the frame is in the desired communication. It will be received by the terminal device. In addition, if the output port corresponding to the destination address of the frame is not learned, the communication terminal device corresponding to the destination address is not necessarily connected to the output port of the frame. When the response frame indicating that the message is correctly received is relayed again by the L2 switch 100, the correspondence relationship between the address and the port is newly learned. Correspondence between address and port is not learned.

  On the other hand, if the flag determination unit 132 determines that the flag “0” is added to the frame (No in step S107), the flag determination unit 132 causes the frame to be registered in the transmission source registration unit 126 of the domain processing unit 120-1. To -1. Then, the transmission source registration unit 126-1 registers the correspondence between the transmission source address of the frame and the input port in the learning table 127-1 as the correspondence between the destination address and the output port (step S110). At this time, the domain ID and the flag added to the frame are also stored in the learning table 127-1 by the transmission source registration unit 126-1. Therefore, since the domain ID remains assigned by the domain ID assigning unit 121-1, the domain ID unique to the subnet # 1 (for example, “sub1”) and the flag “0” are stored in the learning table 127-1. Remembered.

  In this way, since the flag added to the frame when the correspondence relationship between the destination address and the output port is learned in the learning table 127-1 is stored at the same time, in the relay of the frame for this destination address, the frame has a subnet. Whether or not the data is transferred across the frame is also stored at the same time. That is, when the flag “0” is stored as described above, it indicates that the relay of the frame for the destination address learned at the same time does not cross the subnet. This is because the learned destination address matches the source address of the frame transmitted from the subnet # 1, so that the frame transferred from the subnet # 1 to this destination address does not cross the subnet. I'm doing it.

  After registration in the learning table 127-1 by the transmission source registration unit 126-1, the frame is output from the output port set in the frame among the ports provided in the input / output port unit 110-1 (step S111). .

  As described above, in this embodiment, a flag indicating whether or not a frame is transferred across subnets when a frame including flooding is relayed is added, and the output port corresponding to the destination address including this flag is added. Are learned in the learning tables 127-1 and 127-2. Therefore, when the output port corresponding to the destination address of the frame has been learned, the flag learned by the learning tables 127-1 and 127-2 is used to determine whether the frame is transferred across the subnet. It can be easily judged. That is, for example, even if the fee structure is different between communication within the subnet and communication between subnets, by referring to the flag acquired from the learning tables 127-1 and 127-2 and added to the frame, It is possible to distinguish between communication and communication between subnets.

  In the present embodiment, since subnet # 1 and subnet # 2 are connected only through L2 switch 100, only one L2 switch is required to connect two subnets, and all ports are connected. Can be used for connection with communication terminal devices belonging to each subnet. As a result, an increase in cost required for the network configuration can be suppressed.

  As described above, according to the present embodiment, the flag indicating whether or not the frame is transferred across the subnet is stored in the learning table simultaneously with the correspondence between the destination address and the output port. When a frame is relayed to a learned destination address, it can be easily determined whether or not this frame is transferred across subnets. In other words, it is possible to discriminate whether or not it is a relay across subnets within one L2 switch, and it is possible to distinguish between communications within a subnet and communications between subnets while suppressing an increase in the cost required for network configuration. it can.

  In the above embodiment, the transmission source registration units 126-1 and 126-2 perform learning in the learning tables 127-1 and 127-2 every time a frame is relayed. Is already registered in the learning tables 127-1 and 127-2 as destination addresses, the source address and input port of this frame are newly registered in the learning tables 127-1 and 127-2. It is not necessary. However, the learning tables 127-1 and 127-2 can always be kept up-to-date by executing learning each time a frame is relayed as in the above embodiment.

  In the above-described embodiment, the learning tables 127-1 and 127-2 are provided in the separate domain processing units 120-1 and 120-2, respectively. Two learning tables may be used. Even if the learning tables are integrated, it is possible to determine whether or not the output port to be referred to has been learned based on the domain ID stored in the learning table. That is, for a frame transmitted from subnet # 1, an output port in which a domain ID corresponding to subnet # 1 is associated and stored is searched, and for a frame transmitted from subnet # 2, it corresponds to subnet # 2. The output port stored in association with the domain ID to be stored may be searched. Similarly, the port tables 123-1 and 123-2 may be integrated to use one port table.

  Further, in the above embodiment, the frame path learning method in the L2 switch 100 has been described. However, the path learning method is described as a computer-readable path learning program, and the computer executes the path learning program. Thus, it is possible to obtain the same effect as that of the above embodiment.

  Regarding the above embodiment, the following additional notes are disclosed.

(Supplementary Note 1) Acquisition means for acquiring a frame from an input port connected to the first subnet;
Setting means for setting an output port corresponding to the destination of the frame acquired by the acquisition means;
First transmission means for transmitting a frame to the first subnet via the output port when the output port of the frame set by the setting means is connected to the first subnet;
Second transmission means for transmitting the frame to the second subnet via the output port when the output port of the frame set by the setting means is connected to a second subnet different from the first subnet; A frame relay device comprising:

(Appendix 2) Acquisition means for acquiring a frame from an input port connected to the first subnet;
Setting means for setting an output port corresponding to the destination of the frame acquired by the acquisition means;
Adding means for adding to the frame a flag indicating whether or not the output port of the frame set by the setting means is connected to a second subnet different from the first subnet;
Registering means for registering a flag added to the frame by the adding means together with a transmission source address and an input port of the frame acquired by the acquiring means in a learning table storing a correspondence relationship between the address and the port. Frame relay device.

(Appendix 3) The setting means includes:
The frame relay device according to appendix 2, wherein a port corresponding to the destination address of the frame is set as an output port of the frame in the learning table when the destination address of the frame has been registered in the learning table by the registration means .

(Appendix 4) The additional means is:
The frame relay apparatus according to appendix 3, wherein a flag stored in the learning table is added to the frame in correspondence with the output port set by the setting means.

(Supplementary Note 5) The setting means includes:
The frame relay device according to appendix 2, wherein all ports other than the input port of the frame are set as output ports when the destination address of the frame is not registered in the learning table by the registration means.

(Appendix 6) The additional means is:
The frame is duplicated according to the output port set by the setting means, and among the obtained duplicate frames, the duplicate frame in which the output port corresponding to the first subnet is set is connected to the first subnet. 6. A frame according to appendix 5, wherein a flag indicating that the connection is made to the second subnet is added to a duplicate frame in which an output port corresponding to the second subnet is set. Relay device.

(Supplementary note 7) Band control for controlling the bandwidth of the frame to a bandwidth suitable for the second subnet when the flag added to the frame by the adding means indicates that the output port of the frame is connected to the second subnet Means,
The frame relay apparatus according to claim 2, further comprising: output means for outputting a frame whose bandwidth is controlled by the bandwidth control means from an output port.

(Supplementary note 8) A route learning program executed by a computer,
An acquisition step of acquiring a frame from an input port connected to the first subnet;
A setting step for setting an output port corresponding to the destination of the frame acquired in the acquisition step;
An adding step of adding to the frame a flag indicating whether the output port of the frame set in the setting step is connected to a second subnet different from the first subnet;
A registration step of registering the flag added to the frame in the addition step together with the transmission source address and input port of the frame acquired in the acquisition step in a learning table that stores a correspondence relationship between the address and the port. A route learning program characterized by that.

(Supplementary Note 9) An acquisition step of acquiring a frame from an input port connected to the first subnet;
A setting step for setting an output port corresponding to the destination of the frame acquired in the acquisition step;
An adding step of adding to the frame a flag indicating whether the output port of the frame set in the setting step is connected to a second subnet different from the first subnet;
A registration step of registering the flag added to the frame in the addition step together with the transmission source address and input port of the frame acquired in the acquisition step in a learning table that stores a correspondence relationship between the address and the port. A route learning method characterized by

  The present invention can be applied to a case where communication within a subnet is distinguished from communication between subnets while suppressing an increase in cost required for the configuration of the network.

110-1, 110-2 Input / output port unit 120-1, 120-2 Domain processing unit 121-1, 121-2 Domain ID assigning unit 122-1, 122-2 Destination determination unit 123-1, 123-2 port Tables 124-1, 124-2 Band control units 125-1, 125-2 ID conversion units 126-1, 126-2 Source registration units 127-1, 127-2 Learning tables 130-1, 130-2 Transfer units 131 Flooding processing unit 132 Flag determination unit

Claims (7)

A first acquisition means for acquiring a frame from an input port connected to the first subnet,
First setting means for setting an output port corresponding to the destination of the frame acquired by the first acquiring means based on a learning table ;
Second acquisition means for acquiring a frame from an input port connected to the second subnet;
Second setting means for setting an output port corresponding to the destination of the frame acquired by the second acquiring means;
And adding means for adding a flag output port of the frame set by the second setting means indicates whether connecting to the first subnet to the frame,
Registering means for registering in the learning table the transmission source address and input port of the frame acquired by the second acquisition means as the transmission destination address and output port together with the flag added to the frame by the adding means; A frame relay apparatus characterized by that. The first setting means includes
2. The frame relay according to claim 1, wherein when the destination address of the frame has been registered in the learning table by the registration means, a port corresponding to the destination address of the frame is set as an output port of the frame in the learning table. apparatus. The first setting means includes
2. The frame relay apparatus according to claim 1, wherein when the destination address of the frame is not registered in the learning table by the registration means, all ports other than the input port of the frame are set as output ports. The adding means includes
The frame is duplicated according to the output port set by the second setting means, and among the obtained duplicate frames, the duplicate frame in which the output port corresponding to the first subnet is set is the first The flag indicating that the connection is to the subnet is added, and the flag indicating that the connection is to the second subnet is added to the duplicate frame in which the output port corresponding to the second subnet is set. 3. The frame relay device according to 3. Bandwidth control means for controlling the bandwidth of the frame to a bandwidth suitable for the second subnet when the flag added to the frame by the addition means indicates that the output port of the frame is connected to the second subnet;
The frame relay apparatus according to claim 1, further comprising: an output unit that outputs from the output port a frame whose band is controlled by the band control unit. A route learning program executed by a computer, the computer
A first acquisition step of acquiring a frame from an input port connected to the first subnet,
A first setting step of setting an output port corresponding to the destination of the frame acquired in the first acquisition step based on a learning table ;
A second acquisition step of acquiring a frame from an input port connected to the second subnet;
A second setting step of setting an output port corresponding to the destination of the frame acquired by the second acquiring step;
An adding step of adding a flag output port of the frame set by the second setting step indicates whether to connect to the first subnet to the frame,
A registration step of registering in the learning table, with the flag added to the frame in the addition step, the transmission source address and input port of the frame acquired in the second acquisition step as a transmission destination address and output port, and A route learning program characterized by being executed. A first acquisition step of acquiring a frame from an input port connected to the first subnet,
A first setting step of setting an output port corresponding to the destination of the frame acquired in the first acquisition step based on a learning table ;
A second acquisition step of acquiring a frame from an input port connected to the second subnet;
A second setting step of setting an output port corresponding to the destination of the frame acquired by the second acquiring step;
An adding step of adding to the frame a flag indicating whether or not the output port of the frame set in the second setting step is connected to the first subnet;
A registration step of registering in the learning table, with the flag added to the frame in the addition step, the transmission source address and input port of the frame acquired in the second acquisition step as a transmission destination address and output port, and A route learning method characterized by comprising:

Images