JP4809758B2 - Network relay device and method for controlling network relay device - Google Patents

Description

  The present invention relates to a network relay system and a method for controlling a network relay device.

  In general, a network relay device such as a repeater hub or a layer 2 switch is used for wiring between network devices. The repeater hub is a device for simply connecting a plurality of network devices electrically. On the other hand, the layer 2 switch has a function of transferring an input packet to an optimum physical port according to the destination address. In recent offices, a layer 2 switch is often used instead of a repeater hub in order to suppress an increase in traffic accompanying an increase in network equipment. In recent years, a network relay device called a layer 3 switch is also used to construct a larger-scale network. The layer 3 switch has a router function in addition to the bridge function of the layer 2 switch.

  Some of the above-described layer 2 switches and layer 3 switches have a function called link aggregation. Such a function is a function that realizes expansion of bandwidth and securing of redundancy by handling a plurality of physical lines connecting between switches as one logical line (see Patent Document 1).

JP 2004-349664 A

  However, the conventional link aggregation function cannot be said to be able to use the physical line and its bandwidth sufficiently efficiently. Such a problem is not limited to link aggregation, but is a common problem when a plurality of logical lines are constructed using a plurality of physical lines. An object of the present invention is to provide a technology for more efficiently using a physical line and its bandwidth when a plurality of logical lines are constructed using a plurality of physical lines.

A first aspect of the present invention provides a network relay device. A network relay device according to a first aspect is a relay control unit capable of constructing a plurality of logical lines using a plurality of ports for connecting physical lines and a plurality of physical lines connected to the plurality of ports. A relay control unit capable of assigning a first physical line of the plurality of physical lines to a plurality of different logical lines.
Each of the plurality of logical lines is a link aggregation group that logically bundles two or more lines of the plurality of physical lines and treats them as one logical line.
The relay control unit
(A) setting a bandwidth ratio for allocating the bandwidth of the first physical line to a plurality of logical lines;
(B) determining bandwidths of a plurality of link aggregation groups including the first physical line based on the bandwidth ratio;
(C) When determining a link aggregation group to be used for data transmission from among the plurality of link aggregation groups, the bandwidth of the link aggregation group to be used is changed according to the frame flow rate of each of the plurality of link aggregation groups. To do.

According to the network relay device according to the first aspect, the first physical line of the plurality of physical lines can be assigned to a plurality of different logical lines. Therefore, it is possible to construct a network relay system that uses a physical line and its bandwidth more efficiently. In addition, flexible link aggregation can be constructed by using the physical line and its bandwidth more efficiently.

  In the network relay device according to the first aspect, allocating the first physical line to a plurality of different logical lines treats the first physical line as a plurality of virtual lines, It may be realized by assigning lines to different logical lines. By so doing, it is possible to easily realize that the first physical line is assigned to a plurality of different logical lines.

  In the network relay device according to the first aspect, the plurality of logical lines may be a plurality of link aggregation groups in which two or more lines of the plurality of physical lines are logically bundled. In this way, it is possible to construct a flexible link aggregation by using the physical line and its bandwidth more efficiently.

  In the network relay device according to the first aspect, the first physical line is a standby line used when a failure occurs in the second physical line belonging to the link aggregation group to which the first physical line belongs. Also good. In this way, it is possible to reduce the number of physical lines necessary as standby lines while ensuring the redundancy of link aggregation.

  In the network relay device according to the first aspect, a bandwidth ratio assigned to a plurality of logical lines is set for the first physical line, and the relay control unit is configured to perform the first physical line based on the bandwidth ratio. A line may be used. In this way, it is possible to construct various link aggregations using the physical line and its bandwidth more efficiently.

  In the network relay device according to the first aspect, the network relay device may be a layer 3 switch or a layer 2 switch.

  The second aspect of the present invention provides a method for controlling a network relay device capable of connecting a plurality of physical lines. In the control method according to the second aspect, when a plurality of logical lines are constructed using a plurality of physical lines, a first physical line of the plurality of physical lines is assigned to a plurality of different logical lines.

  According to the control method according to the second aspect, it is possible to obtain the same operation and effect as the network relay device according to the first aspect. Further, the control method according to the second aspect can be realized in various aspects, like the network relay device according to the first aspect.

  The present invention can be further realized in various aspects, for example, a network relay system including the network relay device according to the first aspect, and a control method for such a network relay system. Can do. In addition to these system, apparatus, and method aspects, a computer program for constructing the system, apparatus, or method, a recording medium that records the computer program, and data embodied in a carrier wave including the computer program It can be realized in the form of a signal or the like.

A. First embodiment:
-Network relay system configuration:
Next, embodiments of the present invention will be described based on examples. The configuration of the network relay system according to the first embodiment will be described with reference to FIGS. FIG. 1 is a block diagram illustrating a schematic configuration of the network relay system according to the first embodiment. FIG. 2 is a block diagram illustrating an internal configuration of the network relay device included in the network relay system according to the first embodiment. FIG. 3 is a diagram conceptually showing the data structure of the Ethernet frame. FIG. 4 is a diagram conceptually illustrating various tables stored in the first relay device. FIG. 5 is a diagram conceptually illustrating various tables stored in the second relay device.

  As shown in FIG. 1, the network relay system 1000 according to the present embodiment includes two network relay devices, that is, a first relay device 100a and a second relay device 100b. In the following, when it is not necessary to distinguish between the two network relay devices, the alphabet at the end of the code is omitted and the network relay device 100 is described.

  As illustrated in FIG. 2, the network relay device 100 includes a plurality of physical ports 110, a transmission / reception processing circuit 120, a frame processing circuit 130, and a device control unit 140.

  The physical port 110 is an interface for connecting to a network via a physical line CV such as a coaxial cable or an optical fiber. In this embodiment, the physical port 110 is a port compliant with the Ethernet (registered trademark) standard.

  The transmission / reception processing circuit 120 is connected to each physical port 110, interprets an electrical signal received via the physical port 110, and collects data used in the data link layer (hereinafter referred to as a frame. In this embodiment). , Ethernet (registered trademark) frame). The transmission / reception processing circuit 120 sends the converted Ethernet frame to the frame processing circuit 130. Further, the transmission / reception processing circuit 120 receives the Ethernet frame to be transferred from the frame processing circuit 130 and port information for specifying the physical port to be transmitted, converts the Ethernet frame into an electrical signal, and specifies based on the port information Process to send from the physical port.

  Here, the Ethernet frame will be described with reference to FIG. The Ethernet frame 800 includes a payload 830, an IP header 820, and an Ethernet header 810. The payload 830 is a data body to be transferred. The IP header 820 is a header in which information used for processing in the network layer is described. As shown in FIG. 3, the IP header 820 includes a TOS (Type of Service) field, a source IP address (hereinafter also referred to as SIP), and a destination IP address (hereinafter also referred to as DIP). . In the present embodiment, information (hereinafter referred to as a TOS value) that specifies the priority of the frame is described in the TOS field. In the present embodiment, when the TOS value is “0”, it indicates that the frame is a low priority frame (low priority frame), and when the TOS value is “1”, the frame is It is assumed that the frame has a high priority (high priority frame).

  The Ethernet header 810 includes a destination MAC address (hereinafter also referred to as DMAC), a transmission source MAC address (hereinafter also referred to as SMAC), a COS (Class of Service) field, and an Ethertype field. The COS field may be omitted.

  The frame processing circuit 130 includes a search processing unit 131 and a memory 134. The memory 134 stores a routing table 135, an address-port correspondence table 136, and a LAG (Link Aggregation Group) setting table 137. The routing table 135 is a table in which information included in an Ethernet frame to be transferred, such as a destination IP address, and information that associates the IP address of the transfer destination device are described. The address-port correspondence table 136 is a table in which information that associates an IP address of a transfer destination device, a MAC address of the transfer destination device, and a port (corresponding port) for transfer to the transfer destination device is described. . The LAG setting table 137 is a table in which the definition of link aggregation is described when link aggregation that handles a plurality of physical lines as one logical line is constructed. The routing table 135, the address-port correspondence table 136, and the LAG setting table 137 will be further described later.

  The search processing unit 131 is an ASIC (Application Specific Integrated Circuit) designed to realize the function of the circuit described later, and executes the function of the circuit described later by hardware processing. The search processing unit 131 is a circuit that performs a transfer process for transferring the Ethernet frame received from the transmission / reception processing circuit 120. Specifically, the search processing unit 131 specifies the IP address of the transfer destination device that transfers the Ethernet frame based on the destination IP address included in the Ethernet frame. The identification of the IP address of the transfer destination device is executed with reference to a routing table 135 stored in the memory 134, as will be described later. The search processing unit 131 further specifies the MAC address of the specified transfer destination device and specifies the corresponding port for transferring the Ethernet frame to the specified transfer destination device. The identification of the MAC address and the corresponding port is executed with reference to the address-port correspondence table 136 stored in the memory 134.

  Here, the corresponding port specified with reference to the address-port correspondence table 136 is not limited to a physical port corresponding to a physical line, but may be a logical port corresponding to a logical line based on the above-described link aggregation. is there. When the specified port is a logical port, the search processing unit 131 refers to the above-described LAG setting table 137 and specifies a physical port for transferring an Ethernet frame.

  The search processing unit 131 changes the destination MAC address included in the Ethernet frame to the MAC address of the specified transfer destination device, and transmits the changed Ethernet frame to the transmission / reception processing circuit 120 by specifying the specified physical port. To do. As a result, the Ethernet frame is transferred from the physical port to the transfer destination device. The transfer performed by specifying the corresponding port by the search processing unit 131 is performed based on the IP address that is the address of the network layer that is the third layer of the OSI (Open Systems Interconnection) reference model. Call it.

  The device control unit 140 controls the entire network relay device 100. The device control unit 140 is a well-known computer, and realizes a function as a device control unit by executing a control program. The device control unit 140 creates a function of processing routing protocols such as RIP (Routing Information Protocol) and OSPF (Open Shortest Path First), and the various tables 135, 136, and 137 described above, and creates a frame processing circuit 130. The function stored in the memory 134 is executed.

  Returning to FIG. 1, the description will be continued. In FIG. 1, black circles indicate ports provided in each device. For example, the ports P1 to P5 of the network relay devices 100a and 100b correspond to any of the physical ports 110 in FIG.

  As shown in FIG. 1, the physical ports P1 to P4 of the first relay device 100a are connected to the physical ports P1 to P4 of the second relay device 100b via physical lines CV1 to CV4, respectively. . The physical port P5 of the first relay device 100a is connected to the external network NT1, and the physical port P5 of the second relay device 100b is connected to the external network NT2. The first relay device 100a and the second relay device 100b may be connected to a larger number of external networks via other physical ports.

  Next, the contents of various tables stored in the first relay device 100a and the second relay device 100b will be described with reference to FIG. 4 and FIG. The table stored in the first relay device 100a is indicated by a at the end of the code, and the table stored in the second relay device 100b is indicated by having b at the end of the code. In this embodiment, as shown in FIG. 4, the routing table 135a of the first relay device 100a contains the destination IP address and the IP address (transfer destination address) of the transfer destination device based on the TOS value described above. Correspondence is described so that it can be identified. In the example shown in FIG. 4, IP_B1 is specified as the IP address of the transfer destination device for the Ethernet frame whose destination IP address is IP_X and whose TOS value is “1”. For an Ethernet frame whose destination IP address is IP_X and whose TOS value is “0”, IP_B2 is specified as the IP address of the transfer destination device. That is, among the Ethernet frames having the same destination address IP_X, the transfer destination IP address of the high priority frame is IP_B1, and the transfer destination IP address of the low priority frame is IP_B2.

  The correspondence relationship is described in the address-port correspondence table 136a of the first relay device 100a so that the MAC address of the transfer destination device and the corresponding port can be identified based on the IP address of the transfer destination device. In the example shown in FIG. 4, MAC_B1 is described as the MAC address corresponding to the IP address IP_B1 of the transfer destination device, and LAG1, which is a logical port corresponding to link aggregation, is described as the corresponding port. Then, MAC_B2 is described as the MAC address corresponding to the IP address IP_B2 of the transfer destination device, and LAG2, which is a logical port corresponding to link aggregation, is described as the corresponding port.

  The LAG setting table 137a of the first relay device 100a describes the physical ports (belonging ports) assigned to the logical ports LAG1 and LAG2. Thereby, link aggregation is defined. Physical ports P1, P2, and P3 are allocated to the logical port LAG1, and physical ports P3 and P4 are allocated to the logical port LAG2. Here, the physical port P3 is assigned to both the logical port LAG1 and the logical port LAG2. As shown in FIG. 4, the physical port P3 is one physical port, but is virtually treated as two ports P3-1 and P3-2. As shown in FIG. 4, the LAG setting table 137a is described so that the virtual port P3-1 is assigned to the logical port LAG1 and the virtual port P3-2 is assigned to the logical port LAG2.

  Similarly to the first relay device 100a, as shown in FIG. 5, the routing table 135b of the second relay device 100b includes the destination IP address and the IP address of the transfer destination device based on the TOS value described above. The correspondence relationship is described so that the (transfer destination address) can be specified. In the example illustrated in FIG. 5, IP_A1 is specified as the IP address of the transfer destination device in the Ethernet frame whose destination IP address is IP_Y and whose TOS value is “1”. For an Ethernet frame whose destination IP address is IP_Y and whose TOS value is “0”, IP_A2 is specified as the IP address of the transfer destination device.

  Similarly to the first relay device 100a, the MAC address of the transfer destination device and the corresponding port can be specified in the address-port correspondence table 136b of the second relay device 100b based on the IP address of the transfer destination device. Describes the correspondence. In the example shown in FIG. 5, MAC_A1 is described as the MAC address corresponding to the IP address IP_A1 of the transfer destination apparatus, and LAG1, which is a logical port corresponding to link aggregation, is described as the corresponding port. Then, MAC_A2 is described as the MAC address corresponding to the IP address IP_A2 of the transfer destination device, and LAG2, which is a logical port corresponding to link aggregation, is described as the corresponding port.

  The LAG setting table 137b of the second relay device 100b describes the physical ports (belonging ports) assigned to the logical ports LAG1 and LAG2. Thereby, link aggregation is defined. Here, the definition of each link aggregation is appropriately set so as not to contradict the definition of the link aggregation (LAG setting table 137a) in the first relay device 100a which is the device on the other side of the link aggregation. As shown in FIG. 5, physical ports P1, P2, and P3 are assigned to the logical port LAG1, and physical ports P3 and P4 are assigned to the logical port LAG2. As in the definition on the first relay device 100a side, the physical port P3 is assigned to both the logical port LAG1 and the logical port LAG2. As shown in FIG. 5, the physical port P3 is one physical port as in the definition on the first relay device 100a side, but is virtually treated as two ports P3-1 and P3-2. This is synonymous with the fact that the physical line CV3 connected to the physical port P3 is handled as two virtual lines. Then, as shown in FIG. 5, the LAG setting table 137b describes that the virtual port P3-1 is assigned to the logical port LAG1 and the virtual port P3-2 is assigned to the logical port LAG2.

  As a result, as shown in FIG. 1, the link aggregation group LAG1 (physical ports corresponding to the link aggregation are used for ease of understanding) that handles the physical lines CV1 to CV3 virtually as one logical line. Therefore, a link aggregation group LAG2 that handles the physical lines CV3 and CV4 virtually as one logical line is defined. The physical line CV3 indicated by hatching belongs to a plurality of link aggregation groups (LAG1 and LAG2). Hereinafter, like the physical line CV3 in the present embodiment, a physical line belonging to a plurality of link aggregation groups is also referred to as a shared physical line.

  Also, the attributes of the physical ports belonging to each link aggregation group are described in the LAG setting table 137a of the first relay device 100a and the LAG setting table 137b of the second relay device 100b. As can be seen from these, the shared physical line CV3 is set as a standby line in two link aggregation groups. The standby line is a line that is used in place of the normal line in which a failure occurs when some failure (for example, disconnection of the line) occurs in another physical line (normal line) belonging to the link aggregation group.

-Network relay system operation:
With reference to FIG. 6, the operation of the network relay system 1000 will be described. FIG. 6 is a flowchart showing the operation steps of each network relay device. As a specific example, the first relay device 100a has a high priority frame addressed to the device X (IP_X) ahead of the external network NT2, that is, an Ethernet frame whose destination IP address is IP_X and whose priority value is “1”. An operation when PX is received via the external network NT1 will be described.

The first relay device 100a receives the Ethernet frame PX having the MAC address destined for itself from the physical port P5 (step S110). When the Ethernet frame is received, in the frame processing circuit 130a of the first relay device 100a, the search processing unit 131a specifies the transfer destination of the Ethernet frame PX (specifically, the IP_address of the transfer destination device) (step S120). ). Specifically, the search processing unit 131a determines the routing table 135a based on the destination IP address “IP_X” described in the IP header of the Ethernet frame PX and the priority value “1” described in the TOS field. To specify the forwarding IP address “IP_A”.

Next, the search processing unit 131a identifies a corresponding port for transferring the Ethernet frame PX to the transfer destination (step S130). Specifically, the search processing unit 131a refers to the address-port correspondence table 136a based on the transfer destination IP address “IP_A” identified in step S120, and identifies the logical port LAG1 as the corresponding port.

  Next, the search processing unit 131a determines whether the identified corresponding port is a logical port or a physical port (step S140). If the identified corresponding port is not a logical port, that is, if it is a physical port (for example, P8 in the address-port correspondence table 136a) (step S140: NO), the process proceeds to step S160, and the transmission / reception processing circuit 120 Then, the Ethernet frame is transferred from the physical port which is the identified corresponding port (step S160).

  On the other hand, if the identified corresponding port is a logical port (step S140: YES), the search processing unit 131a should transfer an Ethernet frame from among the physical ports assigned to the identified logical port. A physical port is specified (step S150). For example, when LAG1 (logical port corresponding to the link aggregation group as described above) is specified as the corresponding port as in the specific example, the LAG setting table 137a is referred to and the belonging port assigned to LAG1 is identified. Specify the physical port to which Ethernet frames should be transferred. In this specific example, among the physical ports assigned to LAG1, P1 and P2 are physical ports to which normal lines CV1 and CV2 are connected, and P3 (described as a virtual port P3-1) is a standby line. Since CV3 is a physical port connected (see LAG setting table 137a), when there is no failure in the normal lines CV1 and CV2, the search processing unit 131a is connected to the normal lines CV1 and CV2. One of the physical ports P1 and P2 is specified as a physical port to which the Ethernet frame is to be transferred. A well-known method may be used to select which of the physical ports P1 and P2. For example, in order to make the bandwidth usage of the normal lines CV1 and CV2 uniform, they may be selected alternately or based on information in the Ethernet frame (for example, the source port number and destination port number in the TCP header) It may be selected according to a predetermined rule.

  Then, when a failure has occurred in the normal line CV1, the search processing unit 131a has a physical port P2 connected to the normal line CV2 and a physical port P3 (virtual port P3-1 connected to the standby line CV3). ) Is specified as a physical port to which an Ethernet frame is to be transferred. Similarly, when a failure has occurred in the normal line CV2, the search processing unit 131a has a physical port P2 connected to the normal line CV1 and a physical port P3 (virtual port P3 connected to the standby line CV3). -1) is specified as a physical port to which an Ethernet frame is to be transferred. The search processing unit 131a transfers an Ethernet frame to the physical port P3 (virtual port P3-1) connected to the standby line CV3 when both the normal lines CV1 and CV2 fail. Identifies the physical port to be used.

  When the physical port to which the Ethernet frame is to be transferred is specified, the transmission / reception processing circuit 120 transfers the Ethernet frame from the physical port to which the Ethernet frame is to be transferred based on the information specifying the physical port to be transferred acquired from the search processing unit 131. (Step S160). At this time, the transmission / reception processing circuit 120 transfers the Ethernet frame from the same physical port P3 regardless of whether the information specifying the physical port to be transferred is “P3-1” or “P3-2”.

  Here, the processing in step S150 when LAG2 is specified as the corresponding port in step S130 will be described. In such a case, the search processing unit 131a refers to the LAG setting table 137a and specifies a physical port to which the Ethernet frame is to be transferred from among the belonging ports assigned to LAG2. In this specific example, among the physical ports assigned to LAG2, P4 is a physical port to which the normal line CV4 is connected, and P3 (described as a virtual port P3-2) is connected to the standby line CV3. If the normal line CV4 has no failure, the search processing unit 131a uses the physical port P4 connected to the normal line CV4 as an Ethernet frame. Identifies the physical port to be transferred. Then, when a failure occurs in the normal line CV4, the search processing unit 131a identifies the physical port P3 (virtual port P3-2) connected to the standby line CV3 as a physical port to which the Ethernet frame is to be transferred. To do.

  As can be understood from the above description, in this embodiment, the transmission / reception processing circuit 120, the frame processing circuit 130, and the device control unit 140 correspond to the relay control unit in the claims.

  According to the present embodiment described above, since the standby line CV3 is set as a shared physical line belonging to a plurality of link aggregation groups, the number of physical lines can be achieved while realizing redundancy of the plurality of link aggregation groups. Can be reduced.

  For easy understanding, a conventional network relay system will be described. FIG. 7 is a diagram showing a configuration of a conventional network relay system. In order to make the figure easy to see, only the connection relation between the two network relay devices is shown, and other connection relations (external network etc.) are omitted. When a plurality of link aggregation groups are set between the conventional network relay device 300a and the network relay device 300b, one physical line can belong to only one link aggregation group. For this reason, as shown in FIG. 7, for example, when two link aggregation groups LAG1 and LAG2 are set and redundancy is to be secured when a physical line fails in both, a standby line is provided for each link aggregation group. It was necessary to belong. In the example shown in FIG. 7, it is understood that the standby line CV3 is prepared for the link aggregation group LAG1, and the standby line CV5 is prepared for the link aggregation group LAG2.

  However, in the present embodiment, the shared physical line CV3 can be used as a substitute line even when a failure occurs in a normal line belonging to either of the link aggregation groups LAG1 and LAG2. In addition, it is possible to reduce the number of physical lines while realizing redundancy of a plurality of link aggregation groups.

B. Second embodiment:
A network relay system according to the second embodiment will be described with reference to FIG. The device configuration and operation of the network relay system in the second embodiment are basically the same as the configuration of the network relay system 1000 in the first embodiment described with reference to FIGS. Will be explained.

The second embodiment differs from the first embodiment in the contents of the LAG setting table and step S150 (FIG. 6) in operation. FIG. 8 is a diagram showing a LAG setting table in the second embodiment. The LAG setting table 137a2 set in the first relay device 100a in the second embodiment is provided with a bandwidth usage rate setting field instead of the attribute setting field of the LAG setting table 137a in the first embodiment. Yes. In the bandwidth usage rate setting field, the bandwidth usage rate (percentage unit) of the physical line is described for each physical port, that is, for each physical line connected to each physical port. As shown in FIG. 8, the physical ports P1, P2, and P4 corresponding to the physical lines CV1, CV2, and C4 belonging to only one link aggregation group are each described as a bandwidth usage rate of 100%. On the other hand, for the physical port P3 (corresponding to virtual ports P3-1 and P3-2) corresponding to CV3, which is a shared physical line belonging to two link aggregation groups, the bandwidth usage rate in a plurality of link aggregations belonging to The bandwidth usage rate is described so that the total is 100%. In the example shown in FIG. 8, the shared physical line CV3 is set so that 50% of the bandwidth is used for the link aggregation group LAG1 and 50% of the bandwidth is used for the link aggregation group LAG2. This bandwidth usage rate can be set arbitrarily. For example, the shared physical line CV3 is set to use 80% bandwidth for the link aggregation group LAG1 and 20% bandwidth for the link aggregation group LAG2. May be.

  As shown in FIG. 8, the LAG setting table 137b2 set in the second relay device 100b in the second embodiment is a link aggregation definition (LAG setting table in the first relay device 100a that is the counterpart device). 137a2) is set appropriately so as not to contradict.

  In the second embodiment, when the physical port to which the Ethernet frame is to be transferred is specified in step S150 described above, the band usage rate column is also referred to. For example, when LAG1 is specified as the corresponding port, the search processing unit 131a refers to the LAG setting table 137a2, and describes the belonging ports P1, P2, and P3 (virtual port P3-1 assigned to LAG1). Specified) as the physical port to which the Ethernet frame should be transferred. At that time, it is assumed that the bandwidth of the physical lines CV1, CV2 connected to the physical ports P1, P2 is 100% used, and the bandwidth of the shared physical line CV3 connected to the physical port P3 is 50% usable. The physical port to which the Ethernet frame is to be transferred is determined. In other words, if each of the physical lines CV1 to CV3 is a physical line having a bandwidth of 10 Gbps (gigabit per second), the search processing unit 131 has two physical lines of 10 Gbps (CV1 and CV2) and one physical line of 5 Gbps. (CV3) is treated as constituting a link aggregation group LAG1.

Similarly, when LAG2 is specified as the corresponding port, the search processing unit 131a refers to the LAG setting table 137a2 and describes the belonging ports P4 and P3 (virtual port P3-2 assigned to LAG2). Specified) as the physical port to which the Ethernet frame should be transferred. At this time, it is assumed that the bandwidth of the physical line CV4 connected to the physical port P4 is 100% used, and the bandwidth of the shared physical line CV3 connected to the physical port P3 is 50% usable. Determine the physical port to be transferred. That is, if each of the physical lines CV3 and CV4 is a physical line having a bandwidth of 10 Gbps (gigabit per second), the search processing unit 131 has one physical line (CV4) of 10 Gbps and one physical line (CV3) of 5 Gbps. Are assumed to constitute the link aggregation group LAG2.

  In the second embodiment, when a failure occurs in any of the physical lines, the transfer is performed using only other physical lines of the link aggregation group to which the failed physical line belongs. become.

  According to the present embodiment described above, the physical line CV3 is set as a shared physical line belonging to a plurality of link aggregation groups. Then, a bandwidth ratio allocated to each link aggregation group in the bandwidth of the shared physical line CV3 is set. This makes it possible to set up a flexible link aggregation group by efficiently using resources such as a physical line and its bandwidth.

  For easy understanding, a conventional network relay system will be described. FIG. 9 is a diagram showing a configuration of a conventional network relay system. In order to make the figure easy to see, only the connection relation between the two network relay devices is shown, and other connection relations (external network etc.) are omitted. When a plurality of link aggregation groups are set between the conventional network relay device 400a and the network relay device 400b, one physical line can belong to only one link aggregation group. Therefore, for example, when it is desired to set a link aggregation group that secures a bandwidth of 25 Gbps and a link aggregation group that secures a bandwidth of 15 Gbps using a physical line having a bandwidth of 10 Gbps, as shown in FIG. The link aggregation group LAG1 of 30 Gbps and the link aggregation group LAG2 of 20 Gbps had to be set using a single physical line.

  However, in this embodiment, the shared physical line CV3 can belong to a plurality of link aggregation groups, and the bandwidth ratio can be set for each link aggregation group. Therefore, using four physical lines having a bandwidth of 10 Gbps, 25 Gbps A link aggregation group having a bandwidth and a link aggregation group having a bandwidth of 15 Gbps can be set. Of course, since the bandwidth ratio can be set freely, a link aggregation group having a bandwidth of 22 Gbps and a link aggregation group having a bandwidth of 18 Gbps can also be set. In this way, a flexible link aggregation group can be set by efficiently using resources.

C. Variations:
・ First modification:
In the above embodiment, two physical lines CV3 are set as shared physical lines. However, the present invention is not limited to this, and two or more physical lines may be set as shared physical lines. An example is shown in FIG. FIG. 10 is a block diagram showing a schematic configuration of the network relay system according to the first modification. In order to make the figure easy to see, only the connection relation between the two network relay devices is shown, and other connection relations (external network etc.) are omitted. In the network relay system 1001 according to the first modification, two link aggregation groups LAG1 and LAG2 are used between the first relay apparatus 101a and the second relay apparatus 101b using seven physical lines CV1 to CV7. Is set. The two physical lines CV4 and CV5 belong to the two link aggregation groups LAG1 and LAG2 as shared physical lines. The two shared physical lines are used as standby lines as in the first embodiment, for example. As a result, it is possible to construct a highly redundant system that can cope with a wider range of line failures with fewer lines.

・ Second modification:
One or a plurality of shared physical lines may belong to three or more link aggregation groups. In such a case, for example, if a physical port connected to a shared physical line is treated as three or more virtual ports such as P3-1, P3-2, and P3-3, and each belongs to a different link aggregation group good.

・ Third modification:
In the above embodiment, the shared physical line CV3 belongs to the two link aggregation groups LAG1 and LAG2, but the shared physical line CV3 belongs not only to the link aggregation group but also to other types of logical lines (logical ports). You may let them. An example is shown in FIG. FIG. 11 is a block diagram illustrating a schematic configuration of a network relay system according to a third modification. In order to make the figure easy to see, only the connection relation between the two network relay devices is shown, and other connection relations (external network etc.) are omitted. In the network relay system 1002 according to the third modification, a link aggregation of 25 Gbps is performed between the first relay device 102a and the second relay device 102b using three physical lines CV1 to CV3 having a bandwidth of 10 Gbps. A logical line of group LAG1 and 5 Gbps is SCV set. That is, in the shared physical line CV3, 50% of the bandwidth is assigned to the link aggregation group LAG1, and 50% of the bandwidth is assigned to the logical line SCV that is not link aggregation. In this way, it is possible to easily construct a 25 Gbps logical line and a 5 Gbps logical line using the three physical lines CV1 to CV3 having a 10 Gbps bandwidth. As a result, it is possible to construct a system that can use the bandwidth more flexibly and effectively.

-Fourth modification:
In the above embodiment, the Ethernet frame is divided into the high priority frame and the low priority frame based on the priority value described in the TOS field, and the link aggregation group used for transfer is selected according to this. The method of dividing the type of the Ethernet frame is not limited to this. For example, it is possible to divide the data into actual data and control data, and select a link aggregation group used for transfer according to this. As a method for determining the type of Ethernet frame, other information may be used instead of the information described in the TOS field. For example, information described in the destination IP address (DIP), the source IP address (SIP), and the protocol field of the IP header 820 among the information included in the Ethernet frame 800 illustrated in FIG. 3 may be used. Alternatively, information described in the traffic class field included in the IPv6 IP header may be used. Alternatively, information described in the destination MAC address (DMAC), the source MAC address (SMAC), and the ether type field of the Ethernet header 810 among the information included in the Ethernet frame 800 may be used.

-5th modification:
In the above-described embodiment, as a method of using the shared physical line CV3, a method of using as a standby line and a method of setting and using a bandwidth usage rate (bandwidth ratio) are shown, but the present invention is not limited to this. For example, the device control unit 140 may constantly monitor the frame flow rate for each link aggregation group and dynamically change the band usage rate according to the fluctuation of the frame flow rate. By so doing, it is possible to allocate the bandwidth of the shared physical line CV3 to the link aggregation group that requires more bandwidth in accordance with the temporal change in the frame flow rate, and the bandwidth can be utilized more efficiently.

-6th modification:
In the above embodiment, each of the network relay devices 100a and 100b is a device that performs layer 3 transfer, but each of the network relay devices 100a and 100b may be a layer 2 switch that performs layer 2 transfer. In such a case, instead of the routing table 135 and the address-port correspondence table 136, the network relay devices 100a and 100b, for example, based on the destination MAC address of the Ethernet frame and the COS value, A table for identifying the corresponding port is provided.

・ Other variations:
In the above embodiment and the modification, the MAC address is used as the address of the data link layer, and the IP address is used as the address of the network layer. This is because Ethernet (registered trademark) is used as the link layer protocol and IP (internet protocol) is used as the network layer protocol. Of course, when other protocols are used for the data link layer and network layer protocols, the addresses used in those protocols may be used. In such a case, the transferred data is not an Ethernet frame but data used for a data link layer protocol.

  In the above embodiments and modifications, a part of the configuration realized by hardware may be replaced with software, and conversely, a part of the configuration realized by software may be replaced by hardware.

  As mentioned above, although this invention was demonstrated based on the Example and the modification, Embodiment mentioned above is for making an understanding of this invention easy, and does not limit this invention. The present invention can be changed and improved without departing from the spirit and scope of the claims, and equivalents thereof are included in the present invention.

1 is a block diagram showing a schematic configuration of a network relay system according to a first embodiment. The block diagram which shows the internal structure of the network relay apparatus contained in the network relay system which concerns on 1st Example. The figure which shows notionally the data structure of an Ethernet frame. The figure which shows notionally the various tables stored in the 1st relay apparatus. The figure which shows notionally the various tables stored in the 2nd relay apparatus. The flowchart which shows the operation | movement step of each network relay apparatus. The figure which shows the structure of the conventional network relay system. The figure which shows the LAG setting table in 2nd Example. The figure which shows the structure of the conventional network relay system. The block diagram which shows schematic structure of the network relay system which concerns on a 1st modification. The block diagram which shows schematic structure of the network relay system which concerns on a 3rd modification.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100, 101, 102 ... Network relay apparatus 110 ... Physical port 120 ... Transmission / reception processing circuit 130 ... Frame processing circuit 131 ... Search processing part 134 ... Memory 135 ... Routing table 136 ... Address-port correspondence table
DESCRIPTION OF SYMBOLS 137 ... LAG setting table 140 ... Device control part 800 ... Ethernet frame 810 ... Ethernet header 830 ... Payload 1000, 1001, 1002 ... Network relay system

Claims (8)

A network relay device,
Multiple ports for connecting physical lines,
A relay control unit capable of constructing a plurality of logical lines using a plurality of physical lines connected to the plurality of ports, wherein a first physical line of the plurality of physical lines is a plurality of different logical lines. A relay control unit that can be assigned to
With
Each of the plurality of logical lines is a link aggregation group that logically bundles two or more lines of the plurality of physical lines and treats them as one logical line.
The relay control unit
(A) setting a bandwidth ratio for allocating the bandwidth of the first physical line to a plurality of logical lines;
(B) determining bandwidths of a plurality of link aggregation groups including the first physical line based on the bandwidth ratio;
(C) When determining a link aggregation group to be used for data transmission from among the plurality of link aggregation groups, the bandwidth of the link aggregation group to be used is changed according to the frame flow rate of each of the plurality of link aggregation groups. Network relay device. The network relay device according to claim 1,
The first physical line is assigned to a plurality of different logical lines by treating the first physical line as a plurality of virtual lines and assigning the plurality of virtual lines to different logical lines. Realized network relay device. In the network relay device according to claim 1 or 2,
The network relay device is a layer 3 switch, which is a layer 3 switch. In the network relay device according to claim 1 or 2,
The network relay device is a layer 2 switch, and is a network relay device. A method for controlling a network relay device capable of connecting a plurality of physical lines,
Assigning a first physical line of the plurality of physical lines to a plurality of different logical lines when constructing a plurality of logical lines using a plurality of physical lines;
Performing data transmission using any of the plurality of physical lines;
Prepared,
Each of the plurality of logical lines is a link aggregation group that logically bundles two or more lines of the plurality of physical lines and treats them as one logical line.
The step of transmitting the data includes
(A) setting a bandwidth ratio for allocating the bandwidth of the first physical line to a plurality of logical lines;
(B) determining a band of a plurality of link aggregation groups including the first physical line based on the band ratio;
(C) When determining a link aggregation group to be used for data transmission from among the plurality of link aggregation groups, the bandwidth of the link aggregation group to be used is changed according to the frame flow rate of each of the plurality of link aggregation groups. And a process of
Including a control method. The control method according to claim 5, wherein
The first physical line is assigned to a plurality of different logical lines by treating the first physical line as a plurality of virtual lines and assigning the plurality of virtual lines to different logical lines. Control method realized. The control method according to claim 5 or 6,
The control method, wherein the network relay device is a layer 3 switch. The control method according to claim 5 or 6,
The control method, wherein the network relay device is a layer 2 switch.

Images