JP2005333220A - Network node device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a network node device capable of facilitating the update of a path information table due to configuration changes of a network, and also, capable of reducing a packet loss. <P>SOLUTION: A path control unit 20 has the path information table 33 composed of a plurality of entries showing path information, a forwarding table 31 for outputting a different memory address corresponding to a packet destination address, and an address conversion table 32 for converting a group of the memory addresses outputted from the forwarding table to an address of a specific entry of the path information table. The network node device is constituted so that one memory address is searched from the forwarding table by using a destination address of the received packet as a search key, and the path information is searched from the path information table on the basis of the entry address converted by the address conversion table. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a network node device, and more particularly to a network node device that facilitates updating of a route information table.

  In the node device constituting the communication network, the routing information table needs to be searched in order to transfer the received packet to a specific output line corresponding to the destination address. In recent years, various high-speed route search methods have been studied in order to speed up packet processing in each node device as the amount of communication increases.

  As an example of a route information search method, for example, Japanese Patent Application Laid-Open No. 11-341076 (Patent Document 1) discloses a binary tree search in a binary tree search in which the destination address of an input packet is checked bit by bit from the upper bit. A network relay device has been proposed in which the route information search is speeded up by performing a search for p stages in a single stage.

  However, in the above prior art, the number of search stages increases in proportion to the bit length of the destination address. Therefore, when the destination address is an IPv6 packet with 128 bits, a route search is required compared with an IPv4 packet with a destination address of 32 bits. There is a problem that time increases remarkably.

  In order to search for route information at high speed regardless of the IP version of the received packet, it has been proposed, for example, in IEICE 2000 General Conference Proceedings, SB-4-2 (Non-patent Document 1). As described above, a two-step table search method for searching a search result holding table formed in a RAM (Random Access Memory) using an output address of an associative memory (CAM) in which a search condition is described is effective.

  In Non-Patent Document 1, a table showing flow identification conditions and forwarding conditions is configured with an associative memory, and a plurality of information items necessary for collation with the flow identification conditions and the forwarding conditions are extracted from the header of the received packet. The associative memory is accessed using the information item as a search key, and the flow control information is read from the search result holding table according to the table address output from the associative memory.

Japanese Patent Laid-Open No. 11-341076 “Network Relay Device and Network Next Transfer Destination Search Method”

Uga et al., “Flow Identification Method Using Associative Memory”, Proceedings of the 2000 IEICE General Conference, SB-4-2, March 2000

FIG. 6 shows an example of a path information search unit including a forwarding table 30 formed with an associative memory and a path information table 33 formed with RAM.
The forwarding table 30 includes a plurality of destination entries DE-i (i = 1 to N) indicating entry addresses 302 corresponding to the search conditions 301. The path information table 33 includes a plurality of entries TE-i (i = 1 to N) having a memory address 331, and path information 322 is stored in each entry.

  In the case of route information search, a packet destination address value is set as a search condition 301 in the forwarding table 30. When the associative memory (forwarding table 30) is accessed using the destination address extracted from the header of the received packet as a search key, the entry address 302 of the destination entry DE-k having the search condition 301 that matches the search key is obtained. Therefore, by accessing the route information table 33 using the entry address 302 as a read address, the route information 332 stored in the entry TE-k of the route information table 33 can be read.

  However, as described below, each node constituting the network needs to dynamically update the route information stored in the route information table in accordance with the change in the network state.

  FIG. 7 shows an example of a communication network including a plurality of node devices each having a route information search unit. Here, a network in which local networks 210, 220, 230, and 240 of a plurality of sites (A 1, A 2, A 3, A 4) belonging to the organization A are interconnected via the public IP network 200 is shown.

  The site A1 is composed of the node device 211 that accommodates the terminals 212 and 213, the site A2 is the node device 221 that accommodates the terminals 222 and 223, the site A3 is the node device 231 that accommodates the terminals 232 and 233, and the site A4 is The node device 241 accommodates the terminals 242, 243, and 244.

  The public IP network 200 includes a node device 202 connected to the node device 211 of the site A1, a node device 203 connected to the node device 241 of the site A4, a node device 221 of the site A2, and a node device 231 of the site A3. It comprises a connected node device 205 and node devices 201 and 204 that connect these node devices.

  The forwarding table 30 and the route information table 33 described with reference to FIG. 6 correspond to the route information search unit of the node device 202 in the network. The route information 322 includes, for example, a next transfer destination MAC address (Next Hop Address), an output port number, transfer control information indicating whether packet transfer is necessary, and the like. In FIG. Route information applied to the node is denoted as RT201, and route information applied to a packet passing through the node device 203 is denoted as RT203. The destination address DA set as the search condition 301 in the forwarding table 30 is expressed as DAxxx with a destination terminal number xxx.

  In the forwarding table 30 included in the node device 202, a plurality of destination entries DE-1 to DE-N corresponding to the value of the destination address included in the received packet from the site A1 are registered, and the route information table 33 also includes the destination entry. And the same number of path information entries TE-1 to TE-N. In this case, since the route information table 33 has a route information entry for each destination address, there are a plurality of route information entries having the same contents corresponding to a plurality of destination addresses having the same next transfer destination node device. become.

  Here, in the network shown in FIG. 7, it is assumed that a failure has occurred in the link between the node device 201 and the node device 205, and packet transfer from the node device 201 to the node device 205 becomes impossible. In this case, in the node device 202, the route is set so that the transmission packet addressed to the site A2 or the site A3 from the site A1 that has been routed through the node device 201 so far reaches the node device 205 via the node devices 203 and 204. It is necessary to change the contents of the information table 33.

  That is, when a failure occurs in the transfer route via the node device 201, the node device 202, as shown by a block 333 in FIG. 6, all the path information entries TE whose next transfer destination MAC address is the failed node device 201. For -1 to TE-4, the route control information needs to be rewritten from RT201 to RT203.

  When there are a plurality of entries whose contents should be changed in this way, it takes time to update the route information table. As a result, a packet erroneously transferred to the failed route is generated by the route control information RT201 before the update. There is a problem that the opportunity for loss increases. This problem occurs even when, for example, a node device address that is a prefix of a plurality of terminal addresses belonging to each site is used as a search condition of the forwarding table 30 in order to reduce the number of entries in the routing information table. To do.

An object of the present invention is to provide a network node device that facilitates updating a route information table accompanying a change in network configuration and reduces the occurrence of packet loss.
Another object of the present invention is to reduce the time required to update the route information table accompanying a change in the network configuration even when the entry size of the route information table is increased, and reduce erroneous transfer of packets during the update period. It is to provide a node device.

In order to achieve the above object, a network node device according to the present invention includes a plurality of interface modules connected to input and output lines, and a route for searching for route information based on a destination address of a packet received by each interface module. A control unit, and a packet relay unit that forwards the received packet to any interface module specified by the route information searched by the route control unit,
The route control unit includes a route information table including a plurality of entries indicating route information, a forwarding table that outputs different memory addresses according to a packet destination address serving as a search condition, and a group output from the forwarding table And an address conversion table comprising a plurality of entries for converting the memory address of the route information table into the address of the specific entry, and using the destination address of the received packet as a search key, one memory address is obtained from the forwarding table. Searching, converting the memory address into an entry address of the routing information table by the conversion table, and searching the routing information from the routing information table based on the entry address.

  More specifically, the route control unit generates an internal header based on the route information retrieved from the route information table, transfers the received packet to the packet relay unit with the internal header added, and transmits the packet relay. The unit transfers the received packet to the interface module specified by the internal header.

  In one embodiment of the present invention, the network node device temporarily stores received packets output from the interface modules, and selectively receives the received packets output from the interface modules. A selector for inputting to the memory; and a buffer control unit for controlling writing and reading of the received packet to and from the buffer memory. The path control unit cooperates with the buffer control unit to relay the received packet as a packet relay. It transfers to a part.

  Further, in one embodiment of the present invention, the routing control unit provides a forwarding table control unit that searches for one memory address from the forwarding table using the destination address of the received packet as a search key, and the forwarding table control unit. A conversion table control unit that retrieves an entry address of the route information table from the conversion table based on the given memory address, and route information from the route information table based on the entry address given from the conversion table control unit A route information table control unit that searches for a packet, and a packet transfer control unit that generates the internal header based on the route information given from the route information table control unit, adds the internal header to a received packet, and outputs the packet. It is characterized by comprising.

Furthermore, in another embodiment of the present invention, the route information table is composed of a plurality of sub-tables, each entry of the address translation table is a group of memory addresses output from the forwarding table, a specific entry of each sub-table. Including a plurality of addresses.
Other problems, features, and operation modes to be solved by the present invention will become apparent from the description of embodiments with reference to the drawings.

  According to the present invention, because the address conversion table associates a plurality of destination addresses having the same next transfer destination node device with a specific entry in the route information table, the route information is changed when a failure occurs in the network. The number of target route information entries can be greatly reduced. Therefore, according to the present invention, since the rewriting of the route information table accompanying the change in the network configuration can be completed instantaneously, a high-quality network in which the number of lost packets in the failed route due to the application of the route information before update is reduced is provided. it can.

  In addition, when the address conversion table is adopted, a plurality of entry addresses can be generated corresponding to the destination address. Therefore, a substantial entry size can be obtained by forming a plurality of route information tables on a RAM having a limited entry size. And the amount of route information can be increased. Therefore, according to the present invention, advanced packet transfer using abundant route information becomes possible.

  Embodiments of the present invention will be described below with reference to the drawings. Here, in order to facilitate comparison with the prior art, a case where the present invention is applied to the node device 202 of the network of FIG. 7 will be described.

FIG. 1 is a block diagram showing an embodiment of a network node device 202 according to the present invention.
The network node device 202 includes a plurality of interface modules 11-i (i = 1 to N) connected to the input line IN-i and the output line OUT-i, and the selector 12 and the packet connected to these interface modules. Relay unit 16, buffer memory 13 connected to selector 12, header extraction unit 14 and buffer control unit (hereinafter referred to as R / W control unit) 15, route control unit 20 connected to packet relay unit 16, Consists of. Reference numeral 60 denotes a management terminal which is located outside the network node device 202 and has the signal line L7 connected to the path control unit 20.

  FIG. 2 shows an example of a format of a variable length MAC frame received from each input line IN-i. The variable-length MAC frame includes an IP packet 100 in the third layer (network layer) and an L2 header 110 in the second layer (data link layer) in the OSI reference model. Reference numeral 120 denotes an internal header added by the route control unit 20 when the IP packet is transferred from the route control unit 20 to the packet relay unit 16.

  The IP packet 100 includes a payload 101 and an IP header 102. The IP header 102 includes control information 103, a source IP address (SA) 104, and a destination IP address (DA) 105. Here, the control information 103 means IP header information other than SA and DA, such as a protocol type, a packet length, and a service type indicating packet priority.

  The format of the L2 header 110 differs depending on the type of input line. For example, when the input line IN-i is Ethernet (registered trademark), the L2 header 110 includes a source MAC address, a destination MAC address, and a frame type. , Other information included.

  In FIG. 1, each interface module 11-i terminates the MAC frame received from the input line IN-i and outputs the received IP packet 100 to the signal line L1-i. The selector 12 selectively outputs the received IP packets input from the signal lines L1-1 to L1-N to the signal line L2 in accordance with a control signal given from the R / W control unit 15 via the signal line L15. The header extraction unit 14 analyzes the IP header 102 of the IP packet 100 output to the signal line L2, and outputs the packet length to the signal line L3 and the destination IP address DA to the signal line L4.

  The R / W control unit 15 writes the IP packet 100 output to the signal line L2 to the buffer memory 13 and reads the IP packet from the buffer memory 13 to the signal line L6. Further, the R / W control unit 15 controls the writing of the IP packet to the buffer memory 13 based on the packet length output from the header extraction unit 14 to the signal line L3, and interlocks with the reading of the packet from the buffer memory 13. Then, a switching control signal for the selector 12 is generated on the signal line L15, and the next packet is written in the buffer memory 13.

  As will be described in detail with reference to FIG. 4, the route control unit 20 searches the route information table for route information based on the destination IP address (DA) output from the header extraction unit 14 to the signal line L4. The route information includes, for example, a MAC address indicating the next transfer destination node device of the packet, an output port number, and transfer control information. The path control unit 20 outputs an internal transmission packet having the internal header 120 including the path information added to the head of the IP packet 100 input from the signal line L6 to the signal line L8.

  Specifically, the path control unit 20 sends the internal header 120 including the path information retrieved from the path information table to the signal line L8, and at a predetermined timing via the signal line L5, the R / W control unit. 15 is requested to read the IP packet. When the R / W control unit 15 reads the IP packet from the buffer memory 13 to the signal line L6, the path control unit 20 sends the IP packet 100 to the signal line L8 following the internal header 120.

  When receiving the internal transmission packet from the signal line L8, the packet relay unit 16 transfers it to the interface module 11-k corresponding to the output port number k indicated by the internal header 120. When receiving the internal transmission packet from the packet relay unit 16, the interface module 11-k separates the internal header 120, adds the L2 header 110 to the IP packet 100, and transmits it as a MAC frame to the output line OUT-k. The MAC address of the interface module 11-k is set as the source MAC address of the L2 header 110, and the next transfer destination MAC address indicated by the internal header 120 is set as the destination MAC address.

FIG. 3 shows the relationship among the forwarding table 31, the address translation table 32, and the route information table 33 applied to the route control unit 20 of the present invention.
In the present invention, an address for associating a plurality of entries in the forwarding table 31 with one entry in the routing information table 33 between the forwarding table 31 composed of the associative memory and the routing information table 33 formed by the RAM. The conversion table 32 is arranged.

  Similar to the forwarding table 30 described with reference to FIG. 6, the forwarding table 31 includes a plurality of entries DE- 1 to DE-N indicating the destination address of a packet as the search condition 311. However, each entry address 312 corresponds to an address 321 in the address conversion table 32.

  The address conversion table 32 includes a RAM and includes the same number of entries PE-1 to PE-N as the forwarding table 31. Each entry PE-i (i = 1 to N) has a memory address 321 and stores, for example, a packet attribute 322A indicating the IP version and packet transfer format, and a path information entry address 322B as the path information identifier 322. doing.

  The path information table 33 includes a plurality of entries TE-1 to TE-M each having a memory address 331. Each entry includes, for example, a next transfer destination MAC address 332A, an output port number 332B, and transfer control information as path information 332. 332C is stored.

  In FIG. 3, a plurality of entries (for example, TE-1 to TE-4) having the same control information exist in the route information table 33. As will be apparent from the following description, the present invention Then, it is possible to reduce the number of registered entries in the routing information table 33 to be smaller than the number of entries in the forwarding table 31 by representing the plurality of entries overlapping in contents as one entry.

  In the present invention, among the entries registered in the address conversion table 32, a plurality of destination addresses in the forwarding table 31 are provided by giving the same route information entry address 322B to a group of entries having the same next transfer destination node device. Is associated with the same entry in the route information table 33.

  Specifically, for the received packet whose next transfer destination is the node device 201, the address conversion table 32 is accessed based on the output address 312 of the forwarding table 30 regardless of the destination address of DA222 to DA233. At this time, the same route information entry address 322B, in this example, the entry address “1” is obtained, and packet transfer is performed according to the specific entry in the route information table 33, in this example, the route information RT201 indicated by the entry TE-1. Is called. The same applies to the received packets of the destination addresses DA242 and DA243, and packet transfer is performed according to a specific entry of the route information table 33, in this example, the route information RT203 read from the entry TE-5.

  In this manner, by interposing the address conversion table 32 between the forwarding table 31 and the route information table 33 and aggregating output addresses of a plurality of entries in the forwarding table 31 into one route information entry in the route information table 33, For example, when a situation where the route information RT201 should be changed to the route information RT203 occurs, the rewriting of the route information table can be completed in a short time.

FIG. 4 shows an embodiment of the route control unit 20.
The route control unit 20 includes a forwarding table control unit 21 for referring to the forwarding table 31, an address conversion table control unit 22 for referring to the address conversion table 32, and a route information table for referring to the route information table 33. It comprises a control unit 23, a packet transfer control unit 24, and a table update control unit 25.

  The entry setting / updating of the tables 31 to 33 is performed by the table update control unit 25. In the case of dynamic routing, the table update control unit 25 automatically sets / updates a table entry according to a predetermined route control protocol (software). In the case of static routing, the system administrator operates the management terminal 60 and gives a control command and data for setting / updating table entries to the table update control unit 25 via the signal line L7.

  The forwarding table control unit 21 accesses the forwarding table (associative memory) 31 using the destination IP address input from the signal line L4 as a search key, and outputs the entry address 312 searched from the forwarding table 31 to the signal line L21. .

  The address translation table control unit 22 accesses the address translation table 32 using the entry address 312 input from the signal line L21 as a read address, and uses the packet attribute 322A and the path information entry address 322B read from the address translation table 32 as the read addresses. The signals are output to the signal lines L22 and L220, respectively.

  The path information table control unit 23 accesses the path information table 33 using the path information entry address 322B input from the signal line L220 as a read address, and outputs the packet attribute 322A received from the signal line L22 to the signal line L23. The route information 332 read from the route information table 33 is output to the signal line L230.

  When receiving the packet attribute 322A and the route information 332 from the route information table control unit 23, the packet transfer control unit 24 generates the internal header 120 including the route information and outputs it to the signal line L8. However, since the arrangement order of the next transfer destination MAC address 332A, the output port number 332B, and the transfer control information 332C included in the path information 332 varies depending on the IP version of the received packet, the packet transfer control unit 24 sets the packet attribute 322A. To identify the configuration of the route information 332 and generate an internal header.

  The packet transfer control unit 24 requests the R / W control unit 15 to read the IP packet via the signal line L5 at a predetermined timing, and outputs the IP packet received from the signal line L6 to the signal line L8. At this time, part of the header information of the IP packet received from the signal line L6 is rewritten as necessary. Note that the packet transfer control unit 24 determines the transfer control information 332C that is a part of the route information, and if the packet transfer control unit 24 indicates that the packet transfer is not required, the packet transfer control unit 24 sends an IP address to the R / W control unit 15 without generating an internal header. Requests reading of the packet, and discards the IP packet received from the signal line L6.

  FIG. 5 shows another embodiment of the address conversion table 32 and the route information table 33 referred to by the route control unit 20. In the present embodiment, the forwarding table 31 is the same as that shown in FIG. This embodiment is characterized in that two entry addresses 322B and 322C are registered in the address conversion table 32, and the first route information table 33 and the second route information table 34 are accessed according to these entry addresses. .

  For example, in an MPLS (Multi Protocol Label Switching) network, a packet is transferred according to label information added to the packet by an edge node. Depending on the packet transfer path, it is necessary to store a plurality of labels corresponding to one destination address, for example, a label for transfer path identification and a label for VLAN identification. If such MPLS label information is to be read from the route information table as part of the route information, the entry size of the route information table needs to be expanded.

  However, since there is a restriction on the memory block size accessible by one memory address in the RAM forming the path information table, the path information table 33 has an upper limit on the data size of path information that can be stored in each table entry. There is.

  In this embodiment, in order to expand the route information entry size, the route information table is divided into a first route information table 33 and a second route information table 34, and according to the first entry address 322B indicated by the address conversion table 32. The first route information table 33 is accessed, and the second route information table 34 is accessed according to the second entry address 322C indicated by the address conversion table 32.

  In the first route information table 33, for example, the route information items 332A to 332C shown in FIG. 3 added with the MPLS label 1 (332D) are registered as the first route information 332, and the second route information table 34 is registered. In this example, MPLS label 2 (342A) and MPLS label 3 (342B) are registered as the second route information 342. In this case, for example, an identification bit is added to the packet attribute 322A of the address conversion table 32, and whether or not valid data (label information) exists in the second route information table 34 is displayed according to the state of the identification bit. .

  In the first route information table 33, only the route information items 332A to 332C are registered as the first route information 332, and all the MPLS labels are set as the second route information 342 in the second route information table 34. You may make it register.

  In this way, by storing a plurality of entry addresses indicating specific entries of the route information table divided into a plurality of sub-tables in each entry PE-i (i = 1 to N) of the address conversion table 32. Corresponding to the search condition 311 indicated by the forwarding table 31, it becomes possible to read out the route information with the expanded information amount.

  Further, as shown in FIG. 5, the first route information table 33 stores basic route information, and the second route information table 34 stores exceptionally necessary route information. By making it possible to identify whether valid information exists in the second route information table 34 from the packet attribute 322A stored in the address conversion table 32, unnecessary access to the second route information table 34 is avoided. In addition, high-speed route information search becomes possible.

  When the embodiment of FIG. 5 is adopted, in the path control unit 20 of FIG. 4, for example, the signal line L220 between the address conversion table control unit 22 and the path information table control unit 23 is used as the signal line for the first entry address. And the second entry address signal line, and the address conversion table control unit 22 outputs the first entry address and the second entry address in parallel.

  The route information table control unit 23 reads the first route information from the first route information table 33 using the first entry address. On the other hand, the reading of the second route information from the second route information table 34 by the second entry address is selectively performed and read according to the state of the identification bit indicated by the packet attribute input from the signal line L22. The first and second route information is supplied to the packet transfer control unit 24 via the signal line L230. In this case, the packet transfer control unit 24 can determine whether or not the second route information is received following the first route information according to the state of the identification bit indicated by the packet attribute received from the signal line L23.

  In the above embodiment, the route control unit 20 searches for route information using the destination address (IP address) of the third layer of the OSI reference model as a search key. For example, in addition to the layer 2 switch function, When the present invention is applied to a node device operating as a layer 3 switch having a packet relay function, a destination address of the second layer (MAC) may be applied as a search key for a forwarding table.

  As is clear from the above description, according to the present invention, in the network node device such as the IP packet router, the layer 3 switch, etc., it is possible to speed up the update of the route information table, so that it dynamically adapts to the network configuration change. Can provide communication services.

The block diagram which shows one Example of the network node apparatus by this invention. The figure which shows an example of the format of the variable length MAC frame input into a network node apparatus. The figure which shows the relationship between the forwarding table 31, the address conversion table 32, and the path | route information table 33 with which the path | route control part 20 is provided in the network node apparatus of this invention. The block diagram which shows one Example of the route control part 20. FIG. The figure which shows the other Example of the address conversion table 32 with which the route control part 20 is provided, and a route information table. The figure which shows the relationship between the forwarding table 30 and the route information table 33 by a prior art. The figure which shows the structural example of the IP network to which the network node apparatus of this invention is applied.

Explanation of symbols

11: Interface module, 12: Selector, 13: Buffer memory,
14: header extraction unit, 15: R / W control unit, 16: packet relay unit,
20: path control unit, 21: forwarding table control unit,
22: Address conversion table control unit, 23: Path information table control unit,
24: Packet transfer control unit, 25: Table update control unit,
30, 31: Forwarding table, 32: Address conversion table,
33, 34: Route information table, 60: Management terminal.

Claims (7)

A plurality of interface modules connected to the input and output lines, a route control unit that searches for route information based on the destination address of the packet received by each interface module, and the route control unit searches for the received packet. In a network node device comprising a packet relay unit that transfers to any one of the interface modules specified by the route information,
The route controller is
A route information table including a plurality of entries indicating route information;
A forwarding table that outputs different memory addresses depending on the packet destination address that is the search condition;
An address conversion table comprising a plurality of entries for converting a group of memory addresses output from the forwarding table into addresses of specific entries of the routing information table;
Using the destination address of the received packet as a search key, one memory address is searched from the forwarding table, the memory address is converted to the entry address of the route information table by the conversion table, and the route information is based on the entry address. A network node device characterized by retrieving route information from a table. The path control unit generates an internal header based on the path information retrieved from the path information table, and forwards the received packet to the packet relay unit with the internal header added;
The network node device according to claim 1, wherein the packet relay unit transfers the received packet to an interface module specified by the internal header. The route control unit is
A forwarding table control unit that searches for one memory address from the forwarding table using the destination address of the received packet as a search key;
A conversion table control unit that searches the conversion table for an entry address of the route information table based on the memory address given from the forwarding table control unit;
A route information table control unit that searches for route information from the route information table based on the entry address given from the conversion table control unit;
3. A packet transfer control unit that generates the internal header based on path information given from the path information table control unit, adds the internal header to a received packet, and outputs the packet. The network node device described in 1.   The network node device according to claim 3, wherein the route control unit includes means for updating the contents of the route information table. A buffer memory for temporarily storing received packets output from the interface modules;
A selector for selectively inputting the received packet output from each interface module to the buffer memory;
A buffer control unit that controls writing and reading of received packets to and from the buffer memory;
5. The network node device according to claim 1, wherein the path control unit forwards a received packet to the packet relay unit in cooperation with the buffer control unit.   6. The network node device according to claim 1, wherein the forwarding table is configured by an associative memory. The route information table is composed of a plurality of sub-tables,
7. Each of the entries in the address conversion table includes a group of memory addresses output from the forwarding table, and includes a plurality of addresses indicating specific entries in the sub-tables. A network node device as described above.

Images