JP2007053467A - Communication system, router position calculating apparatus and program, and recording medium - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a communication system capable of dynamically establishing a disjointed multicast tree (so independent as to have no common router on a route), and to provide a router position calculating apparatus, a router position calculating program and a recording medium. <P>SOLUTION: Two sub-nets (network addresses NA1 and NA2) are set for a physical link for connecting a downlink router X to an uplink router A, and two sub-nets (network addresses NB1 and NB2) are set for a physical link for connecting the downlink router X to an uplink router B. NA1 and NB1 are assigned for transfer of main multicast packets, and NA2 and NB2 are assigned for transfer of backup multicast packets. NA1 and NB1, and NA2 and NB2 satisfy a predetermined relation (e. g. NA1>NB1 and NA2<NB2), respectively. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a communication system including a plurality of routers that transfer a broadcast main multicast packet transmitted by a distribution server and a backup multicast packet transmitted by a backup server. The present invention also relates to a router position calculation device that performs processing related to the logical positional relationship of routers constituting the communication system, a router position calculation program, and a recording medium on which the router position calculation program is recorded.

  In order to realize services such as video distribution, multicast (see Non-Patent Document 1) is defined in IP (Internet Protocol) of OSI (Open Systems Interconnection) Layer 3, and various protocols (Non-Patent Document 2 and 3) is implemented and operated. However, there is a problem that a communication failure of several seconds to several tens of seconds is involved in detection and recovery of a network failure (based on the IP routing protocol) in the IP layer. Therefore, in order to compensate for this communication interruption time and make the service uninterrupted for an IP network having a redundant configuration, the present inventors have made Japanese Patent Application Nos. 2004-060528, 2004-349665, and 2005. -062293 is proposed.

The multicast distribution route is determined based on unicast route information from the multicast receiver to the caller (or a special node called a rendezvous point (RP)) (see Non-Patent Document 2). Do. Unicast path information is generated by a unicast path control protocol such as OSPF (Open Shortest Path First) (see Non-Patent Document 3).
Beau Williamson, “IP Multicast Network Development Guide Vo1.1”, Softbank Publishing Co., Ltd., July 3, 2001 D.Estrin, 9 others, RFC2362-Protocol Independent Multicast-Sparse Mode (PIM-SM): Protocol Specification, [online], 1998, The Internet Society, [searched July 6, 2005], Internet <URL: http://www.faqs.org/rfcs/rfc2362.html> J.Moy, RFC2328-OSPF Version 2, [online], 1998, The Internet Society, [searched July 6, 2005], Internet <URL: http://www.faqs.org/rfcs/rfc2328.html > Load Splitting IP Multicast Traffic, [online], 1998, The Internet Society, [Search July 6, 2005], Internet <URL: http://www.cisco.com/univercd/cc/td/doc/product /software/ios124/124cg/himc_c/ch05/mcbsplit.htm>

  FIG. 10 shows the configuration of a communication network according to the technique proposed by the present inventor. The broadcast distribution server 101 transmits a multicast packet in which stream data to be distributed is stored. The transmitted multicast packet is transferred to the client terminal 110 via the L2 switch 102 and each communication device (routers 104b, 105b, 106b, 107 and L2 switch 109) as a normal distribution (main) multicast Gm. The multicast packet is duplicated by the L2 switch 102 and transmitted to the backup server 103. The backup server 103 duplicates the received multicast packet and transmits it as two preliminary delivery (backup) multicasts Gb1 and Gb2.

  A certain delay is added to the backup multicast Gb2 by the backup server 103. The backup multicasts Gb1 and Gb2 are discarded by the router 104a at the normal time. A communication control device 108 is provided in the vicinity of the router 107 at the edge closest to the client terminal 110. When the communication control device 108 detects the distribution interruption of the main multicast Gm, the communication control device 108 participates in the reception of the backup multicasts Gb1 and Gb2. The backup multicasts Gb1 and Gb2 are received by the communication control device 108 from the router 104a via the routers 105a, 106a, and 107.

  The communication control device 108 converts the received backup multicasts Gb1 and Gb2 into main multicasts and outputs them as a proxy (multicast Gm ′ in FIG. 10). At this time, in order to avoid the failure point (router link), the distribution route of the main multicast and the backup multicast needs to be disjoint (not sparse to each other = not via the same router link).

  According to Non-Patent Document 2, a multicast tree is constructed hop-by-hop by each router according to unicast route information from a receiver to a caller (or RP). Therefore, if a plurality of multicast trees are set from a certain receiver to the same (or network topology adjacent) sender, there is a tendency to go through the same router link. That is, in normal operation, as shown in FIG. 11, the main multicast and the backup multicast go through the same router.

  In a network topology with a redundant configuration that is targeted by this technology, there are multiple equal-cost unicast paths (equal-cost RPF (Reverse Path Forwarding) paths), so there are at least two types of inherently disjoint multicast trees. It is an environment that can be constructed. However, in a general PIM-SM implementation, when there are multiple equal-cost RPF paths (ie upstream routers), the upstream router with the largest address is uniquely selected (Non-Patent Document 1). (See FIG. 11).

  On the other hand, from the viewpoint of load distribution, there is an implementation of PIM-SM that selects different upstream routers by hash calculation using a source IP address as a key (see Non-Patent Document 4). However, this function does not guarantee disjoint tree configuration.

  The present invention has been made in view of the above-described problems, and is a communication system and a router location calculation device capable of dynamically constructing a disjoint (independent) router-free multicast tree. An object is to provide a router position calculation program and a recording medium.

  The present invention has been made to solve the above-described problems, and the invention according to claim 1 is a communication system including a plurality of routers having a hierarchical structure for transferring multicast packets transmitted by a distribution server. In any router except the router closest to the distribution server, the first and second subnets are set for the physical link connected to the first upstream router provided on the distribution server side. And the third and fourth subnets are set for the physical link connected to the second upstream router provided on the distribution server side, and the first and third subnets are the first Assigned to transfer the second multicast packet, and the second and fourth subnets are used to transfer the second multicast packet. Ri devoted, said first, said second, said third, and network address of the fourth subnet is a communication system characterized by a predetermined relationship.

  The invention according to claim 2 is the communication system according to claim 1, wherein the network address of the first subnet is NA1, the network address of the second subnet is NA2, the third network address is NB1, The fourth network address is NB2, and the predetermined relationship is NA1> NB1 and NA2 <NB2 or NA1 <NB1 and NA2> NB2.

  According to a third aspect of the present invention, in the communication system according to the first or second aspect, the logical addition position when an arbitrary router other than the router closest to the distribution server is added has a hierarchical structure. Network topology information indicating the logical positional relationship of the plurality of routers having the above, and the logical position of the upstream router and the additional router provided on the distribution server side connected to the newly added additional router It is calculated based on the additional router connection information indicating the relationship and the identification value assigned to each of the plurality of routers corresponding to the logical position of each router.

  The invention according to claim 4 is a communication system comprising a plurality of routers having a hierarchical structure for transferring multicast packets transmitted by the first distribution server and the second distribution server, wherein the plurality of routers The multicast packet transmitted by the first distribution server is forwarded from a first upstream router provided on the first distribution server side by hash processing of each router constituting the first distribution server. An identification value of the distribution server has been determined, and a multicast packet transmitted by the second distribution server is provided on the second distribution server side by hash processing of each router constituting the plurality of routers. An identification value of the second distribution server is determined to be transferred from the second upstream router, and the first distribution server And a logical addition position when an arbitrary router other than the router closest to the second distribution server is added, network topology information indicating a logical positional relationship of the plurality of routers having a hierarchical structure; Additional router connection information indicating a logical positional relationship between the upstream router and the additional router provided on the side of the first distribution server and the second distribution server, which is connected to the newly added additional router; The communication system is calculated based on an identification value assigned to each of the plurality of routers corresponding to a logical position of each router.

  According to the fifth aspect of the present invention, network topology information indicating a logical positional relationship between a plurality of routers having a hierarchical structure for transferring multicast packets transmitted by a distribution server, and connection with a newly added additional router are provided. The additional router connection information indicating the logical positional relationship between the upstream router and the additional router provided on the distribution server side, and the logical position of each router for each of the plurality of routers A router position calculation device that calculates a logical position of the additional router based on the identification value assigned in this way.

  According to the sixth aspect of the present invention, network topology information indicating a logical positional relationship between a plurality of routers having a hierarchical structure, which forwards multicast packets transmitted by a distribution server, and a connection with a newly added additional router The additional router connection information indicating the logical positional relationship between the upstream router and the additional router provided on the distribution server side, and the logical position of each router for each of the plurality of routers A router position calculation program for causing a computer to execute a process of calculating a logical position of the additional router based on the identification value given in the above.

  The invention described in claim 7 is a computer-readable recording medium in which the router position calculation program according to claim 6 is recorded.

  According to the present invention, it is possible to dynamically construct a disjoint multicast tree.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. First, the following assumptions are assumed. An example of the network configuration is the same as that in FIG. 10, and is a communication system including a plurality of routers for transferring broadcast multicast packets transmitted by the distribution server and backup multicast packets transmitted by the backup server. The communication system has a hierarchical network topology such as a multicast source (source host or RP (Rendezvous Point), DC (Data Center) edge router, core router, customer side edge router). The source (root) and the backup multicast source (root) are each only one fixed node, and there are two equal-cost upstream routers from any router (section) to the root. Then, join (join multicast packet reception) toward the upstream router having the largest interface address (see Non-Patent Document 1 described above).

(Another upstream router selection method)
In an environment that satisfies the above, each downstream router sets a plurality of logical connections to each of the two upstream routers so that the main multicast and the backup multicast can be acquired from different upstream routers. Hereinafter, the principle of the setting method will be described with reference to FIG. It is assumed that the network administrator makes settings for the routers constituting this communication system in the following procedure.

1. Two (logical) subnets (network addresses: NA1 and NA2) are set for the physical link connecting the downstream router X and the upstream router A.
2. Two (logical) subnets (network addresses: NB1 and NB2) are set for the physical link connecting the downstream router X and the upstream router B.
3. Classify subnets for unicast route calculation into two groups. NA1 and NB1 belong to group 1, and NA2 and NB2 belong to group 2.
4). The subnet to which the main multicast source belongs belongs to group 1, and the subnet to which the backup multicast source belongs belongs to group 2.

  5. Group 1 and group 2 perform unicast route calculation independently. Thus, the unicast route from the router X to the main multicast source is NA1 and NB1, and the unicast route to the backup multicast source is NA2 and NB2. Thus, NA1 and NB1 are assigned for transferring the main multicast packet, and NA2 and NB2 are assigned for transferring the backup multicast packet.

  6). For NA1 and NB1 and NA2 and NB2, the magnitude relationship of the addresses is reversed (e.g. NA1> NB1, NA2 <NB2). As a result, the upstream routers to be joined (participation in multicast packet reception) are separated into main and backup (e.g. main is upstream router A (NA1), backup is upstream router B (NB2)). Hereinafter, a physical link in which a tree toward the main multicast source (first tree) is set is defined as a first link, and a physical link in which a tree toward the backup multicast source (second tree) is set is defined as a second link. When there are two upstream router candidates that are equal in cost to two multicast sources, the route toward each multicast source can be branched by the above method.

In addition to the above-described method, the “upstream router selection method based on hash calculation using a multicast source IP address as a key” shown in Non-Patent Document 4 may be used to set the first link and the second link. Good. That is, assuming that the hash function used in the router is unique, the same effect as in FIG. 2 can be obtained by doing the following.
1. The tree toward the main multicast source (first distribution server) selects the first upstream router (first link) by the hash processing of each router (the multicast packet transmitted by the main multicast source is The IP address of the main multicast source is determined to be forwarded from the first upstream router to each router).
2. As a result of the hash processing of each router, the tree toward the backup multicast source (second distribution server) selects the second upstream router (second link) (the multicast packet transmitted by the backup multicast source is To determine the IP address of the backup multicast source.

  By applying one of the two methods described above as another upstream router selection method and introducing the following downstream router arrangement method, any router except for the router directly connected to the multicast source can be selected. A disjoint multicast tree can be constructed from the router toward the two multicast sources.

(Downstream router placement method)
Furthermore, the selection policy of the upstream router is made to coincide with all the downstream routers. In addition, the main multicast tree and backup multicast can be used in the network topology that satisfies the preconditions described at the beginning by appropriately determining the logical position of the downstream router to be added according to the following policy. You can set the tree to disjoint. Hereinafter, the downstream router addition algorithm will be described with reference to FIG.

1. In each hierarchy l (l ≧ 0), each router is 3. Based on the policy shown below, it shall be arranged in order from the left side on the network topology.
2. Assume that there are two upstream (l = 0) routers connected to a multicast source as an initial state.
3. For the downstream router to be added, the main multicast is acquired from the upstream router on the right side (first upstream router) and the backup multicast is acquired from the upstream router on the left side (second upstream router) according to the setting method described with reference to FIG. Set up.

4). In order to uniquely represent each router, the level 1 where the router is located and the order j from the left in the level 1 are used. That is, the j (j ≧ 0) -th router from the left in the hierarchy l is described as R [l, j].
5. For routers in layer 0 (l = 0), the router that accommodates the main multicast source is defined as R [0, 1], and the router that accommodates the backup multicast source is defined as R [0, 0].
6). Consider a case where a downstream router R [l, j] is newly added to a certain hierarchy l (l ≧ 1). First, another router belonging to the same layer as the new downstream router R [l, j] is described as R [l, i]. Here, i (i ≧ 0) represents the router order (order before addition of R [l, j]) counted from the left side on the layer l. However, i = 0 is the leftmost side.

7). The value of j (meaning jth from the left in hierarchy l) is determined so as to satisfy the following condition. However, j = 0 is the leftmost side. The second and first upstream routers of R [l, j] are R [l-1, j '] and R [l-1, j "], respectively, and the second and first upstreams of R [l, i] Assuming that the routers are R [l-1, i '] and R [l-1, i "], respectively, for all i in layer l and its associated i', i" in layer l-1, J is determined according to the following conditions.
(I '<j') or (i '= j' and i "≤j") ⇒ i <j
(I '>j') or (i '= j' and i ">j") ⇒ i> j

  In the example shown in FIG. 3, i ′ <j ′ is satisfied for each router constituting the router group 301, and i ′ = j ′ and i ″ ≦ j ″ for each router constituting the router group 302. Is satisfied. Therefore, the router R [l, j] added to the hierarchy l is added on the right side of the router groups 301 and 302 on the network topology. Further, i ′ = j ′ and i ″> j ″ are satisfied for each router constituting the router group 303, and i ′> j ′ is satisfied for each router constituting the router group 304. Therefore, the router R [l, j] added to the hierarchy l is added on the left side of the router groups 303 and 304 on the network topology. From the above, the router R [l, j] is added at the position shown in FIG.

  FIG. 1 shows the configuration of the router location calculation apparatus according to the present embodiment. This router location calculation device calculates the logical location on the network topology of a new router to be added to a communication system having a plurality of routers having a hierarchical structure, using the downstream router placement method described above. In FIG. 1, a storage unit 1 stores various types of information and data. The information and data stored by the storage unit 1 include various parameters, initial NW topology information, additional router connection information, logical NW topology information, and the like.

  Various parameters include VLAN ID / IP address / prefix information (including the ID value of each router) set in a communication system that has already been established. The initial NW topology information indicates the logical positional relationship (network topology) of a plurality of routers constituting the current communication system. The additional router connection information indicates the logical positional relationship between the upstream router provided on the distribution server or backup server side connected to the newly added additional router and the new additional router. It is assumed that the upstream router of the additional router has been set in advance by the network administrator in accordance with the above-described another upstream router selection method. Logical NW topology information includes the logical location relationship (network topology) and VLAN ID / IP address of multiple routers that make up the communication system with a new additional router placed at the location calculated by the method described above. / Indicates a prefix.

  The storage unit 1 is connected to the calculation unit 2. The computing unit 2 performs computation using the initial NW topology information, additional router connection information, VLAN ID / IP address / prefix information, etc. as inputs, and outputs logical NW topology information. The calculation unit 2 includes a router logical position calculation unit 21, a router ID calculation unit 22, a VLAN ID / IP address / prefix generation unit 23, and an NW topology output display unit 24.

  The router logical position calculation unit 21 uses the initial NW topology information and the additional router connection information to calculate the logical position of the additional router according to the “downstream router arrangement method” described above. The router ID calculation unit 22 calculates and assigns a 32-bit ID value (Router ID) within the range of VLAN ID / IP address / prefix information to each router whose logical position is calculated. The VLAN ID / IP address / prefix generation unit 23 generates the VLAN ID / IP address / prefix of the VLAN to be set to the interface of each router based on the instructions of the network administrator according to the above-mentioned “different upstream router selection method”. And grant. These results are stored in the storage unit 1 as output data (logical NW topology information). The output data can be visualized through the NW topology output display unit 24 based on the logical position or physical position of the router.

Next, an implementation example of the above-mentioned “different upstream router selection method” and “downstream router arrangement method” will be described.
(Realization example of another upstream router selection method)
1. Two 802.1q tagged VLANs (VLAN A1 and VLAN A2) are set for the physical link connecting the downstream router X and the upstream router A. A network address NA1 is assigned to VLAN A1, and a network address NA2 is assigned to VLAN A2.
2. Two 802.1q tagged VLANs (VLAN B1 and VLAN B2) are set for the physical link connecting downstream router X and upstream router B. A network address NB1 is assigned to VLAN B1, and a network address NB2 is assigned to VLAN B2.

3. Start two OSPF processes (instances) (process 1 and process 2). Each process independently performs unicast route calculation for each target subnet group.
4). Classify the target subnets into two groups and make them belong to OSPF process 1 and OSPF process 2, respectively. NA1 and NB1 belong to OSPF process 1, and NA2 and NB2 belong to OSPF process 2. Thus, NA1 and NB1 are assigned for transferring the main multicast packet, and NA2 and NB2 are assigned for transferring the backup multicast packet.

5. The subnet to which the main multicast source belongs belongs to OSPF process 1. The subnet to which the backup multicast source belongs belongs to the OSPF process 2.
6). For NA1 and NB1 and NA2 and NB2, the magnitude relationship of the addresses is reversed (eg NA1> NB1, NA2 <NB2). As a result, the upstream routers to be joined (participation in multicast packet reception) are different between the main and backup (eg, upstream is the upstream router A (NA1) and backup is the upstream router B (NB2)).

(Realization example of downstream router placement method)
1. A data structure as shown in FIG. 4 is created. The characteristics are shown below.
(A) Each router (indicated by a circle in FIG. 4) is assigned (at least) an ID value that is unique in the hierarchy to identify each other. Hereinafter, the ID value of router R [l, i] in layer 1 is described as ID [l, i]. Let ID [l, i] <ID [l, j] for any i, j (i <j) (the router on the left has a small ID value). The ID value holding method may be explicitly assigned to each router as a router ID or the like, or may be centrally managed by the network operator as an implicit value.

(B) Each router has pointer information (* prev and * next in FIG. 4) indicating routers before and after (left and right) in the same hierarchy.
(C) Each router has pointer information (or ID value) indicating the first upstream router (one) for one upstream layer (* ru in FIG. 4).
(D) Each router has pointer information (or ID value) pointing to the second upstream router (one) for one upstream layer (* bu in FIG. 4).

(E) Each router has pointer information indicating the first downstream router (downstream router of the main multicast distribution route) (0 to plural) for one downstream layer (** rdh, ** rdt in FIG. 4). , * Rd0 to * rdm).
(F) Each router has pointer information indicating the second downstream router (downstream router of the backup multicast distribution route) (0 to plural) for one downstream layer (** bdh, ** bdt in FIG. 4). , * Bd0 to * bdm).
(G) Holds management information (eg, a pointer indicating the first router in each layer) for managing router groups in each layer (from * L_0 in FIG. 4).

  2. The state in which two routers R [0,0] and R [0,1] accommodating the multicast source (main backup) are installed in the hierarchy 0 (the uppermost stream) is the initial state.

  3. FIG. 5 to FIG. 7 show examples in which a router is added to the level 1 with respect to FIG. The router addition algorithm will be described with reference to FIGS. 5 to 7, the downstream router and the first upstream router are connected by a solid line, and the downstream router and the second upstream router are connected by a one-dot chain line. That is, the solid line is the main multicast distribution route, and the one-dot chain line is the backup multicast distribution route. The following processes (A) to (H) are mainly performed by the router logical position calculation unit 21, and the ID value finally given to the additional router is calculated by the router ID calculation unit 22. In order to perform the following processing, the router logical position calculation unit 21 has a function of selecting the first upstream router and the second upstream router of the additional router, a function of selecting the second downstream router of the second upstream router, A function of selecting a first upstream router, a function of selecting an upstream router adjacent to the second upstream router (on the left side in this implementation), a function of comparing ID values of the selected routers, and the like are provided.

(A) The additional router R [l, j] has a pointer to the second upstream router R [l-1, j ′] and the first upstream router R [l-1, j ″] and an ID value (ID [l −1, j ′] and ID [l−1, j ″] (or obtainable).
(B) The second downstream router R [l, b (k)] of the second upstream router R [l−1, j ′] (k = 0 to m is a natural number, and b (k) is bdk = bd0 to The second downstream router (ie R [l, b (k)]) is checked for the presence or absence of the function b (k) from the left in the hierarchy l. Go to (C). If there is no R [l, b (k)] for R [l-1, j '], the process proceeds to (F).

(C) The second downstream router R [l, b (k)] of the second upstream router R [l-1, j ′] is selected in ascending order of ID values (in ascending order from ie k = 0) and sequentially ( D) is executed. When (D) is executed up to router R [l, b (m)] with the maximum ID value (k = m) (ie ID [l-1, {b (m)} "] ≤ID [l-1 , J ″]) proceeds to (H).
(D) ID value of the first upstream router R [l-1, {b (k)} "] of R [l, b (k)] (ID [l-1, {b (k)}"]) And is compared with ID [l-1, j ″]. If ID [l−1, {b (k)} ″] ≦ ID [l−1, j ″] as a result of the comparison, (C If ID [l-1, {b (k)} "]> ID [l-1, j"], the process proceeds to (E).

(E) R [l, j] is inserted between R [l, b (k) -1] and R [l, b (k)] (determination of the logical position of the additional router). ID [l, j] that satisfies ID [l, b (k) -1] <ID [l, j] <ID [l, b (k)] is assigned, and the process ends (eg ID [l, j] = (ID [l, b (k) -1] + ID [l, b (k)]) / 2).
(F) Select upstream router R [l-1, j'-1] on the left side of second upstream router R [l-1, j '], and proceed to (G). If there is no R [l-1, j'-1] (j '= 0: R [l-1, j'] is the first router in layer l-1, R [l, j] is in layer l Insert as the first router (j = 0) (determine logical position of additional router). In this case, an appropriate ID [l, 0] is assigned and the process ends (eg ID [l, 0] = 0, ID [l, 1] = ID [l, 2] / 2).

(G) Select R [l, b (m)] (the router with the largest ID value) from the second downstream routers of R [l-1, j'-1], and proceed to (H). When there is no second downstream router, j ′ = j′−1 is set and the process returns to (F).
(H) R [l, j] is inserted between R [l, b (m)] and R [l, b (m) +1] (determination of the logical position of the additional router). ID [l, j] satisfying ID [l, b (m)] <ID [l, j] <ID [l, b (m) +1] is assigned, and the process ends (eg ID [l, j] = (ID [l, b (m)] + ID [l, b (m) +1]) / 2).

  For example, in FIG. 5, the arrangement ends in (B) → (C) → (D) → (C) → (D) → (E). First, in (B), the presence / absence of the second downstream router of the second upstream router R [l-1, 0] is checked. Since the second downstream router exists, the process proceeds to (C). In (C), the second downstream router R [l, 0] of R [l-1,0] is selected, and the process proceeds to (D). In (D), the ID value (ID [l-1,1]) of the first upstream router R [l-1,1] of R [l, 0] and the first upstream router R of R [l, j] The ID values of [l-1, 2] (ID [l-1, 2]) are compared. Since ID [l-1,1] <ID [l-1,2], the process returns to (C).

  In (C), the second downstream router R [l, 1] next to R [l-1,0] is selected, and the process proceeds to (D). In (D), the ID value (ID [l-1,3]) of the first upstream router R [l-1,3] of R [l, 1] and the first upstream router R [ The ID values of l-1, 2] (ID [l-1, 2]) are compared. Since ID [l-1,3]> ID [l-1,2], the process proceeds to (E). In (E), it is decided to insert R [l, j] between R [l, 0] and R [l, 1], and the ID value is calculated.

  In FIG. 6, the arrangement ends in (B) → (C) → (D) → (C) → (D) → (C) → (H). First, in (B), the presence / absence of the second downstream router of the second upstream router R [l-1, 1] is checked. Since the second downstream router exists, the process proceeds to (C). In (C), the second downstream router R [l, 3] of R [l-1,1] is selected, and the process proceeds to (D). In (D), the ID value (ID [l-1,2]) of the first upstream router R [l-1,2] of R [l, 3] and the first upstream router R of R [l, j] The ID values of [l-1, 4] (ID [l-1, 4]) are compared. Since ID [l-1,2] <ID [l-1,4], the process returns to (C).

  In (C), the second downstream router R [l, 4] next to R [l-1,1] is selected, and the process proceeds to (D). In (D), the ID value (ID [l-1,4]) of the first upstream router R [l-1,4] of R [l, 4] and the first upstream router R [R [l, j] of R [l, 4] The ID values (ID [l-1, 4]) of l-1, 4] are compared. Since the ID values are the same, the process returns to (C). In (C), since (D) has been executed up to the second downstream router R [l, 4] having the maximum ID value of R [l−1,1], the process proceeds to (H). In (H), it is determined to insert R [l, j] between R [l, 4] and R [l, 5], and the ID value is calculated.

  In FIG. 7, the arrangement ends in (B) → (F) → (G) → (H). First, in (B), the presence / absence of the second downstream router of the second upstream router R [l-1,3] is checked. Since there is no second downstream router, the process proceeds to (F). In (F), the router R [l-1,2] on the left side of R [l-1,3] is selected, and the process proceeds to (G). In (G), the router R [l, 6] having the maximum ID value is selected from the second downstream routers of R [l-1,2], and the process proceeds to (H). In (H), it is determined to insert R [l, j] between R [l, 6] and R [l, 7], and the ID value is calculated.

  An example in which the additional steps described above are repeated is shown in FIG. It can be seen that two types of disjoint multicast distribution trees can be set at any router in the hierarchy of l ≧ 1. Regardless of which router is set as the starting point, the main and backup multicast distribution routes can be set so that the distribution routes of both do not overlap.

  FIG. 9 shows the procedure of the downstream router arrangement method described above. Hereinafter, the processing shown in FIG. 9 will be described in correspondence with the above processing. First, 0 is set in the argument k, the second upstream router of the additional router R [l, j] is set in the argument R [l-1, j '] of the second upstream router, and the argument of the first upstream router is set. The first upstream router of the additional router R [l, j] is set in R [l−1, j ″] (step S901).

  Subsequently, it is determined whether or not a second downstream router exists in R [l−1, j ′] (step S902). This step corresponds to (B). When the second downstream router exists, the process proceeds to step S903, and when it does not exist, the process proceeds to step S908. In step S903, the second downstream router of R [l−1, j ′] is set in the argument R [l, b (k)] of the second downstream router, and the argument R [l− of the first upstream router is set. 1, {b (k)} ″] is set to the first upstream router of R [l, b (k)]. This step corresponds to (C).

  Subsequently, it is determined whether or not ID [l−1, {b (k)} ″] ≦ ID [l−1, j ″] (step S904). This step corresponds to (D). If ID [l-1, {b (k)} "] ≤ID [l-1, j"], the process proceeds to step S906, where ID [l-1, {b (k)} "]> If it is ID [l-1, j ″], the process proceeds to step S905. In step S905, ID [l, j] satisfying ID [l, b (k) -1] <ID [l, j] <ID [l, b (k)] is assigned, and the process ends. This step corresponds to (E).

  In step S906, it is determined whether or not the ID value has been compared up to the router having the largest ID value among the second downstream routers of R [l−1, j ′]. This step corresponds to (C). If the ID value is not compared up to the router with the largest ID value, the value of k is incremented by 1 to perform the next second downstream router process (step S907), and the process returns to step S903. If the ID value is compared up to the router with the largest ID value, the argument R [l, b (m)] of the second downstream router with the largest ID value is set (step S910). ID [l, j] satisfying l, b (m)] <ID [l, j] <ID [l, b (m) +1] is assigned, and the process ends (step S911). Steps S910 to S911 correspond to (H).

  On the other hand, in step S908, it is determined whether or not a left router exists in the second upstream router R [l-1, j ']. This step corresponds to (F). If there is no router on the left side, that is, R [l-1, j '] is the first router in layer l-1, R [l, j] is the first router in layer l and an appropriate ID [l, 0] is given and the process is terminated (step S912). This step corresponds to (F). If there is a router on the left side, it is determined whether or not a second downstream router exists in R [l−1, j ′] (step S909). If the second downstream router exists, the process proceeds to step S910. If the second downstream router does not exist, the process of step S908 is performed again on the left router.

  As described above, in the other upstream router selection method of this embodiment, two subnets are provided for each physical link connected to the two upstream routers in any router except the router closest to the distribution server or the backup server. In addition, each subnet is appropriately grouped for main multicast and backup multicast, and the network address of each subnet is set to satisfy a predetermined relationship. As a result, when there are two upstream router candidates having the same cost for the two multicast sources, the route toward each multicast source can be branched.

  Further, in the downstream router arrangement method of the present embodiment, initial NW topology information (network topology information) indicating the logical positional relationship between a plurality of routers having a hierarchical structure, and the upstream router and additional router connected to the additional router Based on the additional router connection information indicating the logical positional relationship and the identification value (ID value) assigned to each of the plurality of routers corresponding to the logical position of each router, By calculating the logical position, establishment of a disjoint multicast distribution path is guaranteed in all routers.

  Therefore, according to the present embodiment, a disjoint multicast tree can be dynamically constructed by combining another upstream router selection method and a downstream router arrangement method. Therefore, it is possible to perform multicast distribution by separating into two disjoint distribution paths in units of multicast sources or groups according to the intention of the network administrator.

  Also, disjoint multicast distribution routes can be secured using only the OSPF and PIM-SM functions of existing routers. Instead of statically fixing the multicast distribution route, the distribution route is dynamically determined by OSPF / PIM-SM as in the conventional case, and distribution route switching by OSPF / PIM-SM functions in the event of a network failure.

  As described above, the embodiments of the present invention have been described in detail with reference to the drawings. However, the specific configuration is not limited to these embodiments, and includes design changes and the like without departing from the gist of the present invention. . For example, in another upstream router selection method, NA1 <NB1 and NA2> NB2 may be used to branch the path, and in this case, the positions of the main multicast source and the backup multicast source are reversed on the network topology.

  Further, the router position calculation apparatus according to the above-described embodiment records a program for realizing the operation and function in a computer-readable recording medium, causes the computer to read the program recorded in the recording medium, and executes the program. You may implement | achieve by doing.

  Here, the “computer” includes a homepage providing environment (or display environment) if the WWW system is used. The “computer-readable recording medium” refers to a storage device such as a portable medium such as a flexible disk, a magneto-optical disk, a ROM, and a CD-ROM, and a hard disk built in the computer. Further, the “computer-readable recording medium” refers to a volatile memory (RAM) in a computer system that becomes a server or a client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. In addition, those holding programs for a certain period of time are also included.

  The above-described program may be transmitted from a computer storing the program in a storage device or the like to another computer via a transmission medium or by a transmission wave in the transmission medium. Here, the “transmission medium” for transmitting the program refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line. Further, the above-described program may be for realizing a part of the above-described function. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer, what is called a difference file (difference program) may be sufficient.

It is a block diagram which shows the structure of the router location calculation apparatus by one Embodiment of this invention. It is a reference figure for demonstrating the another upstream router selection method in one Embodiment of this invention. It is a reference figure for demonstrating the downstream router selection method in one Embodiment of this invention. It is a reference figure for demonstrating the data structure in one Embodiment of this invention. It is a reference figure for demonstrating the example of the addition of the router in one Embodiment of this invention. It is a reference figure for demonstrating the example of the addition of the router in one Embodiment of this invention. It is a reference figure for demonstrating the example of the addition of the router in one Embodiment of this invention. It is a reference diagram showing a network configuration example in an embodiment of the present invention. It is a flowchart which shows the procedure of the downstream router selection method in one Embodiment of this invention. It is a block diagram which shows the structure of the conventional communication network. It is explanatory drawing for demonstrating the problem of a prior art.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Memory | storage part, 2 ... Operation part, 21 ... Router logical position calculation part, 22 ... Router ID calculation part, 23 ... VLAN ID / IP address / prefix generation part, 24 ... NW topology output display unit 101... Broadcast distribution server 102 102 109 L2 switch 103 backup server 104a 104b 105a 105b 106a 106b 107 router 108 ... Communication control device, 110 ... Client terminal

Claims (7)

A communication system comprising a plurality of routers having a hierarchical structure for transferring multicast packets transmitted by a distribution server,
In any router except the router closest to the distribution server, the first and second subnets are set for the physical link connected to the first upstream router provided on the distribution server side, and The third and fourth subnets are set for the physical link connected to the second upstream router provided on the distribution server side,
The first and third subnets are assigned for transferring a first multicast packet, the second and fourth subnets are assigned for transferring a second multicast packet, and the first, The communication system, wherein network addresses of the second, third, and fourth subnets satisfy a predetermined relationship.   Assuming that the network address of the first subnet is NA1, the network address of the second subnet is NA2, the third network address is NB1, and the fourth network address is NB2, the predetermined relationship is NA1> NB1 The communication system according to claim 1, wherein NA2 <NB2 or NA1 <NB1 and NA2> NB2. The logical addition position when any router except the router closest to the distribution server is added is:
Network topology information indicating a logical positional relationship between the plurality of routers having a hierarchical structure;
Additional router connection information indicating a logical positional relationship between the upstream router provided on the distribution server side and the additional router connected to the newly added additional router,
An identification value assigned to each of the plurality of routers corresponding to the logical position of each router;
The communication system according to claim 1, wherein the communication system is calculated based on the above. A communication system comprising a plurality of routers having a hierarchical structure for transferring multicast packets transmitted by a first distribution server and a second distribution server,
The multicast packet transmitted by the first distribution server is forwarded from the first upstream router provided on the first distribution server side by the hash processing of each router constituting the plurality of routers. An identification value of the first distribution server is determined;
A multicast packet transmitted by the second distribution server is forwarded from a second upstream router provided on the second distribution server side by hash processing of each router constituting the plurality of routers. An identification value of the second distribution server is determined;
The logical addition position when any router except the router closest to the first distribution server and the second distribution server is added is
Network topology information indicating a logical positional relationship between the plurality of routers having a hierarchical structure;
Additional router connection information indicating a logical positional relationship between the upstream router and the additional router provided on the side of the first distribution server and the second distribution server, which is connected to the newly added additional router;
An identification value assigned to each of the plurality of routers corresponding to the logical position of each router;
A communication system characterized by being calculated based on the above. Network topology information indicating a logical positional relationship between a plurality of routers having a hierarchical structure that forwards multicast packets transmitted by a distribution server;
Additional router connection information indicating a logical positional relationship between the upstream router provided on the distribution server side and the additional router connected to the newly added additional router,
An identification value assigned to each of the plurality of routers corresponding to the logical position of each router;
A router position calculation device that calculates a logical position of the additional router based on Network topology information indicating a logical positional relationship between a plurality of routers having a hierarchical structure that forwards multicast packets transmitted by a distribution server;
Additional router connection information indicating a logical positional relationship between the upstream router provided on the distribution server side and the additional router connected to the newly added additional router,
An identification value assigned to each of the plurality of routers corresponding to the logical position of each router;
A router position calculation program for causing a computer to execute a process of calculating a logical position of the additional router based on the above. A computer-readable recording medium on which the router position calculation program according to claim 6 is recorded.

Images