JP2009017230A - Communication system, redundant node, and program for redundant node - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To make an available communication band fair between terminals connected to each node. <P>SOLUTION: When an upstream adjacent node is judged to be a redundant node and it is judged that a control object traffic can be received from an upstream redundant node capable of receiving the control object traffic from a terminal, a fair rate notification means 730 provided in a redundant node 700 notifies the upstream adjacent node of a fair rate notified by a downstream adjacent node. A communication band determination means 740 determines an upper limit value of the communication band of the node so that the sum total of upper limit values of the communication band of a prescribed control object traffic, which are set to respective redundant nodes capable of receiving control object traffics from terminals and transmitting the control object traffics into a communication system, may be equal to the fair rate. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a communication system, a redundant node, a redundant node program, and a communication band upper limit value determining method applied to the redundant node.

  In the Internet, unlike a time division multiplexing method in which a predetermined communication band is assigned to traffic in advance, a multiplexing method called a packet multiplexing method is used.

  In the packet multiplexing method, the communication band of the communication system is shared by all traffic. A network node (hereinafter simply referred to as a node) can transmit traffic when there is a communication band that is not used by other traffic. That is, if the rate of other traffic is low, there is a sufficient amount of data that can be transmitted and received between nodes, and traffic can be transmitted.

  A traffic is a set of frames. The communication band used by the traffic is a value obtained by dividing the sum of the data amount (frame size) of the frames included in the traffic by the time required for transmission, and is synonymous with the rate.

  In the time division multiplexing method, even if a communication band assigned to a certain traffic is unused, other traffic is prohibited from using the communication band. On the other hand, in the packet multiplexing method, it is possible to multiplex other traffic to an empty band. Therefore, the bandwidth utilization efficiency of the packet multiplexing scheme is higher than that of the time division multiplexing scheme. This effect is called the statistical multiplexing effect of the packet multiplexing method.

  However, in the packet multiplexing method, because of the above-described characteristics, when a large number of traffic is intensively transmitted to a specific link of the communication system within the same time, a situation (congestion) exceeding the communication band of the link occurs. , There may be a delay in forwarding traffic or a part of the traffic may be discarded. In this case, the packet multiplexing method has a disadvantage that communication becomes unstable or communication is interrupted.

  In order to solve this problem, the packet multiplexing method guarantees a predetermined communication band for specific traffic as in the time division multiplexing method, and adds a parameter called priority to each traffic prioritization. Various communication band control methods have been devised, such as preferentially transmitting high traffic.

  As one of such bandwidth control methods, there is a bandwidth control method in which all traffic passing through a certain link in the communication system shares the communication bandwidth of the link fairly. “Traffic sharing the communication bandwidth of a link fairly” means that each traffic is communicated so that the communication bandwidth that can be transmitted by the link does not exceed the value that is equally divided by the number of traffic that uses the link. It is to control the bandwidth. For example, assume that the communication bandwidth of the link is 10 Gbps, and there are four types of traffic A, B, C, and D as traffic whose communication bandwidth is controlled. In this case, control is performed so that the communication bandwidth of each traffic does not exceed 10 Gbps / 4 = 2.5 Gbps. Controlling the communication bandwidth of the traffic in this way is called “traffic sharing of the link communication bandwidth fairly”.

  As an example of such control, there is a band control called fairness used in RPR (Resilient Packet Ring) standardized in 2004 in the Institute of Electrical and Electronics Engineers (IEEE). Fairness is described in Non-Patent Document 1.

  Hereinafter, traffic whose communication bandwidth is controlled is referred to as fairness control target traffic.

  A communication network to which RPR is applied (hereinafter referred to as an RPR network) includes a ring (ringlet) for transferring a frame in a clockwise direction and a ring (ringlet) for transferring a frame in a counterclockwise direction. It is a heavy ring network. A ringlet is a communication path of an RPR network.

  FIG. 7 is an explanatory diagram illustrating an example of an RPR network. FIG. 7 illustrates a case where eight nodes (hereinafter referred to as RPR nodes) that operate in conformity with IEEE 802.17 are included. In the RPR network 10 illustrated in FIG. 7, the RPR nodes 100 to 170 are connected to different terminals 200 to 270, respectively.

  The RPR node and its subordinate terminals (terminals not belonging to the ring network) transmit and receive Ethernet frames. Ethernet is a registered trademark. When the RPR node receives an Ethernet frame from a subordinate terminal, the RPR node generates a frame (denoted as an RPR frame) in which the Ethernet frame is stored in a payload, and transfers the RPR frame to another RPR node. Specifically, the RPR frame is transferred to the RPR node whose terminal is the terminal serving as the destination of the Ethernet frame. Further, when the RPR node receives the RPR frame destined for the own node, the RPR node extracts the Ethernet frame from the RPR frame and transmits the Ethernet frame to the terminal under control.

  Each of the nodes 100 to 170 has ports P1 to P3, respectively. Ports P1 and P2 are ports for transmitting and receiving RPR frames to and from adjacent RPR nodes. The port P3 is a port for transmitting / receiving an Ethernet frame with a subordinate terminal.

  In the traffic of the RPR network, bandwidth guaranteed traffic in which the communication bandwidth of traffic is guaranteed and best effort traffic in which the bandwidth is not guaranteed are defined. The fact that the traffic bandwidth is guaranteed means that the traffic bandwidth is secured.

  Bandwidth-guaranteed traffic is traffic that is assigned a predetermined communication band exclusively in advance and is preferentially transmitted over best-effort traffic. If the guaranteed bandwidth traffic has a rate below the guaranteed bandwidth, communication is always guaranteed. However, if multiple bandwidth-guaranteed traffic exists in the communication system and they share the same link, the sum of the communication bandwidths assigned to these traffic is the communication bandwidth of the shared link. It is necessary to set the guaranteed bandwidth of each bandwidth-guaranteed traffic so that it is less than or equal to the minimum value.

  On the other hand, the best effort type traffic is normal traffic in the packet multiplexing method, and can be transmitted only when an unused communication band exists as described above.

  In the RPR network, best-effort traffic is a target of fairness bandwidth control processing (fairness control target traffic).

  Hereinafter, as a technique related to the present invention, a general operation of a node when congestion occurs in a link constituting an RPR network will be described. A node existing immediately upstream of the congestion occurrence point is referred to as a head node. When congestion occurs in the link constituting the RPR network, the head node controls the fairness control that passes through the node (the node upstream of the head node) with respect to the node arranged upstream of the own node (that is, the head node itself). A fair rate indicating the upper limit value of the communication band permitted as the communication band of the target traffic is advertised. Specifically, the head node transmits a control frame (fairness frame) storing the fair rate to an adjacent node upstream of the own node (head node itself). The upstream adjacent node that receives the fairness frame from the head node so that the communication band used by the traffic passing through the congested portion of the fairness controlled traffic transmitted from the node itself into the RPR network is equal to or less than the fair rate. Perform bandwidth control processing.

  Each node arranged upstream of the head node operates as follows. If the communication bandwidth of the traffic subject to fairness control received from the upstream adjacent node is equal to or higher than the fair rate, each node arranged upstream of the head node will receive a fairness frame storing the fair rate advertised by the head node. Send to upstream neighbor. On the other hand, if the communication bandwidth of the fairness controlled traffic received from the upstream adjacent node is less than the fair rate, the upstream adjacent node is notified with a fairness frame that it is not necessary to suppress the communication bandwidth of the fairness controlled traffic. To do. Specifically, a fairness frame with the fair rate as a full rate is transmitted to the upstream adjacent node. Each node that receives a fairness frame in which the fair rate is not full rate is set so that the communication bandwidth used by the traffic passing through the congestion point is equal to or less than the fair rate among the fairness controlled traffic transmitted from the node itself into the RPR network. Perform bandwidth control processing.

  A node that receives a fairness frame in which the fair rate is not full rate and transmits a fairness frame in which the fair rate is full rate is called a tail node. The communication path from the head node to the tail node is called a congestion domain.

  A node that receives a fairness frame whose fair rate is full rate from a downstream adjacent node transmits a similar fairness frame upstream. Accordingly, the nodes upstream from the tail node do not execute the bandwidth control process on the fairness controlled traffic transmitted by the nodes.

  FIG. 8 is a flowchart of processing in which a general RPR node determines whether or not its own node is a tail node. When a general RPR receives a fairness frame from an adjacent RPR node, it determines whether or not the communication band of the fairness controlled traffic received from the upstream adjacent node is less than the fair rate (step S1). Then, if the communication band of the fairness control target traffic received from the upstream adjacent node is less than the fair rate (Yes in step S1), it is determined that the own node is a tail node (step S2). On the other hand, if it is equal to or higher than the fair rate (No in step S1), it is determined that the own node is not a tail node (step S3).

  With the above-described control, even in the case of congestion in the RPR network, only a specific node can occupy the communication bandwidth of the link at the congested part, and a plurality of nodes can be shared fairly. .

  Next, a high-speed protection function that realizes high reliability, which is another feature of RPR, will be described. In the RPR network, when a link between RPR nodes is disconnected, the RPR nodes on both sides of the link detect the link disconnection. The RPR node that detects the disconnection of the link immediately notifies all other RPR nodes that the link has failed. The other RPR node that has received the notification of the occurrence of the failure makes a transition to an operation state in which traffic is transmitted so as to bypass the link disconnection point. As a result, the RPR network can continue communication.

  RPR is designed to be used in backbone communication systems where large volumes of traffic flow, such as urban networks, and communicates in a short time within 50 ms, equivalent to SDH (Synchronous Digital Hierarchy) or SONET (Synchronous Optical Network). Designed to recover. Therefore, it is possible to construct a highly reliable communication system.

  However, IEEE 802.17 does not describe a method for continuing communication when an RPR node that is a traffic destination fails. Therefore, various RPR node redundancy techniques have been proposed that enable communication to be continued even when one RPR node fails by duplicating the RPR node.

  FIG. 9 is an explanatory diagram showing an example of an RPR network including redundant RPR nodes. In the RPR network 20 illustrated in FIG. 9, the RPR nodes 100 and 110 are virtualized as one RPR node. A common terminal 200 is connected to the RPR nodes 100 and 110. Note that separate terminals 220 to 270 are connected to the other RPR nodes 120 to 170, respectively. As illustrated in FIG. 9, by connecting a common terminal 200 to the RPR nodes 100 and 110, even if one of the RPR nodes 100 and 110 fails, the terminal 200 transmits traffic to the other RPR node. Thus, communication can be continued. As in this example, a pair of RPR nodes 100 and 110 made redundant by being connected to the same terminal are called a pair node.

  In the RPR node redundancy technology, a method for preventing a frame from being received multiple times, a method for preventing a broadcast storm, a method for determining an output port to which a subordinate terminal transmits a frame, a method for restoring communication in the event of a failure, etc. are defined. Is done.

  An example of a node redundancy technique is described in Patent Document 1, for example.

  By combining the above-mentioned RPR node redundancy technology with the high-speed protection technology described in IEEE 802.17, it is possible to further improve the reliability of the RPR network.

JP 2006-245780 A “IEEE Standards 802.17 Part 17: Resilient Packet Ring Access Method and Physical Layer Specification (Part 17: Resident Packet Ring (RPR) access method and physical layer specifications). overview) ", IEEE (Institute of Electrical and Electronics Engineers, Inc), 2004, p. 27-54

  As a technique related to the present invention, an RPR node having the configuration illustrated in FIG. 10 can be considered. The RPR node illustrated in FIG. 10 will be described. The nodes illustrated in FIG. 10 include input ports 400-1 to 400-3, frame analysis units 410-1 to 410-2, an RPR switch processing unit 420, a TDB (Topology Database) storage unit 430, and a frame multiplexing unit. 440-1-2, output ports 450-1-3, fairness control units 460-1-2, and traffic rate measurement units 470-1-2.

  Input ports 400-1 to 400-3 of the RPR node are ports on the receiving side in ports P1 to P3 (see FIG. 7). The input ports 400-1 and 400-2 are ports that receive RPR frames transmitted from adjacent RPR nodes. The input port 400-1 is a port that receives an RPR frame transmitted from one adjacent RPR node. The input port 400-2 is a port that receives an RPR frame transmitted from the other adjacent RPR node. The input port 400-3 is a port that receives an Ethernet frame transmitted from a subordinate terminal.

  Output ports 450-1 to 450-3 are ports on the transmission side in ports P1 to P3 (see FIG. 7). The output ports 450-1 and 450-2 are ports that transmit RPR frames to adjacent RPR nodes. The output port 450-1 is a port that transmits an RPR frame to one adjacent RPR node. The output port 450-2 is a port that transmits an RPR frame to the other adjacent RPR node. The output port 450-3 is a port that transmits an Ethernet frame to a subordinate terminal.

  If the RPR frame received from the adjacent RPR node via the input port 400-1 is a fairness frame, the frame analysis unit 410-1 sends the fairness frame to the fairness control unit 460-2. If the received RPR frame is not a fairness frame, the RPR frame is sent to the RPR switch processing unit 420. Similarly, if the RPR frame received from the other adjacent RPR node via input port 400-2 is a fairness frame, frame analysis unit 410-2 sends the fairness frame to fairness control unit 460-1. . If the received RPR frame is not a fairness frame, the RPR frame is sent to the RPR switch processing unit 420.

  The RPR switch processing unit 420 executes various processes related to RPR defined in “IEEE Standards 802.17”. For example, a process of transferring an RPR frame received from an adjacent RPR node, a process of generating an RPR frame from an Ethernet frame received from a subordinate terminal and transmitting it within the RPR network, an Ethernet frame stored in an RPR frame addressed to the own node Is extracted and transmitted to the subordinate terminals, the topology information of the RPR network is managed by Topology Discovery Protocol (TDP), the RPR network is managed by OAM (Operations, Administration, Maintenance), and the like.

  The RPR switch processing unit 420 holds the fairness control target traffic among the traffic scheduled to be transmitted from the output port 450-1. Similarly, the traffic subject to fairness control is held among the traffic scheduled to be transmitted from the output port 450-2. When the amount of data held as the traffic scheduled to be transmitted from the output port 450-1 becomes larger than a predetermined amount, the fairness control unit 460-1 is notified of this. Further, when the amount of data held as traffic scheduled to be transmitted from the output port 450-2 becomes larger than a predetermined amount, the fairness control unit 460-2 is notified of this.

  The TDB storage unit 430 is a storage device that stores a TDB (topology database). The TDB is a database in which topology information of the RPR network to which the own node belongs is registered. As the topology information, the number of RPR nodes included in the RPR network, the position of each RPR node, the failure information of each link, and the like are registered. The TDB is referred to by the fairness control units 460-1 and 460-2 and the RPR switch processing unit 420.

  The frame multiplexing unit 440-1 multiplexes the data frame sent from the traffic rate measuring unit 470-1 and the fairness frame sent from the fairness control unit 460-2, and the multiplexed frame is output from the output port 450-1. Send. Similarly, the frame multiplexing unit 440-2 multiplexes the data frame sent from the traffic rate measuring unit 470-2 and the fairness frame sent from the fairness control unit 460-1, and the multiplexed frame is output to the output port 450. -2.

  The fairness control unit 460-1 and the fairness control unit 460-2 execute various processes related to fairness defined in “IEEE Standards 802.17”. For example, detection of congestion at the output port 450-1 or output port 450-2 of the own node, calculation of fair rate, generation and transmission of a fairness frame, analysis of a fairness frame received from an adjacent node, RPR switch processing unit of the own node The bandwidth control information is notified to 420.

  Each of the fairness control target traffics 460-1 and 460-2 determines that congestion has occurred when the communication band used by the fairness control target traffic exceeds a predetermined communication band. Further, it is also determined that congestion has occurred when a notification that the data amount of the fairness control target traffic held by the RPR switch processing unit 430 has exceeded a predetermined amount has been received.

  The traffic rate measurement unit 470-1 executes various traffic rate measurement processes defined in IEEE 802.17, and then displays the measurement results and the results of computations performed on the measurement results as fairness of the own node. Notify controller 460-1. Similarly, the traffic rate measurement unit 470-2 also executes various traffic rate measurement processes and notifies the fairness control unit 460-2 of the own node of the measurement results.

  Examples of traffic rates measured by the traffic rate measuring units 470-1 and 470-2 include traffic rates transmitted from the own node and subjected to fairness control, traffic transmitted from the upstream node, and passing through the own node. Late etc.

  Each traffic rate measuring unit 470-1-2 also performs calculations such as smoothing processing and normalization processing using a low-pass filter on the measured traffic rate.

  The RPR node having the configuration illustrated in FIG. 10 can be used as each node illustrated in FIG. 7, and a RPR node made redundant by applying the RPR node redundancy technique (for example, illustrated in FIG. 9). Nodes 100 and 110) can also be used.

  When the RPR node is made redundant, there is a problem that unfairness occurs in the use of the communication band for the fairness control target traffic between the terminals under the RPR node. Specifically, the communication bandwidth of the fairness controlled traffic that can be transmitted from the terminals under the redundant RPR node into the RPR network, and the terminals under the non-redundant RPR node can be transmitted into the RPR network. There is a problem that a difference occurs in the communication band of the traffic subject to fairness control.

  For example, in the example illustrated in FIG. 9, the communication band that can be used by the terminal 200 for transmitting the traffic subject to fairness control is a communication band assigned to two nodes, the RPR node 100 and the RPR node 100. Accordingly, the communication band that can be used by the terminal 200 for transmitting the traffic subject to fairness control is twice the communication band that can be used by the terminals 220 to 270 under the non-redundant RPR nodes 120 to 170.

  Further, when one of the RPR nodes 100 and 110 becomes a tail node, the other RPR node does not execute processing for suppressing the communication band of the fairness control target traffic. Then, the difference between the communication band that can be used by the terminal 200 for transmitting the fairness control target traffic and the communication band that can be used by the other terminals 220 to 270 increases.

  Therefore, the present invention provides a communication system, a redundant node, a redundant node program, and a communication band applied to the redundant node that can make the available communication band fair among terminals connected to each node. An object is to provide a method for determining an upper limit value.

  The communication system of the present invention is a communication system including a node that notifies the other node of a fair rate that is an upper limit value of a communication band permitted to another node, and is connected to a common terminal. Multiple redundant nodes, each redundant node is placed adjacent to each other, and when each redundant node transmits traffic to the congestion location side, when the traffic transmission direction is the downstream side In addition, the redundant node determining means for determining whether or not the upstream adjacent node is a redundant node, and control from the upstream redundant node capable of receiving the control target traffic whose communication bandwidth is controlled from the terminal Receivability judging means for judging whether or not the target traffic can be received, and it is determined that the upstream adjacent node is a redundant node. A fair rate notification that notifies the upstream neighboring node of the fair rate notified from the downstream adjacent node when it is determined that the control target traffic can be received from the redundant node on the upstream side that can receive the network. And a sum of upper limits of communication bandwidths of predetermined control target traffic set in a redundant node capable of receiving control target traffic from the terminal and transmitting the control target traffic into the communication system. Communication bandwidth determining means for determining an upper limit value of the communication bandwidth of the predetermined control target traffic transmitted from the own node so as to be equal to the fair rate notified from the adjacent node on the downstream side, and the predetermined control The target traffic is the control target traffic received from the terminal, and the control target traffic that passes through the congestion point. Characterized in that it is a Fick

  The redundant node according to the present invention is connected to a common terminal in a communication system including a node for notifying the other node of a fair rate that is an upper limit value of a communication band permitted to another node. A redundant node used as one of a plurality of nodes, and when transmitting traffic to the congestion location side, when the traffic transmission direction is the downstream side, the adjacent node on the upstream side is the redundant node A redundancy node determination means for determining whether or not the control target traffic can be received from an upstream redundant node capable of receiving the control target traffic whose communication bandwidth is controlled from the terminal. It is determined that the determination means and the upstream adjacent node are redundant nodes, and control is performed from the upstream redundant node that can receive the control target traffic from the terminal. When receiving the traffic to be notified from the downstream adjacent node to the upstream adjacent node, and receiving the control target traffic from the terminal. The total of the upper limit values of the communication bandwidths of the predetermined control target traffic set in the redundant node capable of transmitting the control target traffic into the communication system is notified from the downstream neighbor node. Communication bandwidth determining means for determining an upper limit value of the communication bandwidth of the predetermined control target traffic transmitted from the node so that the predetermined control target traffic is the control target traffic received from the terminal. And it is the control object traffic which passes a congestion location, It is characterized by the above-mentioned.

  The redundant node program according to the present invention is connected to a common terminal in a communication system including a node that notifies the other node of a fair rate that is an upper limit value of a communication band permitted to the other node. When transmitting traffic to the congestion location side to the computer provided in the redundant node used as one of a plurality of nodes, when the traffic transmission direction is the downstream side, the adjacent node on the upstream side Redundant node determination processing for determining whether or not the node is a redundant node, whether or not control target traffic can be received from an upstream redundant node capable of receiving the control target traffic whose communication bandwidth is controlled from the terminal Receivability determination process for determining, it is determined that the upstream adjacent node is a redundant node, and the control target traffic can be received from the terminal. A fair rate notification process for notifying an upstream adjacent node of a fair rate notified from a downstream adjacent node when it is determined that control target traffic can be received from the redundant node on the side, and the terminal The sum of the upper limit values of the communication bandwidths of the predetermined control target traffic set in the redundant node that can receive the control target traffic and transmit the control target traffic into the communication system is A communication band determining process for determining an upper limit value of a communication band of the predetermined control target traffic transmitted from the own node so as to be equal to the fair rate notified from the node, and a communication band of the predetermined control target traffic As the upper limit value of the traffic to be controlled received from the terminal and passing through the congestion point Characterized in that to determine the upper limit value of the communication bandwidth of the control target traffic.

  A method for determining an upper limit value of a communication band according to the present invention is a communication system including a node that notifies the other node of a fair rate that is an upper limit value of a communication band permitted to another node. Communication bandwidth upper limit value applied to each redundant node connected to the network, when transmitting traffic to the congestion location side, when the traffic transmission direction is downstream, Determining whether the node is a redundant node, determining whether the control target traffic can be received from the upstream redundant node capable of receiving the control target traffic whose communication bandwidth is controlled from the terminal; It is determined that the upstream adjacent node is a redundant node, and it is determined that the control target traffic can be received from the upstream redundant node that can receive the control target traffic from the terminal. The downstream is notified from the downstream adjacent node to the upstream adjacent node, receives the control target traffic from the terminal, and transmits the control target traffic into the communication system. The predetermined value transmitted from the own node is set so that the sum of the upper limits of the communication bandwidths of the predetermined control target traffic set in the redundant node that can perform the same as the fair rate notified from the adjacent node on the downstream side. An upper limit value of a communication band of the control target traffic is determined, and the control target traffic that is received from the terminal and that passes through a congestion point is used as the predetermined control target traffic.

  ADVANTAGE OF THE INVENTION According to this invention, the terminal under the node contained in a congestion area can share the communication band of the communication system to which a node belongs fairly.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.
In the following description, when traffic is transmitted to the congestion location side, the transmission direction of the traffic is assumed to be the downstream side (sometimes simply referred to as downstream). The direction on the opposite side is the upstream side (sometimes simply referred to as upstream). Each node receives a fairness frame from a neighboring node on the downstream side.

  FIG. 1 is an explanatory diagram showing an example of a communication system according to the present invention. FIG. 1 illustrates a case where the communication system 15 of the present invention is an RPR network. The communication path in the RPR network is a duplex ringlet. The nodes 10 and 11 of the present invention are redundant nodes (redundant nodes). That is, the nodes 10 and 11 of the present invention are arranged adjacent to each other, and a common terminal 200 is connected as a subordinate terminal. The terminal 200 is connected to both the nodes 10 and 11 of the present invention.

  The number of nodes of the present invention to be made redundant may be plural. In other words, FIG. 1 illustrates a case where two nodes 10 and 11 are made redundant, but three or more nodes may be made redundant. Also in this case, the redundant nodes are arranged adjacent to each other. That is, the redundant nodes are arranged in a row.

  The RPR network 15 is a communication system in which the nodes 10 and 11 of the present invention and the nodes 120 to 170 are connected in a ring shape. The nodes 120 to 170 are general nodes used in a general RPR network. The nodes 10 and 11 and the nodes 120 to 170 according to the present invention are RPR nodes that operate in accordance with “IEEE Standards 802.17”. However, the operations when the RPR nodes 10 and 11 receive the fairness frame are different from those of other general RPR nodes 120 to 170.

  As described above, the common terminal 200 is connected to the nodes 10 and 11 of the present invention. Subordinate terminals 220 to 270 are connected to other general RPR nodes 120 to 170, respectively. Terminals 220 to 270 under the general RPR nodes 120 to 170 are not connected to other RPR nodes.

  Further, a redundancy technique for virtualizing a plurality of nodes as one RPR node is applied to the nodes 10 and 11 of the present invention. For example, an operation is applied in which a plurality of redundant nodes do not transmit Ethernet frames redundantly to the terminal 200. An example of this operation will be described later.

  Further, when transmitting an Ethernet frame to a node, the terminal 200 connected to the plurality of redundant nodes transmits the Ethernet frame to only one of the plurality of nodes. For example, terminal 200 responds to the destination MAC address and source MAC address of the Ethernet frame that terminal 200 itself intends to transmit from among a plurality of redundant nodes (nodes 10 and 11 in the example shown in FIG. 1). One node is selected and an Ethernet frame is transmitted to that node. However, the method by which the terminal 200 selects one node from a plurality of redundant nodes is not limited to the above example.

  Each of the nodes 10, 11, 120 to 170 has ports P1 to P3, respectively. Ports P1 and P2 are ports for transmitting and receiving RPR frames to and from adjacent RPR nodes. The port P3 is a port for transmitting / receiving an Ethernet frame with a subordinate terminal.

  For example, RPR nodes having the configuration shown in FIG. 10 may be used as the RPR nodes 120 to 170 that are not made redundant.

  Each terminal 200 to 270 transmits traffic to the RPR node. Among the traffic transmitted by the terminal to the RPR node, there is fairness controlled traffic. What kind of traffic is to be fairness controlled traffic is variously determined by the network administrator. The RPR switch processing 420 (see FIGS. 2 and 10) included in each RPR node 10, 11, 120 to 170 includes, for example, a destination address, a transmission source address, a VLAN identifier, which are stored in a header of a frame received from a terminal. Based on the priority (priority) or the like, it is determined whether or not the traffic received from the terminal is fairness control target traffic. Further, the RPR switch processing 420 of each node controls the communication band when the fairness control target traffic received from the terminal is transmitted into the RPR network, and when the fairness control target traffic passes through the congestion point. When the fairness control target traffic received from the terminal does not pass through the congestion point, the communication band is not controlled. Therefore, among the fairness control target traffic received from the terminal, the traffic that passes through the congestion point becomes the fairness control target traffic whose communication band is actually controlled.

  Next, the configuration of the node of the present invention will be described. FIG. 2 is a block diagram illustrating a configuration example of a node according to the present invention. Constituent elements similar to those of the RPR node illustrated in FIG. 10 are denoted by the same reference numerals as in FIG. Hereinafter, the configuration of the node will be described by taking the RPR node 10 as an example with reference to FIG. 2, but the configuration of another redundant RPR node 11 (see FIG. 1) is the same as that of the RPR node 10.

  The RPR node 10 of the present invention includes input ports 400-1 to 400-3, frame analysis units 410-1 to 410-2, an RPR switch processing unit 420, a TDB storage unit 530, frame multiplexing units 440-1 to 4-2, Output ports 450-1 to 3, fairness control units 560-1 and 2, traffic rate measurement units 470-1 and 2, and a setting interface 580 are provided. The input ports 400-1 to 400-3, the frame analysis units 410-1 to 42, the RPR switch processing unit 420, the frame multiplexing units 440-1 to 2, the output ports 450-1 to 450, and the traffic rate measurement units 470-1 to These are the same as the constituent elements shown in FIG.

  400-1 to 400-3 of the RPR node 10 are reception-side ports (ports for receiving frames) in the ports P1 to P3 of the RPR node 10 shown in FIG.

  The input ports 400-1 and 400-2 are ports that receive RPR frames transmitted from adjacent RPR nodes. Specifically, the input port 400-1 is a port that receives an RPR frame transmitted from an RPR node adjacent in the clockwise direction (in this example, the RPR node 11). In addition, the input port 400-2 is a port that receives an RPR frame transmitted from an RPR node adjacent in the counterclockwise direction (in this example, the RPR node 170). The input port 400-3 is a port that receives an Ethernet frame transmitted from a subordinate terminal (the terminal 200 in this example).

  The output ports 450-1 to 450-3 of the RPR node 10 are transmission-side ports (ports for transmitting frames) in the ports P1 to P3 of the RPR node 10 shown in FIG.

  The output ports 450-1 and 450-2 are ports that transmit RPR frames to adjacent RPR nodes. Specifically, the output port 450-1 is a port that transmits an RPR frame to an RPR node (the RPR node 11 in this example) adjacent in the clockwise direction. The output port 450-2 is a port that transmits an RPR frame to an RPR node (an RPR node 170 in this example) adjacent in the counterclockwise direction. The output port 450-3 is a port that transmits an Ethernet frame to a subordinate terminal (the terminal 200 in this example).

  The frame analysis units 410-1 and 410-2 each determine whether or not the RPR frame received by the input port is a fairness frame. In this determination, for example, with reference to the header of the received RPR frame, if the information indicating the fairness frame is described in the header, it may be determined that the received RPR frame is a fairness frame. If information indicating another type of frame is described in the header, it may be determined that the received RPR frame is not a fairness frame. If the received RPR frame is a fairness frame, the frame analysis unit 410-1 and the frame analysis unit 410-2 correspond to the fairness frame in the fairness control unit 560-1 and the fairness control unit 560-2. Send to If the received RPR frame is not a fairness frame, the RPR frame is sent to the RPR switch processing unit 420.

  Specifically, the frame analysis unit 410-1 determines whether or not the RPR frame received by the input port 400-1 from the adjacent RPR node (in this example, the RPR node 11) is a fairness frame. Is determined, the RPR frame is sent to the fairness control unit 560-2. If it is determined that the frame is not a fairness frame, the RPR frame is sent to the RPR switch processing unit 420.

  Similarly, the frame analysis unit 410-2 determines whether or not the RPR frame received by the input port 400-2 from the adjacent RPR node (in this example, the RPR node 170) is a fairness frame, and is a fairness frame. Is determined, the RPR frame is sent to the fairness control unit 560-1. If it is determined that the frame is not a fairness frame, the RPR frame is sent to the RPR switch processing unit 420.

  The RPR switch processing unit 420 executes various processes related to RPR defined in “IEEE Standards 802.17”.

  As an example of processing executed by the RPR switch processing unit 420, transfer processing of an RPR frame received from an adjacent RPR node (in this example, the RPR node 11 and the RPR node 170), received from a subordinate terminal (in this example, the terminal 200) A process of generating an RPR frame from an Ethernet frame and transmitting it within the RPR network, a process of extracting an Ethernet frame stored in the RPR frame addressed to its own node and transmitting it to the subordinate terminal 200, and RPR using Topology Discovery Protocol (TDP) There are a process for managing network topology information, a process for managing an RPR network by OAM (Operations, Administration, Maintenance), and the like.

  The node of the present invention is applied with an RPR node redundancy technique that prevents redundant nodes from overlapping and transmitting Ethernet frames to subordinate terminals. For example, as described later, the TDB storage unit 530 of the node of the present invention stores in advance the RPR MAC address of each redundant node. The RPR switch processing unit 420 extracts the Ethernet frame from the RPR frame not only to the RPR frame addressed to the own node but also to the other redundant node and extracts the Ethernet frame from the RPR frame to the subordinate terminal 200. A process of transmitting and discarding the RPR frame is performed. Further, when it is detected that the link between the own node and the terminal 200 under its control is disconnected, the RPR switch processing unit 420 receives the RPR frame addressed to the own node or another redundant node. For example, the RPR frame is transferred to the next RPR node. For example, when the RPR node 10 receives an RPR frame addressed to itself or RPR 11 from the RPR node 170, the RPR node 10 transmits the Ethernet frame stored in the RPR frame to the terminal 200 and discards the RPR frame. At this time, if the link between the RPR node 10 and the terminal 200 is disconnected, the RPR frame addressed to the own node or the RPR 11 is transferred to the next RPR node 11. In this case, the RPR node 11 transmits an Ethernet frame to the terminal 200.

  The RPR switch processing unit 420 sends a frame to be transferred to an RPR node (RPR node 11 in this example) adjacent in the clockwise direction to the frame multiplexing unit 440-1 via the traffic measurement unit 470-1. Similarly, a frame to be transferred to the RPR node adjacent in the counterclockwise direction (RPR node 170 in this example) is sent to the frame multiplexing unit 440-2 via the traffic measurement unit 470-2.

  Hereinafter, the details of the processing of the RPR switch processing unit 420 are omitted except for the operations deeply related to the operation of the present invention.

  The RPR switch processing unit 420 includes a buffer corresponding to the output port 450-1 and a buffer corresponding to the output port 450-2. The buffer corresponding to the output port 450-1 holds the fairness control target traffic among the traffic scheduled to be transmitted from the output port 450-1. Similarly, the buffer corresponding to the output port 450-2 holds the fairness control target traffic among the traffic scheduled to be transmitted from the output port 450-2. When congestion occurs, the use amount of these buffers (capacity of fairness control target traffic) increases. Therefore, it can be determined that congestion has occurred when the amount of use of these buffers is equal to or greater than the threshold. For example, when the usage amount of the buffer corresponding to the output port 450-1 exceeds a threshold, it can be determined that congestion has occurred in the link from the RPR node 10 to the RPR node 11. Similarly, when the usage amount of the buffer corresponding to the output port 450-2 exceeds the threshold, it can be determined that congestion has occurred in the link from the RPR node 10 to the RPR node 170. When the usage amount of the buffer corresponding to the output port 450-1 exceeds the threshold value, the RPR switch processing unit 420 notifies the fairness control unit 460-1 to that effect. Further, when the usage amount of the buffer corresponding to the output port 450-2 exceeds the threshold value, the fairness control unit 460-2 is notified to that effect.

  The TDB storage unit 530 is a storage device that stores a TDB (topology database). The TDB is a database in which the topology information of the RPR network to which the own node belongs is registered as a result of the TDP process executed by the RPR switch processing unit 420 of the own node (here, the RPR node 10). In the TDB, the number of RPR nodes constituting the RPR network, the position of each RPR node, the failure information of each link, and the like are registered as topology information by the RPR switch processing unit 420. Further, the TDB is referred to by each fairness control unit 560-1 and 560-2 and the RPR switch processing unit 420.

  The RPR switch processing unit 420 periodically generates a control frame (Topology Protection frame; hereinafter referred to as a TP frame) for each RPR node to create a TDB, and broadcasts it to each ringlet. When generating the TP frame, the RPR switch processing unit 420 stores the RPR MAC address of its own node in the TP frame as a transmission source address. In addition, the RPR switch processing unit 420 has a failure in a link between TTL (Time To Live), an initial value of TTL (hereinafter referred to as TTL base), and each adjacent RPR node. Information indicating whether or not is stored in the TP frame. At this time, the TTL base and TTL may be determined based on the number of RPR nodes registered in the TDB when the TP frame is created. The information stored in the TP frame is not limited to the above information.

  When receiving a TP frame transmitted from another RPR node, the RPR switch processing unit 420 subtracts 1 from the TTL in the TP frame. The RPR switch processing unit 420 transfers the TP frame to the next RPR node if the subtraction result is not 0, and discards the TP frame if the subtraction result is 0.

  In addition, when the RPR switch processing unit 420 receives a TP frame from another node, the RPR switch processing unit 420 creates a TDB as follows. The received TP frame includes a TTL and a TTL base (an initial value of TTL). Therefore, a hop count (a value indicating the number of hops from which the TP packet transmission source node exists) can be obtained from the TTL and the TTL base. For each TP packet received from the input port 400-1, the RPR switch processing unit 420 obtains a value obtained by subtracting TTL from the TTL base as a hop count, and the RPR MAC address of the transmission source of the TP packet in the order of hop count Are stored in the TDB storage unit 530 as topology information. Similarly, for each TP packet received from the input port 400-2, for each TP packet, a value obtained by subtracting TTL from the TTL base is obtained as the hop count, and the RPR MAC address of the transmission source of the TP packet in the hop count order. Are stored in the TDB storage unit 530 as topology information.

  Further, the TP frame includes information indicating whether or not a failure has occurred in the link between the transmission source node and each RPR node adjacent to the transmission source node. In other words, when a failure has occurred, it can be determined from this information which link among the links between the RPR nodes has caused the failure. The RPR switch processing unit 420 refers to the information in the TP frame, and when a failure occurs in the link between the RPR nodes, the failure information indicating which link has the failure as topology information as TDB The data is stored in the storage unit 530.

  The RPR MAC address and link failure information stored in this way are the TDB.

  In addition to the RPR switch processing unit 420 of the RPR node 10 illustrated in FIG. 2, the RPR switch processing unit 420 of each node periodically transmits a TP frame. Then, the RPR switch processing unit 420 of each node periodically recreates the TDB using TP frames received from the other RPR nodes. The failed RPR node does not transmit a TP frame. That is, it is possible to determine whether the RPR node is normal or failed depending on whether or not the RPR MAC address is registered in the TDB. The RPR switch processing unit 420 and each fairness control unit 560-1 and 2 determine that the RPR node in which the RPR MAC address is registered in the TDB is normal, and the RPR node in which the RPR MAC address is not registered in the TDB Judge that it is broken.

  Further, the RPR switch processing unit 420 determines whether or not a failure has occurred in the link between the own node and the subordinate terminal 200. For example, the RPR switch processing unit 420 and the terminal 200 transmit keep alive frames to each other at regular intervals. When a certain number of keep-alive frames have not arrived from the terminal (when no keep-alive frames have been received from the terminal for a certain period), the RPR switch processing unit 420 has a failure in the link between the own node and the terminal. It is determined that it has occurred. Also, the RPR switch processing unit 420 acquires whether or not a failure has occurred in the link between the own node and the terminal by acquiring information on the link status in the physical layer via the input port 400-3. You may judge.

  In addition, the RPR switch processing unit 420 notifies each of the other redundant nodes whether or not a failure has occurred in the link between the own node and the terminal under its control. For example, the RPR switch processing unit 420 of each redundant node sends a special frame indicating whether or not a failure has occurred in the link between the own node and the subordinate terminal to each of the other redundant nodes at a constant cycle. Notify by sending. If a failure has occurred in the link between the own node and the terminal under its control, a special frame indicating that fact is transmitted to each of the other redundant nodes. A TP frame may be used as this special frame.

  Further, when the frame analysis unit 410-1 and the frame analysis unit 410-2 receive the special frame (which may be a TP frame) from another redundant node, the special frame is subjected to fairness control. To each of the unit 460-1 and the fairness control unit 460-2. The fairness control unit 460-1 and the fairness control unit 460-2 receive a special frame indicating a link failure between the subordinate terminals, and a link failure occurs between which redundant node and the subordinate terminal. Identify whether or not In addition, the number of redundant nodes in which a link failure has occurred with the terminal can be counted.

  In addition, as an RPR node redundancy technique, when not only an RPR frame addressed to the own node but also an RPR frame addressed to another redundant node is received, an Ethernet frame is extracted from the RPR frame and transmitted to the subordinate terminal. If a redundant technology for transmission is used, a frame is received that has another redundant node as the transmission source and is destined for the local node or another redundant node (redundancy node different from the transmission source). Then, it can be determined that a failure has occurred in the link between the redundant node of the transmission source and the terminal. Also in this case, the frame analysis unit 410-1 and the frame analysis unit 410-2 send such a frame to the fairness control unit 460-1 and the fairness control unit 460-2. The fairness control unit 460-1 and the fairness control unit 460-2 may determine that a failure has occurred in the link between the redundant node that is the transmission source of the frame and the terminal.

  The setting interface unit 580 is an interface to which position information of each redundant node is input from a network administrator. The setting interface unit 580 stores the input position information in the TDB storage unit 530. Therefore, the TDB storage unit 530 stores the position information of each redundant node in addition to the TDB. Here, the position information of each redundant node is information that lists, for example, the RPR MAC addresses of the redundant nodes in the order of arrangement of the redundant nodes in the RPR network. The addresses of the respective redundant nodes are listed in the arrangement order, and the fairness frame is sent from the adjacent node on the downstream side. It is possible to determine the presence / absence of a redundant node and the arrangement order of redundant nodes on the upstream side. For example, the RPR MAC address of the downstream adjacent node that has transmitted the fairness frame can be specified from the topology database. Then, with reference to the position information, it is possible to determine that the address after the next address arranged next to the address of the adjacent node on the downstream side and the address of the own node is the address of the redundant node on the upstream side of the own node. If there is an address of the redundant node on the upstream side, it can be determined that there is a redundant node on the upstream side. If there is no next address arranged next to the address of the adjacent node on the downstream side and the address of the own node, it can be determined that the own node is the most upstream address and there is no redundant node on the upstream side. If the address information of the adjacent node on the downstream side is not included in the position information, it can be determined that the own node is the most downstream redundant node. Further, the total number of RPR MAC addresses included in the location information represents the number of redundant nodes.

  The frame multiplexing unit 440-1 multiplexes the data frame sent from the traffic rate measuring unit 470-1 and the fairness frame sent from the fairness control unit 560-2. Then, the frame multiplexing unit 440-1 transmits the multiplexed frame to the RPR node (the RPR node 11 in this example) adjacent in the clockwise direction via the output port 450-1.

  Similarly, frame multiplexing section 440-2 multiplexes the data frame sent from traffic rate measuring section 470-2 and the fairness frame sent from fairness control section 560-1. Then, the frame multiplexing unit 440-2 transmits the multiplexed frame to the adjacent RPR node (RPR node 170 in this example) via the output port 450-2 in the counterclockwise direction.

  The traffic rate measurement unit 470-1 executes various traffic rate measurement processes defined in IEEE 802.17, and then displays the measurement results and the results of computations performed on the measurement results as fairness of the own node. Notify controller 560-1. Similarly, the traffic rate measuring unit 470-2 also executes various traffic rate measuring processes and notifies the fairness control unit 560-2 of the own node of the measurement results.

  Examples of traffic rates measured by the traffic rate measuring units 470-1 and 470-2 include traffic rates transmitted from the own node and subjected to fairness control, traffic transmitted from the upstream node, and passing through the own node. Late etc.

  Each traffic rate measuring unit 470-1-2 also performs calculations such as smoothing processing and normalization processing using a low-pass filter on the measured traffic rate.

  The fairness control unit 560-1 and the fairness control unit 560-2 execute various processes related to fairness defined in “IEEE Standards 802.17”.

  Examples of processing executed by the fairness control unit 560-1 and the fairness control unit 560-2 include detection of congestion at the output port 450-1 or output port 450-2 of the own node, calculation of fair rate, and generation of a fairness frame. Transmission, analysis of fairness frames received from adjacent nodes, notification of bandwidth control information to the RPR switch processing unit 420 of the own node, and the like.

  The fairness control unit 560-1 and the fairness control unit 560-2 included in the node 10 of the present invention perform the following process when receiving the fairness frame. In the following description, the fairness control unit 560-1 will be described as an example, but the operation of the fairness control unit 560-2 is the same.

  When the fairness control unit 560-1 receives the fairness frame from the RPR node (herein, RPR node 170) downstream of the own node (here, the RPR node 10), the fairness control unit 560-1 sets the own node as the tail node. And a process for determining the upper limit value of the communication bandwidth of the fairness control target traffic that is received by the node from the terminal and that passes through the congestion point. These two processes are different from the processes performed by the general RPR nodes 120 to 170 when a fairness frame is received. The general RPR nodes 120 to 170 perform the processing shown in FIG. 8 and set the fair rate stored in the fairness frame as the upper limit value of the communication bandwidth of the fairness control target traffic.

  FIG. 3 is a flowchart illustrating an example of processing in which the fairness control unit included in the node of the present invention determines whether or not the own node is a tail node. When the frame analysis unit 410-2 of the RPR node 10 receives the fairness frame from the downstream RPR node 170 and sends the fairness frame to the fairness control unit 560-1 of its own node, the fairness control unit 560-1 It operates as follows.

  The fairness control unit 560-1 determines whether or not the adjacent node on the upstream side of the own node is a redundant node (step S21). This determination may be made by referring to the position information stored in the TDB storage unit 530.

  When the upstream adjacent node is a redundant node (Yes in step S21), the fairness control unit 560-1 receives the redundant node from any of the redundant nodes on the upstream side, from the terminal under its control. It is determined whether the fairness controlled traffic transmitted to the downstream side can be received (step S22). That is, it is determined whether the control target traffic can be received from the redundant node on the upstream side that can receive the fairness control target traffic from the terminal. More specifically, in any of the redundant nodes on the upstream side, the link to the terminal is normal, there is no failure, and each link to the local node and each redundant node are all normal. It is determined whether there is a redundant node. The determination process in step S22 will be described later.

  When it is determined that the control target traffic can be received from the upstream redundant node that can receive the fairness control target traffic from the terminal (Yes in step S22), the process proceeds to step S24.

  When the upstream adjacent node is not a redundant node (No in step S21), or when it is determined that the control target traffic cannot be received from the upstream redundant node that can receive the fairness control target traffic from the terminal (step S22). No), the process proceeds to step S23.

  When the process proceeds to step S23, the fairness control unit 560-1 determines whether the communication band of the fairness control target traffic received from the upstream adjacent node is less than the fair rate (step S23). This fair rate is a fair rate stored in a fairness frame received from the downstream side. If the communication band of the fairness control target traffic received from the upstream adjacent node is equal to or higher than fair rate (No in step S23), the process proceeds to step S24. If the communication band of the fairness control target traffic received from the upstream adjacent node is less than the fair rate (Yes in step S23), the process proceeds to step S25.

  If it transfers to step S24, the fairness control part 560-1 will judge that a self-node is not a tail node (step S24). In step S24, the fairness control unit 560-1 transmits the fairness frame received from the downstream side to the upstream RPR node.

  If it transfers to step S25, the fairness control part 560-1 will judge that an own node is a tail node (step S25). In step S25, the fairness control unit 560-1 transmits a fairness frame (a fairness frame indicating that communication band suppression is not required) storing the full rate to the upstream RPR node.

  As described above, when the fairness control unit 560-1 determines Yes in step S21 and determines Yes in step S22, the fairness control unit 560-1 does not depend on the communication band of the fairness control target traffic received from the upstream adjacent node. Notification of the fair rate notified from the downstream side to the upstream adjacent node.

  FIG. 4 shows that in any one of the redundant nodes on the upstream side, the link to the terminal is normal, there is no failure, and each link to the own node and each redundant node are all normal. It is a flowchart which shows process progress of the process (step S22) which determines whether the redundant node which is is.

  When the process proceeds to step S22 shown in FIG. 3, the fairness control unit 560-1 selects redundant nodes upstream from the own node in order of proximity to the own node, and for each selected redundant node, the step A3, etc. Judgment processing is performed. Let n be a variable that represents the number of redundant nodes on the upstream side. For example, if n = 1, it means that the redundant node closest to the own node is selected from the upstream redundant nodes. When the process proceeds to step S22, the fairness control unit 560-1 sets n to an initial value 0 (step A1). Next, n is incremented by 1 (step A2).

  After step A2, the fairness control unit 560-1 determines the link (n between the selected redundant node (nth redundant node) and the redundant node adjacent to the downstream side of the selected redundant node. It is determined whether or not a failure has occurred (step A3). The fairness control unit 560-1 may refer to the TDB stored in the TDB storage unit 530 to determine whether or not a failure has occurred in the upstream n-th link.

  If there is no failure in the nth link on the upstream side (No in step A3), the fairness control unit 560-1 determines whether the selected redundant node (nth redundant node) has failed. Is determined (step A4). The address of the nth redundant node can be specified as the nth address arranged upstream from the address of the own node in the position information. If the address of the nth redundant node is registered in the topology database, the fairness control unit 560-1 determines that the redundant node is normal (No in step A4). If the address of the selected nth redundant node is not registered in the topology database, it is determined that the redundant node has failed (Yes in step A4).

  When a failure has occurred in the nth link on the upstream side (Yes in Step A3), or when the nth redundant node on the upstream side has failed (Yes in Step S4), the fairness control unit 560 -1 sets No as the determination result in step S22 (step A8).

  If the selected nth redundant node is normal, it is determined whether or not a failure has occurred in the link between the nth redundant node and the terminal under it (step A5). For example, if a special frame indicating that a failure has occurred in the link with the subordinate terminal is received from the selected redundant node, it may be determined that there is a failure in the link. If a special frame indicating that no failure has occurred in the link with the subordinate terminal is received from the selected redundant node, it may be determined that there is no failure in the link. Moreover, you may perform determination of step A5 with another method.

  If no failure has occurred in the link between the selected nth redundant node and the terminal under the node (No in step A5), the fairness control unit 560-1 sets the determination result in step S22 to Yes. (Step A7).

  On the other hand, when a failure has occurred in the link between the selected nth redundant node and the terminal under it (Yes in step A5), the nth redundant node is the most upstream redundant node. It is determined whether or not there is (step A6). This determination may be performed with reference to position information stored in advance in the TDB storage unit 530. If there is no address of a redundant node upstream from the selected redundant node, it may be determined that the selected redundant node is the most upstream redundant node. In that case, the determination result of step S22 is No (step A8).

  If the position information includes the address of the redundant node upstream from the selected redundant node, the selected redundant node is not the most upstream redundant node (No in step A6). In this case, the process proceeds to step A2, and the processes after A2 are repeated. That is, in step A2, n is incremented by 1, the next redundant node on the upstream side is selected, and determination such as step A3 is performed.

  Next, a process for determining the upper limit value of the communication bandwidth of the fairness control target traffic that passes through the congestion point will be described. FIG. 5 is a flowchart showing the progress of this process. When the fairness frame is transmitted, the fairness control unit 560-1 receives the fairness control target traffic from the terminal and counts the number of redundant nodes that can be transmitted in the RPR network (step S31). The fairness control unit 560-1 is based on the location information stored in the TDB storage unit 530, the topology database, and the information on the redundant node in which a failure has occurred in the link with the terminal. A counting process may be performed. For example, first, among the RPR MAC addresses of the redundant nodes listed in the position information, the addresses included in the topology database are counted. This count value is the number of redundant nodes that have not failed. In other words, if traffic can be received from a terminal, this is the number of redundant nodes that can transmit the traffic into the RPR network. By subtracting the number of nodes of other redundant nodes that have notified that a failure has occurred in the link with the terminal from this count value, the traffic for fairness control is received from the terminal and entered into the RPR network. The number of redundant nodes that can be transmitted can be determined. The number of other redundant nodes that have notified that a failure has occurred in the link with the terminal, for example, received a special frame indicating that a failure has occurred in the link with the terminal Is a number.

  Subsequently, the fairness control unit 560-1 is the number of redundant nodes that can receive the fairness control target traffic from the terminal and transmit it in the RPR network, and the fairness control unit 560-1 stores the fairness stored in the fairness frame received from the downstream side. Divide the rate. Then, the division result is determined as the upper limit value of the communication bandwidth of the fairness control target traffic received from the terminal and passing through the congestion point (step S32).

  In step S31, the fairness control unit 560-1 determines whether there is a failed redundant node on the upstream side of the own node, and the failed redundancy on the upstream side of the own node. If there is a node, it receives the fairness controlled traffic from the terminal from the most downstream redundant node to the redundant node adjacent to the downstream side of the failed redundant node, and enters the RPR network. Redundant nodes that can be transmitted may be counted. For example, the fairness control unit 560-1 identifies the address of the redundant node on the upstream side of the own node from the position information, and if there is an address that is not included in the topology database among the addresses, What is necessary is just to determine with the redundant node having a failure in the upstream. The fairness control unit 560-1 then counts the number of addresses from the most downstream redundant node to the redundant node adjacent to the downstream side of the failed redundant node with reference to the position information. Then, from the counted value, a failure has occurred in the link between the redundant node from the most downstream redundant node to the redundant node adjacent to the downstream side of the failed redundant node. What is necessary is just to subtract the number of the other redundant node which notified that it is. In step S32, the fair rate may be divided by the number of nodes and determined as the upper limit value of the communication band. However, even if there is a failed redundant node upstream from the own node, the number of nodes is not counted by the counting method described in step S31 described above, but by the counting method described above. Also good.

  In the above description, the fairness control unit 560-1 has been described as an example, but the operation of the fairness control unit 560-2 is the same.

  When the RPR switch processing unit 420 transmits fairness control target traffic that is received from a terminal and passes through a congestion point, the RPR switch processing unit 420 determines the traffic bandwidth of the fairness control unit 560-1 and the fairness control unit. 560-2 controls below the upper limit determined.

  Note that the RPR switch processing unit 420 may determine whether or not the traffic transmitted from the own node passes through the congestion point, for example, as follows. The source address of the fairness frame stores the RPR MAC address of the RPR node immediately upstream of the congestion point. The RPR switch processing unit 420 refers to this address and the TDB and determines how many hops ahead the congestion has occurred. Alternatively, the congestion location may be determined by calculating the number of hops to the congestion location based on the difference between the TTL stored in the fairness frame and its initial value. Then, it is only necessary to determine whether or not the traffic passes through the congestion point by referring to the destination address of the data frame constituting the traffic and the TDB. Further, it may be determined whether the traffic passes through the congestion point by comparing the TTL (current value) of the data frame constituting the traffic with the number of hops to the congestion point.

  Further, when a fairness frame storing a full rate is received, the fairness control units 560-1 and 560-2 do not determine the upper limit value of the communication band, and the RPR switch processing unit 420 does not perform band suppression on the fairness control target traffic. .

  Hereinafter, the operation will be described by taking as an example a case where congestion occurs in the link from the RPR node 150 to the RPR node 140 in the RPR network 15 illustrated in FIG.

  The fairness control unit 460-2 (see FIG. 10) of the RPR node 150 responds when the communication band used by the traffic at the output port 450-2 of the own node exceeds a predetermined communication band. It is determined that congestion has occurred in the link with the adjacent node 140 in the clockwise direction.

  Further, the RPR switch processing unit 420 of the RPR node 150, when the amount of use of the buffer holding the fairness control target traffic among the traffic scheduled to be transmitted from the output port 450-2 exceeds the threshold value, indicates the fairness. Notify controller 460-2. Even when the fairness control unit 460-2 of the RPR node 150 receives this notification from the RPR switch processing unit 420, the fairness control unit 460-2 determines that congestion has occurred in the link with the adjacent node 140 in the counterclockwise direction of the own node. .

  The fairness control unit 460-2 of the RPR node 150 that has determined that the congestion has occurred advertises the fair rate to the RPR node arranged upstream of the own node (RPR node 150), thereby causing the RPR node 150 to the RPR node 140. Suppresses the rate of fairness-controlled traffic that passes through the outgoing link.

  In “IEEE Standards 802.17”, the value of the first fair rate to be advertised (the initial value of the fair rate) is the rate of the fairness controlled traffic transmitted from the RPR node 150 into the RPR network immediately before the congestion is detected. It is stipulated that there is. The traffic rate measuring unit 470-2 of the RPR node 150 measures the rate of fairness control target traffic that is the source of the own node and the rate of fairness control target that is transmitted from the upstream RPR node and passes through the own node. ing. The fairness control unit 460-2 uses the sum of the rates immediately before the congestion as a fair rate and transmits a fairness frame storing the fair rate to the upstream side.

  Even after first advertising the fair rate, the traffic rate measuring unit 470-2 of the RPR node 150 calculates the fair rate according to the provisions of “IEEE Standards 802.17” at a predetermined time interval and transmits it upstream. To do.

  When the RPR node 160, which is an adjacent node upstream of the RPR node 150, receives the fairness frame from the RPR node 150, the traffic passing through the congestion point among the fairness control target traffic received from the terminal 260 under its own node is displayed. The communication band to be used is controlled to be equal to or less than the fair rate stored in the fairness frame.

  Further, the RPR node 160 stores the fair rate notified from the RPR node 150 when the communication band of the fairness control target traffic received from the RPR node 170 which is an adjacent node upstream of the own node is equal to or higher than the fair rate. A fairness frame is transmitted to the RPR node 170. On the other hand, when the communication band of the fairness control target traffic received from the RPR node 170 is less than the fair rate, the RPR node 160 determines that the own node is a tail node. Then, the RPR node 170 is notified by a fairness frame that it is not necessary to suppress the communication bandwidth of the fairness control target traffic. That is, the fairness frame storing the full rate is transmitted to the RPR node 170.

  The operation of another general RPR node 170 on the upstream side is the same as the operation of the RPR node 160 described above.

  Next, the operation when the RPR nodes 10 and 11 of the present invention receive a fairness frame storing a fair rate that is not full-rate from a downstream RPR node will be described. Here, it is assumed that none of the redundant nodes 10 and 11 has failed and no failure has occurred in any link.

  When receiving the fairness frame from the RPR node 170, the frame analysis unit 410-2 of the RPR node 10 of the present invention sends it to the fairness control unit 560-1. The fairness control unit 560-1 performs a communication band upper limit determination process (see FIG. 5). First, the fairness control unit 560-1 receives the fairness control target traffic from the terminal and counts the number of redundant nodes that can be transmitted in the RPR network (step S31). As described above, this counting process is performed based on the location information stored in the TDB storage unit 530, the topology database, and the information of the redundant node in which a failure occurs in the link with the terminal. Just do it. In this example, the redundancy nodes 10 and 11 are normal, and the links between the terminal 200 and the redundancy nodes 10 and 11 are also normal. Therefore, the fairness control unit 560-1 counts the number of nodes as 2. To do. Then, the fairness control unit 560-1 is a fairness control target traffic received by the own node from the terminal, which is a value obtained by dividing the fair rate stored in the fairness frame by the number of nodes of 2, and passes through the congestion point. It is determined as the upper limit value of the communication bandwidth of the control target traffic. That is, 1/2 of the fair rate is set as the upper limit value of the communication band. Then, the RPR switch processing unit 420 of the RPR node 10 suppresses the communication band of the fairness control target traffic that is received from the terminal 200 and passes through the congestion point to ½ or less of the fair rate.

  In addition, the fairness control unit 560-1 of the RPR node 10 performs processing (see FIG. 3) for determining whether or not the own node is a tail node. In this example, the fairness control unit 560-1 determines that the upstream adjacent node is a redundant node (Yes in step S21), and proceeds to step S22. The redundant node 11 has not failed, and no failure has occurred in the link between the redundant nodes 10 and 11 and the link between the redundant node 11 and the terminal 200. Therefore, the fairness control unit 560-1 performs processing in the order of steps A1, A2, A3, A4, A5, and A7 in the processing of step S22 (see FIG. 4). Then, the process of step S22 is terminated, and the link with the terminal is normal in any one of the redundant nodes on the upstream side, is not broken, and each link and It is determined that there is a redundant node in which all redundant nodes are normal (Yes in step S22). After the process of step S22, the fairness control unit 560-1 determines that the own node is not a tail node (step S24), and transmits the fairness frame received from the downstream side to the upstream RPR node 11.

  When the RPR node 11 receives the fairness frame from the RPR node 10, the fairness control unit 560-1 of the RPR node 11 determines the upper limit value of the communication band in the same manner as the fairness control unit 560-1 of the RPR node 10 described above. . That is, the upper limit value of the communication bandwidth of the fairness control target traffic that is received by the own node from the terminal and passes through the congestion point is determined to be 1/2 of the fair rate. Then, the RPR switch processing unit 420 of the RPR node 11 suppresses the communication band of the fairness control target traffic that is received from the terminal 200 and passes through the congestion point to ½ or less of the fair rate.

  In addition, the fairness control unit 560-1 of the RPR node 11 performs processing (see FIG. 3) for determining whether or not the own node is a tail node. In this case, the fairness control unit 560-1 determines that the upstream adjacent node is not a redundant node, and proceeds to step S23. If the rate of the fairness control target traffic received from the adjacent node upstream of the own node is equal to or higher than the fair rate (No in step S23), it is determined that the own node is not a tail node (step S24), and the RPR node 10 The fairness frame storing the notified fair rate is transmitted to the upstream RPR node 120. If the rate of the fairness control target traffic received from the adjacent node upstream of the own node is less than the fair rate (Yes in step S23), it is determined that the own node is the tail node (step S25), and the fairness storing the full rate is stored. The frame is transmitted to the upstream RPR node 120.

  As a result of the above operation, when the RPR node 10 on the downstream side of the redundant RPR nodes 10 and 11 belongs to the congestion area, the upstream RPR node 11 is always included in the congestion area. The total communication bandwidth that can be used by each of the redundant RPR nodes 10 and 11 for transmitting the traffic subject to fairness control is equal to the fair rate. Therefore, the communication bandwidth that can be used by the terminals 250 to 270 under the general RPR nodes 150 to 170 included in the congestion area to transmit the traffic subject to fairness control, and the terminals 200 under the redundancy nodes 10 and 11 are fairness. The communication bandwidth that can be used for transmission of the control target traffic is the same. As a result, the communication band of the RPR network can be shared fairly among those terminals.

  Next, an operation when one of the redundant RPR nodes 10 and 11 fails will be described. Here, the operation of the RPR node 10 when the RPR node 11 fails will be described, but the operation of the RPR node 11 when the RPR node 10 fails is the same.

  When the terminal 200 detects a failure of the RPR node 11, the terminal 200 transmits the traffic transmitted to the RPR node 11 to the RPR node 10.

  In this state, it is assumed that the RPR node 10 receives a fairness frame storing a fair rate that is not full rate from the downstream RPR node 170. When the fairness frame is sent to the fairness control unit 560-1 of the RPR node 10, the fairness control unit 560-1 performs a process of determining whether or not the own node is a tail node (see FIG. 3). In this example, the fairness control unit 560-1 determines that the upstream adjacent node is a redundant node (Yes in step S21), and proceeds to step S22. In this example, the first redundant node 11 on the upstream side has failed. Accordingly, the fairness control unit 560-1 proceeds in the order of steps A1, A2, A3, A4, and A8 in the process of step S22 (see FIG. 4), finishes the process of step S22, and then proceeds to step S23 (see FIG. 3). .) In addition, since the redundant node 11 adjacent to the upstream side is out of order, no traffic is received from the redundant node 11. Accordingly, the communication band of the fairness control target traffic from the upstream is always less than the fair rate, and it is determined that the own node is the tail node.

  In addition, the fairness control unit 560-1 performs a communication band upper limit determination process (see FIG. 5). The fairness control unit 560-1 receives the fairness control target traffic from the terminal and counts the number of redundant nodes that can be transmitted in the RPR network (step S31). In this example, since the RPR node 11 is out of order, 1 is counted as the number of redundant nodes that can receive the fairness control target traffic from the terminal and transmit it within the RPR network. The fairness control unit 560-1 is the fairness control target traffic received by the own node from the terminal, which is a value obtained by dividing the fair rate stored in the fairness frame by the number of nodes 1 (ie, the fair rate). It is determined as the upper limit value of the communication bandwidth of the fairness control target traffic that passes through the location. The RPR switch processing unit 420 of the RPR node 10 suppresses the communication band of the fairness control target traffic received from the terminal 200 and passing through the congestion location to be equal to or lower than the fair rate.

  Note that the operation in the case where a failure has occurred in the link between the redundant nodes 10 and 11 is the same as that of the node 11 except that the processing in step S22 is shifted to steps A1, A2, A3, and A8 in order. This is the same as the operation in the case of failure.

  Next, an operation when a failure occurs in a link between one of the redundant RPR nodes 10 and 11 and the terminal 200 will be described. Here, the operation of the RPR nodes 10 and 11 when a failure occurs in the link between the RPR node 11 and the terminal 200 will be described. However, when a failure occurs in the link between the RPR node 10 and the terminal 200. The operation of is the same.

  When the terminal 200 detects a failure of the link between the RPR node 11 and the terminal 200, the terminal 200 transmits the traffic transmitted to the RPR node 11 to the RPR node 10.

  In this state, it is assumed that the RPR node 10 receives a fairness frame storing a fair rate that is not full rate from the downstream RPR node 170. When the fairness frame is sent to the fairness control unit 560-1 of the RPR node 10, the fairness control unit 560-1 performs a process of determining whether or not the own node is a tail node (see FIG. 3). In this example, the fairness control unit 560-1 determines that the upstream adjacent node is a redundant node (Yes in step S21), and proceeds to step S22. In this example, a failure has occurred in the link between the first redundant node 11 on the upstream side and the terminal 200. Accordingly, the fairness control unit 560-1 proceeds in the order of steps A1, A2, A3, A4, A5, A6, and A8 in the process of step S22 (see FIG. 4), and ends the process of step S22. (See FIG. 3). In this example, the case where the redundant node on the upstream side of the RPR node 10 is only the RPR node 11 is illustrated, but when there is a redundant node further on the upstream side, the steps A6 to A2 are performed. And the process after step A2 is repeated.

  The fairness control unit 560-1 compares the rate of the fairness control target traffic received from the RPR node 11 with the fair rate (step S23). If the rate of the fairness control target traffic received from the RPR node 11 is equal to or higher than the fair rate, the own node is not the tail node (step S24), and transmits the fairness frame storing the fair rate to the upstream RPR node 11. . On the other hand, if the rate is less than the fair rate, the own node is a tail node (step S25), and the fairness frame storing the full rate is transmitted to the upstream RPR node 11.

  In addition, the fairness control unit 560-1 performs a communication band upper limit determination process (see FIG. 5). The fairness control unit 560-1 performs the processing of steps S31 and S32 in the same manner as when the RPR node 11 fails, and receives the fair rate from the terminal 200 and passes the traffic band of the fairness control target traffic. Determine the upper limit of.

  With the above operation, even if the RPR node 11 fails or a link failure occurs between the RPR node 11 and the terminal 200, the terminals 250 to 270 and the terminal 200 included in the congestion area are subject to fairness control. The communication bandwidth that can be used for traffic transmission can be made the same.

  In the above example, each redundant node 10 and 11 performs the process shown in steps S31 and S32 to obtain the upper limit of the communication band. The process of determining the upper limit value of the communication bandwidth of the fairness control target traffic received from the terminal 200 and passing through the congestion point is not limited to steps S31 and S32.

  For example, whether or not other redundant nodes have failed in advance, whether or not a link between the other redundant node and the terminal has failed, and each link between redundant nodes has failed. Depending on whether or not there is an upper limit value of the communication bandwidth determined by the own node, it may be determined. In the example illustrated in FIG. 1, the fairness control units 560-1 and 560-2 of the node 10 have no failure in the redundant node 11, and no failure has occurred in the link between the redundant node 11 and the terminal 200. If there is no failure in the link between the redundant nodes 10 and 11, the upper limit value of the communication band is set to ½ of the fair rate, the redundant node 11 is out of order, the redundant node 11 and the terminal 200 In the case where a failure has occurred in the other link or a failure has occurred in the link between the redundant nodes 10 and 11, the upper limit value of the communication band may be determined to be fair. Then, the upper limit value of the communication band may be determined according to the status of other redundant nodes and their links. The same applies to the RPR node 11. At this time, the upper limit value of the communication band determined by each redundant node may be determined so that the sum is equal to the fair rate. For example, the upper limit value of the communication band determined by the RPR node 10 and the upper limit value of the communication band determined by the RPR node 11 are determined so that the sum of the upper limit value and the upper limit value of the communication band determined by the RPR node 11 is equal to the fair rate. It only has to be done.

  In each of the above embodiments, the case where the communication network to which the nodes are connected is an RPR network has been described as an example. The communication system of the present invention may be a communication system other than the RPR network as long as the communication system includes a node that notifies the other node of the fair rate. For example, an MPLS (Multi Protocol Label Switching) network using an LSP (Label Switched Path) as a communication path may be used. Further, the node of the present invention may be a redundant node used in MPLS using the LSP as a communication path.

  In the above embodiment, the nodes are the frame analysis units 410-1 and 410-2, the RPR switch processing unit 420, the frame multiplexing unit 440-1 and 2, the fairness control unit 560-1 and the traffic rate measurement unit 470-1. 2. Although described as a configuration including each processing unit such as the setting interface 580, the node is provided with a computer and a storage device in advance, and the computer operates according to the program for the node stored in the storage device. The structure which implement | achieves the operation | movement similar to each process part may be sufficient. Further, a storage device may be used as the TDB storage unit 530.

  Further, the redundant node included in the communication system of the present invention only needs to include the components shown in FIG. The plurality of redundant nodes 700 included in the communication system of the present invention are arranged adjacent to each other. FIG. 6 shows one of them. Each redundant node 700 includes redundant node determination means (for example, fairness control units 560-1 and 2 that perform the processing of step 21) for determining whether or not the upstream adjacent node is a redundant node. Receivability determination means for determining whether or not the control target traffic can be received from the redundant node on the upstream side capable of receiving the control target traffic whose communication band is controlled from the terminal (for example, fairness control for performing the processing of step 22 560-1 and 2) and the upstream adjacent node is determined to be a redundant node, and it is determined that the control target traffic can be received from the upstream redundant node that can receive the control target traffic from the terminal. Sometimes, the fair rate notification means for notifying the upstream side adjacent node of the fair rate notified from the downstream side adjacent node (for example, Comprising a fairness control unit 560-1～2) and for processing the step 24.

  Each redundant node 700 receives the control target traffic from the terminal and transmits the control target traffic to the communication system, and the upper limit of the communication bandwidth of the predetermined control target traffic set in the redundancy node Communication band determining means (for example, fairness control) that determines the upper limit value of the communication band of the predetermined control target traffic transmitted from the own node so that the sum of the values is equal to the fair rate notified from the adjacent node on the downstream side Part 560-1 to 2).

  By providing such a redundant node, the available communication bandwidth can be made fair between a terminal connected to the redundant node and a terminal connected to another node.

  Further, in the above description, the communication band determining means notifies the downstream node of the number of redundant nodes that can receive the control target traffic from the terminal and transmit the control target traffic into the communication system. It is disclosed that the obtained fair rate is divided and the division result is determined as an upper limit value of a communication band of a predetermined control target traffic transmitted from the own node.

  In the above description, the communication band determining means receives the control target traffic from the terminal and counts the number of redundant nodes that can transmit the control target traffic into the communication system (for example, step The fairness control units 560-1 to 2) that perform the processing of 31 and the number of nodes counted by the counting unit divide the fair rate notified from the downstream adjacent node, and the division result is transmitted from the own node. It is disclosed that it has division means (for example, fairness control units 560-1 and 560-2 that perform the processing of step 32) that is determined as an upper limit value of a communication band of predetermined control target traffic.

  In the above description, the position information storage unit (for example, the TDB storage unit 530) that stores in advance the position information that lists the addresses of the redundant nodes in the arrangement order, the link between the node and the address of the node that has not failed Link failure information indicating that a failure has occurred in the link between the own node and the terminal, and the topology database storage means (for example, TDB storage unit 530) that stores the topology database including information on the failure occurrence location of Failure information notification means (for example, RPR switch processing unit 420) for notifying each other redundant node, and the counting means includes addresses included in the topology database among the addresses listed in the position information. And subtract the number of other redundant nodes that have reported inter-terminal link failure information from the address count result. Therefore, it has been disclosed for counting the number of nodes of redundant nodes that can transmit a control target traffic receives the control target traffic from a terminal in the communication system.

  Further, in the above description, when there is a redundant node that has failed on the upstream side of its own node, the counting means moves from the most downstream redundant node to the downstream side of the failed redundant node. It is disclosed to count the number of redundant nodes that can receive control target traffic from a terminal and transmit control target traffic into a communication system from among adjacent redundant nodes.

  In the above description, the position information storage unit (for example, the TDB storage unit 530) that stores in advance the position information that lists the addresses of the redundant nodes in the arrangement order, the link between the node and the address of the node that has not failed Link failure information indicating that a failure has occurred in the link between the own node and the terminal, and the topology database storage means (for example, TDB storage unit 530) that stores the topology database including information on the failure occurrence location of Failure information notifying means (for example, RPR switch processing unit 420) for notifying each other redundant node, and the reception availability determining means includes one redundant node on the upstream side of the own node in order from the closest to the own node. The link between the selected redundant node and the redundant node adjacent to the downstream side of the selected redundant node has failed. Is determined based on the topology database, whether the selected redundant node is failed or not is determined based on the topology database, and a failure occurs in the link between the selected redundant node and the terminal. Is determined based on whether or not the inter-terminal link failure information is received from the selected redundant node, and the redundant nodes on the upstream side of the own node are selected one by one in order of proximity to the own node It is determined that a failure has not occurred in the link between the selected redundant node and the redundant node adjacent to the downstream side of the selected redundant node, and the selected redundant node has failed. If the next redundant node is selected on the condition that it is determined that there is a failure in the link between the selected redundant node and the terminal, and one of the redundant nodes is selected. If it is determined that no failure has occurred in the link between the selected redundant node and the redundant node adjacent to the downstream side of the selected redundant node, and the selected redundant node has not failed. From the upstream redundant node that can receive the control target traffic whose communication bandwidth is controlled from the terminal when it is determined that there is no failure in the link between the selected redundant node and the terminal. When it is determined that the control target traffic can be received and one of the redundant nodes is selected, it is determined that a failure has occurred in the link with the redundant node adjacent to the downstream side of the selected redundant node. If it is determined that the selected redundant node has failed, or if it is determined that a failure has occurred in the link between each redundant node and the terminal selected in order. It is determined that the control target traffic cannot be received from the redundant node on the upstream side that can receive the control target traffic whose communication band is controlled from the terminal.

  Moreover, the case where a communication system is an RPR network which uses a ringlet as a communication path is disclosed. Furthermore, it is disclosed that the communication system may be an MPLS network using the LSP as a communication path.

  The present invention is preferably applied to a redundant node that is connected to a common terminal and made redundant, and a communication system including the redundant node.

It is explanatory drawing which shows the example of the communication system of this invention. It is a block diagram which shows the structural example of the node of this invention. It is a flowchart which shows the example of the process which determines whether the node of this invention makes a self node a tail node. Redundancy in which any link to the terminal is normal in any of the redundant nodes on the upstream side, is not broken, and each link to each node and each redundant node are all normal It is a flowchart which shows the example of the process which determines whether a node exists. It is a flowchart which shows the example of the process which determines the upper limit of the communication band of the fairness control object traffic which passes a congestion location. It is a block diagram which shows the structural example of the node of this invention. It is explanatory drawing which shows the example of an RPR network. It is a flowchart of a process in which a general RPR node determines whether or not its own node is a tail node. It is explanatory drawing which shows the example of the RPR network containing the RPR node made redundant. It is a block diagram which shows the structural example of a RPR node.

Explanation of symbols

10,11 RPR node (redundant node)
120 to 170 RPR node 200 terminal 400-1 to 3 input port 410-1 to 2 frame analysis unit 440-1 to 2 frame multiplexing unit 450-1 to 3 output port 470-1 to 2 traffic rate measurement unit 530 TDB storage unit 560-1 to 2 Fairness control unit 580 Setting interface 700 Redundant node 710 Redundant node determining unit 720 Receivability determining unit 730 Fairy notify unit 740 Communication band determining unit

Claims (18)

A communication system comprising a node for notifying the other node of a fair rate that is an upper limit value of a communication bandwidth permitted to another node,
A plurality of redundant nodes, which are nodes connected to a common terminal,
Each redundant node is placed adjacent to each other,
Each redundant node
When transmitting traffic to the congestion location side, when the transmission direction of the traffic is the downstream side, a redundant node determination unit that determines whether the upstream adjacent node is a redundant node; and
Receivability determining means for determining whether or not the control target traffic can be received from the redundant node on the upstream side capable of receiving the control target traffic whose communication band is controlled from the terminal;
When it is determined that the upstream adjacent node is a redundant node and it is determined that the control target traffic can be received from the upstream redundant node capable of receiving the control target traffic from the terminal, the downstream adjacent node A fair rate notification means for notifying an upstream adjacent node of the fair rate notified from
The sum of the upper limit values of the communication bandwidths of the predetermined control target traffic set in the redundant node that can receive the control target traffic from the terminal and transmit the control target traffic into the communication system is the downstream side. Communication bandwidth determining means for determining an upper limit value of the communication bandwidth of the predetermined control target traffic transmitted from the own node so as to be equal to the fair rate notified from the adjacent node of
The predetermined control target traffic is control target traffic received from the terminal, and is control target traffic that passes through a congestion point. The communication band determination means is a number of redundant nodes capable of receiving the control target traffic from the terminal and transmitting the control target traffic into the communication system, and determining the fair rate notified from the downstream adjacent node. The communication system according to claim 1, wherein division is performed and the division result is determined as an upper limit value of a communication band of predetermined control target traffic transmitted from the own node. Communication bandwidth determination means
Counting means for receiving the control target traffic from the terminal and counting the number of redundant nodes capable of transmitting the control target traffic into the communication system;
Dividing means for dividing the fair rate notified from the downstream adjacent node by the number of nodes counted by the counting means and determining the division result as an upper limit value of the communication bandwidth of the predetermined control target traffic transmitted from the own node The communication system according to claim 1 or 2. Position information storage means for storing in advance position information listing addresses of each redundant node in the arrangement order;
Topology database storage means for storing a topology database including the address of a node that has not failed and information on the location of the failure of the link between the nodes;
A failure information notification means for notifying each other redundant node of link failure information between terminals indicating that a failure has occurred in the link between the own node and the terminal;
The counting means counts the addresses included in the topology database among the addresses listed in the position information, and subtracts the number of other redundant nodes that have reported the inter-terminal link failure information from the counting result of the addresses. The communication system according to claim 3, wherein the number of redundant nodes capable of receiving the control target traffic from the terminal and transmitting the control target traffic into the communication system is counted. When there is a failed redundant node on the upstream side of the own node, the counting means is from the most downstream redundant node to the redundant node adjacent to the downstream side of the failed redundant node. 5. The communication system according to claim 3, wherein the number of redundant nodes capable of receiving the control target traffic from the terminal and transmitting the control target traffic into the communication system is counted. Position information storage means for storing in advance position information listing addresses of each redundant node in the arrangement order;
Topology database storage means for storing a topology database including the address of a node that has not failed and information on the location of the failure of the link between the nodes;
A failure information notification means for notifying each other redundant node of link failure information between terminals indicating that a failure has occurred in the link between the own node and the terminal;
Receivability determination means
Select one redundant node on the upstream side of its own node one by one in order of proximity to its own node,
Based on the topology database, whether or not a failure has occurred in the link between the selected redundant node and the redundant node adjacent to the downstream side of the selected redundant node,
Determine whether the selected redundant node has failed based on the topology database,
It is determined whether or not a failure has occurred in the link between the selected redundant node and the terminal based on whether or not the inter-terminal link failure information is received from the selected redundant node,
When selecting the redundant nodes on the upstream side of the own node one by one in the order close to the own node, the link between the selected redundant node and the redundant node adjacent to the downstream side of the selected redundant node It is determined that a failure has not occurred, it has been determined that the selected redundant node has not failed, and it has been determined that a failure has occurred in the link between the selected redundant node and the terminal. Select the next redundant node,
When any redundant node is selected, it is determined that a failure has not occurred in the link between the selected redundant node and the redundant node adjacent to the downstream side of the selected redundant node. If it is determined that the redundant node has not failed, and it is determined that no failure has occurred in the link between the selected redundant node and the terminal, the traffic to be controlled whose communication bandwidth is controlled from the terminal It is determined that the control target traffic can be received from the upstream redundant node that can be received,
If any redundant node is selected and it is determined that a failure has occurred in the link with the redundant node adjacent to the downstream side of the selected redundant node, or the selected redundant node has failed If it is determined that a failure has occurred in the link between each redundant node selected in order and the terminal, the traffic to be controlled whose communication bandwidth is controlled from the terminal is determined. The communication system according to any one of claims 1 to 5, wherein it is determined that control target traffic cannot be received from a redundant node on the upstream side that can be received. The communication system according to any one of claims 1 to 6, wherein the communication system is an RPR network using a ringlet as a communication path. The communication system according to any one of claims 1 to 6, wherein the communication system is an MPLS network using an LSP as a communication path. Used as one of a plurality of nodes connected to a common terminal in a communication system including a node that notifies the other node of a fair rate that is an upper limit value of a communication band permitted to another node A redundant node,
When transmitting traffic to the congestion location side, when the transmission direction of the traffic is the downstream side, a redundant node determination unit that determines whether the upstream adjacent node is a redundant node; and
Receivability determining means for determining whether or not the control target traffic can be received from the redundant node on the upstream side capable of receiving the control target traffic whose communication band is controlled from the terminal;
When it is determined that the upstream adjacent node is a redundant node and it is determined that the control target traffic can be received from the upstream redundant node capable of receiving the control target traffic from the terminal, the downstream adjacent node A fair rate notification means for notifying an upstream adjacent node of the fair rate notified from
The sum of the upper limit values of the communication bandwidths of the predetermined control target traffic set in the redundant node that can receive the control target traffic from the terminal and transmit the control target traffic into the communication system is the downstream side. Communication bandwidth determining means for determining an upper limit value of the communication bandwidth of the predetermined control target traffic transmitted from the own node so as to be equal to the fair rate notified from the adjacent node of
The redundancy node, wherein the predetermined control target traffic is control target traffic received from the terminal, and is control target traffic that passes through a congestion point. The communication band determination means is a number of redundant nodes capable of receiving the control target traffic from the terminal and transmitting the control target traffic into the communication system, and determining the fair rate notified from the downstream adjacent node. The redundancy node according to claim 9, wherein division is performed and the division result is determined as an upper limit value of a communication band of predetermined control target traffic transmitted from the own node. Communication bandwidth determination means
Counting means for receiving the control target traffic from the terminal and counting the number of redundant nodes capable of transmitting the control target traffic into the communication system;
Dividing means for dividing the fair rate notified from the downstream adjacent node by the number of nodes counted by the counting means and determining the division result as an upper limit value of the communication bandwidth of the predetermined control target traffic transmitted from the own node The redundant node according to claim 9 or 10. Position information storage means for storing in advance position information listing addresses of each redundant node in the arrangement order;
Topology database storage means for storing a topology database including the address of a node that has not failed and information on the location of the failure of the link between the nodes;
A failure information notification means for notifying each other redundant node of link failure information between terminals indicating that a failure has occurred in the link between the own node and the terminal;
The counting means counts the addresses included in the topology database among the addresses listed in the position information, and subtracts the number of other redundant nodes that have reported the inter-terminal link failure information from the counting result of the addresses. The redundant node according to claim 11, wherein the number of redundant nodes capable of receiving the control target traffic from the terminal and transmitting the control target traffic into the communication system is counted. When there is a failed redundant node on the upstream side of the own node, the counting means is from the most downstream redundant node to the redundant node adjacent to the downstream side of the failed redundant node. The redundant node according to claim 11 or 12, wherein the number of redundant nodes capable of receiving the control target traffic from the terminal and transmitting the control target traffic into the communication system is counted. . Position information storage means for storing in advance position information listing addresses of each redundant node in the arrangement order;
Topology database storage means for storing a topology database including the address of a node that has not failed and information on the location of the failure of the link between the nodes;
A failure information notification means for notifying each other redundant node of link failure information between terminals indicating that a failure has occurred in the link between the own node and the terminal;
Receivability determination means
Select one redundant node on the upstream side of its own node one by one in order of proximity to its own node,
Based on the topology database, whether or not a failure has occurred in the link between the selected redundant node and the redundant node adjacent to the downstream side of the selected redundant node,
Determine whether the selected redundant node has failed based on the topology database,
It is determined whether or not a failure has occurred in the link between the selected redundant node and the terminal based on whether or not the inter-terminal link failure information is received from the selected redundant node,
When selecting the redundant nodes on the upstream side of the own node one by one in the order close to the own node, the link between the selected redundant node and the redundant node adjacent to the downstream side of the selected redundant node It is determined that a failure has not occurred, it has been determined that the selected redundant node has not failed, and it has been determined that a failure has occurred in the link between the selected redundant node and the terminal. Select the next redundant node,
When any redundant node is selected, it is determined that a failure has not occurred in the link between the selected redundant node and the redundant node adjacent to the downstream side of the selected redundant node. If it is determined that the redundant node has not failed, and it is determined that no failure has occurred in the link between the selected redundant node and the terminal, the traffic to be controlled whose communication bandwidth is controlled from the terminal It is determined that the control target traffic can be received from the upstream redundant node that can be received,
If any redundant node is selected and it is determined that a failure has occurred in the link with the redundant node adjacent to the downstream side of the selected redundant node, or the selected redundant node has failed If it is determined that a failure has occurred in the link between each redundant node selected in order and the terminal, the traffic to be controlled whose communication bandwidth is controlled from the terminal is determined. The redundant node according to any one of claims 9 to 13, wherein it is determined that the control target traffic cannot be received from the upstream redundant node capable of receiving. Used as one of a plurality of nodes connected to a common terminal in a communication system including a node that notifies the other node of a fair rate that is an upper limit value of a communication band permitted to another node The computer with the redundant node
When transmitting traffic to the congestion location side, when the transmission direction of the traffic is the downstream side, a redundant node determination process for determining whether the upstream adjacent node is a redundant node,
Receivability determination process for determining whether or not the control target traffic can be received from the upstream redundant node capable of receiving the control target traffic whose communication band is controlled from the terminal,
When it is determined that the upstream adjacent node is a redundant node and it is determined that the control target traffic can be received from the upstream redundant node capable of receiving the control target traffic from the terminal, the downstream adjacent node A fair rate notification process for notifying the upstream neighbor node of the fair rate notified from
The sum of the upper limit values of the communication bandwidths of the predetermined control target traffic set in the redundant node that can receive the control target traffic from the terminal and transmit the control target traffic into the communication system is the downstream side. A communication bandwidth determination process for determining an upper limit value of the communication bandwidth of the predetermined control target traffic transmitted from the own node so as to be equal to the fair rate notified from the adjacent node of
A redundancy node program for determining an upper limit value of a communication band of a control target traffic received from the terminal and passing through a congested portion as an upper limit value of the communication band of the predetermined control target traffic . On the computer,
In the communication bandwidth determination process,
Divide the fair rate notified from the downstream adjacent node by the number of redundant nodes that can receive the control target traffic from the terminal and transmit the control target traffic into the communication system, and the division result The redundant node program according to claim 15, wherein a process for determining the upper limit value of the communication bandwidth of predetermined control target traffic transmitted from the own node is executed. Applied to each redundant node connected to a common terminal in a communication system including a node for notifying the other node of a fair rate that is an upper limit value of a communication band permitted to another node. A method for determining a communication bandwidth upper limit value,
When transmitting traffic to the congestion location side, determine whether the adjacent node on the upstream side is a redundant node when the traffic transmission direction is the downstream side,
Determining whether or not the control target traffic can be received from the redundant node on the upstream side capable of receiving the control target traffic whose communication band is controlled from the terminal;
When determining that the upstream adjacent node is a redundant node and determining that the control target traffic can be received from the upstream redundant node capable of receiving the control target traffic from the terminal, the downstream adjacent node Notify the notified fair array to the adjacent node on the upstream side,
The sum of the upper limit values of the communication bandwidths of the predetermined control target traffic set in the redundant node that can receive the control target traffic from the terminal and transmit the control target traffic into the communication system is the downstream side. Determining the upper limit value of the communication band of the predetermined control target traffic transmitted from the own node so as to be equal to the fair rate notified from the adjacent node of
A method of determining a communication band upper limit value, wherein the control target traffic received from the terminal and passing through a congestion point is used as the predetermined control target traffic. Divide the fair rate notified from the downstream adjacent node by the number of redundant nodes that can receive the control target traffic from the terminal and transmit the control target traffic into the communication system, and the division result The method of determining a communication band upper limit value according to claim 17, wherein the communication band upper limit value is determined as a communication band upper limit value of predetermined control target traffic transmitted from the own node.

Images