JP2010226318A - Ring network - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To accomplish a high-reliability network by avoiding communication from being disabled all over the network due to any malfunction or failure. <P>SOLUTION: A slave node 202 transmits a closed state notification packet 301 notifying presence of a closed port 51, and a master node 100 receives the closed state notification packet 301 and recognizes the presence of a closed port 50 of the master node 100 itself to recognize presence of two closed ports on a network. In the case that the number of closed ports present on the network is 2 or more, a node to cancel its closed state and the port are determined. In the case that the node is the master node 100 itself, the closed state of the closed port 50 is canceled. In the case that the node is the slave node 202, a closing control packet 401 notifying closing cancel is transmitted to the slave node 202, and upon receipt of the closing control packet 401, the slave node 202 cancels the closed state of the closed port 51. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a ring network, and more particularly to a ring network that can detect and recover from a failure in a ring network.

  As means for realizing high reliability of a network, a ring network that constructs a network in a loop shape is widely adopted in order to provide communication path redundancy. In a ring network, in order to avoid a broadcast storm, ring control is required to perform path switching by logically cutting a loop (hereinafter referred to as blocking) and releasing the blocking in the event of a failure.

Therefore, STP (Spanning) described in Non-Patent Document 1, which is a conventional ring control method.
(Tree Protocol) is blocked by putting a single port on the ring network into a blocking state. In the STP, network routes are monitored by periodically exchanging control frames called BPDU (Bridge Protocol Data Unit) between network devices. Here, when a failure occurs in the network, periodic BPDUs do not arrive, and when the maximum age timer elapses, it is detected that a failure has occurred in the network. In the above STP, after detecting a failure, it is assumed that the loop has been disconnected at the failed port, and the blocked port is released without blocking the failed port. Therefore, if BPDU is not received due to transfer delay or transfer buffer leak in the network device, the cause of the loop is detected by erroneously detecting as a failure and releasing the blocked port even though no failure has actually occurred. It was.

Therefore, ITU-T G.I. In ERP (Ethernet (registered trademark) Ring Protection Switching) defined in 8032, when a failure is detected, the port at the failure point is blocked, and then the blocked port is released by a control frame that notifies the release of the blocking. , Avoiding the occurrence of loops. Also, when a failure is recovered, the occurrence of a loop is avoided by closing the port of the master node and then releasing the blocked port at the time of failure. Note that there are Patent Literature 1 and Patent Literature 2 that disclose related technologies.

IEEE 802.1D, 1998 Edition MAC bridges (P58-P109)

JP 2007-259406 A JP 2004-201209 A

  Here, in the conventional ring control system, there is a problem that a loop occurs due to an accidental release of a blocked port (hereinafter referred to as an erroneous release of a blocked port) due to a failure or malfunction of a semiconductor component. In addition, there is a problem that the communication path is disconnected due to the erroneous setting of the blocked port (hereinafter referred to as erroneous setting of the blocked port). In Patent Document 1, although it is possible to detect the occurrence of multiple failures and at least one recovery from multiple failures, there is no disclosure regarding the correspondence between erroneous release of a blocked port and erroneous setting of a blocked port. Further, in Patent Document 2, although there is no blocked port on the ring and a problem of occurrence of a loop can be avoided, there is no disclosure regarding handling of erroneous setting of the blocked port.

  The present invention has been made in order to solve the above-described problems, avoids a problem that a loop occurs due to an accidental release of a blocked port, and sets blocked ports incorrectly at a plurality of locations. By avoiding the problem that the communication path is divided by this, it is possible to avoid communication failure in the entire network due to malfunction or failure, and to realize a highly reliable network.

  In order to achieve the above object, a ring network according to the present invention includes a node that forwards a plurality of packets and has a plurality of ports, the plurality of nodes are connected in a ring shape, and at least one of the nodes Is a master node that monitors the state of the network, and in the ring network in which the other nodes are slave nodes that follow instructions from the master node, the slave node is a blocked port that detects the presence or absence of the slave node's own blocked port When the blocked port is detected by the detecting unit, the blocked state notification packet generating unit for generating a blocked state notification packet for notifying the master node of the presence of the blocked port, and when the blocked port is detected by the blocked port detecting unit, the blocked state notification A control unit that controls the packet generation unit to generate a blocking state notification packet; A transmission unit that transmits a blocking state notification packet generated by the blocking state notification packet generation unit, wherein the master node receives the blocking state notification packet transmitted from the slave node, and the blocking state A blocking state notification packet detecting unit for detecting a notification packet, a blocking port detecting unit for detecting the presence or absence of a blocking port of the master node itself, a ring state monitoring database unit for managing the presence or absence of a blocking port for each port for each node; The ring state monitoring database unit is updated based on the presence or absence of the blocked port of each node in the blocked port detection unit and the blocked state notification packet detection unit of the master node, and the information of the ring state monitoring database unit is Next, determine the node to be blocked and its port according to the number of blocked ports in the ring. When the port to be blocked is the master node itself, the corresponding port of the master node itself is blocked. When the port to be blocked is the slave node, the port to be blocked for the corresponding slave node and the blocking control of the port. A control unit that controls to generate a block control packet that notifies the contents of the block, a block control packet generation unit that generates the block control packet, and a transmission unit that transmits the block control packet. The slave node has a receiving unit that receives the blocking control packet transmitted from the master node, a blocking control packet detecting unit that detects a blocking control packet addressed to itself, and whether or not a blocking control packet that is not addressed to itself exists in its blocking port. Regardless of the block, the relay unit that relays the block control packet and the block control added to the block control packet And a control unit that performs blockage control of the corresponding port based on the contents of the port and the port that performs blockage control.

  According to the present invention, it is possible to detect a failure caused by erroneous setting or erroneous release of a blocked port and automatically restore it, thereby providing a highly reliable ring network.

It is a structural example of the ring type network which concerns on Embodiment 1 of this invention. It is a structural example of the ring type network which concerns on Embodiment 1 of this invention. It is a packet format of the obstruction | occlusion state notification packet which concerns on Embodiment 1 of this invention. It is a packet format of the obstruction | occlusion control packet which concerns on Embodiment 1 of this invention. It is a flowchart of the master node which concerns on Embodiment 1 of this invention. It is an internal block diagram of the master node which concerns on Embodiment 1 of this invention. It is a flowchart of the slave node which concerns on Embodiment 1 of this invention. It is an internal block diagram of the slave node which concerns on Embodiment 1 of this invention. It is a structural example of the ring type network which concerns on Embodiment 2 of this invention. It is a packet format of the obstruction | occlusion state notification packet which concerns on Embodiment 2 of this invention. It is a flowchart of the master node which concerns on Embodiment 2 of this invention. It is an internal block diagram of the master node concerning Embodiment 2 of this invention. It is a flowchart of the slave node which concerns on Embodiment 2 of this invention. It is an internal block diagram of the slave node which concerns on Embodiment 2 of this invention. It is a structural example of the ring type network which concerns on Embodiment 3 of this invention. It is an internal block diagram of the master node concerning Embodiment 3 of this invention. It is a flowchart of the slave node which concerns on Embodiment 3 of this invention. It is an internal block diagram of the slave node which concerns on Embodiment 3 of this invention. It is a structural example of the ring type network which concerns on Embodiment 4 of this invention. It is a flowchart of the slave node which concerns on Embodiment 4 of this invention. It is an internal block diagram of the master node concerning Embodiment 4 of this invention. It is a flowchart of the slave node which concerns on Embodiment 4 of this invention. It is an internal block diagram of the slave node which concerns on Embodiment 4 of this invention.

  Preferred embodiments of a ring network according to the present invention will be described below with reference to the accompanying drawings. The present invention is not limited to the embodiments.

Embodiment 1 FIG.
A ring network according to Embodiment 1 of the present invention will be described below with reference to FIGS. 3 and 6 based on FIGS. In this embodiment, a single ring network will be described. However, the device to be applied, the type of packet, and the number of devices are not limited to this.

  FIG. 1 is a diagram illustrating a configuration example of a normal ring network according to the first embodiment. An operation for recognizing that the master node 100 is a normal ring network without erroneous setting or erroneous release of a blocked port will be described. An operation of recognizing that a loop has occurred on the ring when there is no blocked port, setting the blocked port, and restoring to a normal ring will also be described.

  FIG. 2 is a diagram illustrating a ring network in which two blocked ports exist on the ring due to an erroneous setting of a blocked port, and a communication path is divided. An operation of detecting that there are two blocked ports on the ring (“blocked port misconfiguration”), notifying the release of the blocked port, and releasing the blocked port will be described to restore the normal ring.

  FIG. 3 is an explanatory diagram showing an example of a packet format of a blocking state notification packet for notifying the master node that a slave node having a blocking port has a blocking port. FIG. 4 is an explanatory diagram showing an example of a packet format for notifying the slave node having the blocked port of the release of the blocked state.

  FIG. 5 is an example of a flowchart showing the operation of the master node, and FIG. 6 is an example of an internal block diagram of the master node showing the operation of the master node.

FIG. 7 is an example of a flowchart showing the operation of the slave node, and FIG. 8 is an example of an internal block diagram of the slave node showing the operation of the slave node.

  Next, referring to FIG. 1, an operation for recognizing a normal ring network will be described.

  In FIG. 1, the ring network includes a master node 100 that monitors the state of the network, and slave nodes 200 to 204 that follow instructions from the master node 100. When there is no failure on the network, the master node 100 logically disconnects the port 2 of the master node 100 at the blocked port 50 (hereinafter referred to as “blocked”). It does not go into the loop state above and avoids a broadcast storm. In the master node 100, for example, the blocked port 50 is canceled by mistake (hereinafter referred to as “blocked port erroneous release”), and the blocked ports are set erroneously at a plurality of locations (hereinafter, “ In order to detect “blocked port misconfiguration”), information on the blocked port added to the blocked state notification packet 301 transmitted when each of the slave nodes 200 to 204 has a blocked port is detected, and The number of blocked ports existing on the ring is obtained by detecting the blocked ports of the master node 100 itself.

  In FIG. 1, the slave nodes 200 to 204 do not have a blocked port and do not transmit the blocked state notification packet 301, so the master node 100 does not recognize the blocked ports of the slave nodes 200 to 204. Furthermore, since the master node 100 itself has the blocked port 50, the master node 100 recognizes the number of blocked ports on the ring as 1, and is a normal ring network in which no route is divided and no loop occurs on the network. It becomes possible to recognize something.

  Next, referring to FIGS. 2, 3, and 4, the master node 100 in the ring network in which a plurality of blocked ports exist on the ring and the communication path is divided due to “erroneous setting of blocked ports”. An operation for detecting and recovering from “incorrect setting of a blocked port” will be described.

  In FIG. 2, the master node 100 recognizes that the blocked port 50 exists in its own port 2. On the other hand, the slave node 202 transmits a blocking state notification packet 301 as shown in FIG. 3 from the port 1 and the port 2 to the master node 100 in order to recognize that the blocking port 51 exists in its own port 2. Here, the blocked state notification packet 301 is transmitted from the port 2 without being discarded at the blocked port 51, and is relayed by slave nodes other than the slave node 202. That is, the blocked state notification packet 301 is relayed without being discarded at the blocked port even if the blocked port exists in the slave node.

  The master node 100 that has received the blocking state notification packet 301 can analyze the blocking state notification packet 301 and recognize that the port 2 of the slave node 202 has a blocking port. In this way, the master node 100 recognizes that there are a total of two blocked ports on the ring, that is, the master node 100 port 2 (blocked port 50) and the slave node 202 port 2 (blocked port 51). To do. That is, it is possible to detect “wrong port setting”.

  Here, the master node 100 determines a port to be unblocked in order to restore the normal ring network, that is, to set the number of blocked ports on the ring to one. For example, a case is considered in which the master node 100 generally has a blocked port, and the port to be released from blocking is port 2 (blocked port 51) of the slave node 202. In that case, the master node 100 transmits a block control packet 401 for releasing the block as shown in FIG. 4 from the port 1 and the port 2 of the master node 100 to the slave node 202. The blocking control packet 401 is received, relayed, and transmitted at each node without being discarded at the blocking port existing on the ring, like the blocking state notification packet 301.

  The slave node 202 that has received the block control packet 401 can release the block port of port 2 of the slave node 202 by analyzing the block control packet 401. Thus, by setting the number of blocked ports on the ring to 1, it is possible to restore a normal ring network.

  Next, the operation of the master node 100 will be described with reference to FIGS. FIG. 5 is a flowchart showing an operation of detecting the number of blocked ports on the ring in the master node 100. FIG. 6 is an internal block diagram of the master node 100.

  The master node 100 manages the state of the blocked port existing on the ring in the ring state monitoring database unit 701. For example, in FIG. 1, the master node 100 uses the ring state monitoring database unit 701 to check whether there is a blocked port of its own port 1 and port 2, the source address of the slave node 202, and the blocked port 1 and port 2. It manages the presence or absence of ports.

  When the master node 100 detects a change in the status of the blocked port of the master node 100 itself in the blocked port detection unit 702 (S501), the control unit 700 updates the presence / absence of the blocked port in the ring status monitoring database unit 701 ( S503).

  Also, when the blocking state notification packet 301 from the slave node is received by the receiving unit 703 (S502) and the blocking port is detected by the blocking state notification packet detection unit 704, the control unit 700 detects the blocking state notification packet 301. The ring state monitoring database unit 701 is updated based on the transmission source MAC address and the presence / absence of blocked ports of port 1 and port 2 (S503).

  The control unit 700 obtains the number of blocked ports existing on the ring network by monitoring the ring state monitoring database unit 701 when the ring state monitoring database unit 701 is updated, or at regular intervals or irregularly. (S500). Note that the control unit 700 can accurately grasp the number of blocked ports existing on the ring immediately after obtaining the number of blocked ports or by deleting the ring state monitoring database unit 701 at a constant cycle.

  When the obtained number of blocked ports on the ring network is 1, the control unit 700 recognizes that it is a normal ring network. When the number of blocked ports is 2 or more, the control unit 700 detects “blocked port misconfiguration” and determines a node to be blocked and its port so that the number of blocked ports on the ring is 1. (S504). When releasing the blocked port of the slave node, the control unit 700 generates the blocked control packet 401 by the blocking control packet generation unit 706 based on the information of the node to be blocked and its port, and blocks by the transmitting unit 705. The blocking control packet 401 is transmitted to the node to be canceled (slave node 202 in the case of FIG. 2) (S505).

  On the other hand, when the number of blocked ports is 0, “blocking port release error setting” is detected, and the node to be blocked and its port are determined (S506). For example, in general, the master node has a blocked port so that the master node has a blocked port (blocked port 50 in the case of FIG. 1) (S507).

  Next, the operation of the slave node will be described with reference to FIGS. FIG. 7 is a flowchart showing an operation of generating a block state notification packet 301 and an operation when receiving the block control packet 401 in the slave node. FIG. 8 is an internal block diagram of the slave node.

  When the slave node detects the blocked port of its own node in the blocked port detection unit 802 (S602), the control unit 800 uses the blocked state notification packet generation unit 806 to add the blocked port notification packet. 301 is generated (S603) and transmitted from the transmission unit 805 (S604).

  On the other hand, when the block control packet 401 addressed to the own node is received by the reception unit 803 (S600) and the block control packet detection unit 804 detects a block port, the control unit 800 releases or blocks the block control packet 401. Based on the port information to be set, release or set the blocked port. Note that when the receiving unit 803 receives the blocking status notification packet 301 of another slave node or the blocking control packet 401 addressed to another slave node, the relay unit 807 determines whether the slave node itself has a blocking port or not. Then, the blocking state notification packet 301 and the blocking control packet 401 are relayed and transmitted by the transmission unit 805.

  In this way, when the master node 100 obtains the number of blocked ports on the ring and detects “erroneous release of blocked port” or “erroneous setting of blocked port”, the number of blocked ports on the ring is 1. As described above, by canceling or blocking the ports on the ring, it avoids the problem of loops caused by accidentally releasing the blocked ports, and by setting the blocked ports incorrectly at multiple locations. The problem that the communication path is divided can be avoided.

  Here, the detection cycle of the blocked port of the master node 100 itself, the transmission cycle of the blocked status notification packet 301 of the slave node, the monitoring cycle of the ring status monitoring database unit 701, and the cycle of deleting the ring status monitoring database unit 701 Therefore, it is possible to detect an erroneous release and an erroneous setting of a blocked port at a higher speed and in real time, and to restore a normal ring network at a higher speed.

Embodiment 2. FIG.
Next, a ring network according to the second embodiment will be described. In the ring network according to the first embodiment, as shown in FIG. 9, when a failure 90 occurs (there is a path where communication between nodes is not possible), and an erroneous setting of a blocked port (blocked port 54). ) May occur, the master node 100 may release the blocked port 52 and the blocked port 53 instead of the blocked port 54. As a result, the communication path may remain disconnected.

  Therefore, in the ring network according to the second embodiment, in consideration of the possibility of the occurrence of the failure 90, the master node 100 has the blocked port 54, the blocked port 52 and the blocked port 53 of the node in which the failure 90 has occurred. For example, a blocked state notification packet 302 in which failure information is added to the blocked state notification packet 301 shown in FIG. 10 is used to release a blocked port other than the blocked port having the failure information, and the communication path is not divided. It is what I did.

  The ring network according to the second embodiment will be described below using FIG. 10, FIG. 11, FIG. 12, FIG. 13, and FIG. The blocking control packet is the same as that in the first embodiment.

  FIG. 10 shows a block state notification packet 302 for notifying the master node that a slave node having a blocked port has a blocked port and the presence or absence of a communication impossible path (hereinafter referred to as a failure). It is explanatory drawing which shows an example of a packet format.

  FIG. 11 is an example of a flowchart showing the operation of the master node. Of the processes in each step, the same processes as those in the first embodiment are denoted by the same reference numerals and the description thereof is omitted.

  FIG. 12 is an example of an internal block diagram of the master node showing the operation of the master node. Among the blocks, the same processes as those in the first embodiment are denoted by the same reference numerals and description thereof is omitted.

  FIG. 13 is an example of a flowchart showing the operation of the slave node. Of the processes in each step, the same processes as those in the first embodiment are denoted by the same reference numerals and the description thereof is omitted. FIG. 14 is an example of an internal block diagram of the slave node showing the operation of the slave node.

  Next, as shown in FIG. 9, in a ring network in which a failure 90 has occurred and a communication path is divided because a plurality of blocked ports exist on the ring due to “erroneous setting of blocked port”, The operation in which the node 100 detects the “blocking port misconfiguration” of the blocked port 54 (communication between nodes is logically possible) in which no failure has occurred will be described. Note that the processing of the recovery operation, that is, the notification of the block control packet to the slave node is the same as that of the first embodiment, and thus description thereof is omitted.

  The slave nodes 200 and 201 detect the blocked ports 52 and 53 that have blocked the port due to the occurrence of the failure 90 as the blocked ports having the failure, and the state of the blocked port (in the case of the slave node 200), as shown in FIG. : In the case of failure, port 1 blocked, slave node 201: failure, port 2 blocked) is transmitted. On the other hand, since the failure has not occurred in the blocked port 54 of the slave node 204, a blocked state notification packet 302 to which a blocked port state (no failure, port 2 is blocked) is added as shown in FIG. The master node 100 receives the blocked state notification packet 302 transmitted from the slave nodes 200, 201, and 204, and analyzes the blocked port state of each blocked state notification packet 302, thereby blocking the blocked port 54 in which no failure has occurred. Detects “blocked port misconfiguration” (communication between nodes is logically possible).

Next, with reference to FIGS. 11 and 12, an operation for detecting “blocking port misconfiguration” of the master node will be described.
The master node 100 manages the status of the blocked port existing on the ring (for example, the MAC address of the node, the presence / absence of the blocked port, the presence / absence of a port failure) in the ring status monitoring database unit 701a. For example, in FIG. 9, the master node 100 uses the ring state monitoring database unit 701a to determine whether there is a blocked port of port 1 and port 2, whether there is a failure, the MAC addresses of the slave nodes 200, 201, 204, The presence / absence of a blocked port of port 1 and port 2 and the presence / absence of a failure are managed.

  When the master node 100 detects a change in the status of the blocked port of the master node 100 itself (the presence or absence of a blocked port, the presence or absence of a failure) at the blocked port detection unit 702a (S501a), the ring status monitoring database at the control unit 700a The presence / absence of a blocked port and the presence / absence of a failure in the unit 701a are updated (S503a). Also, when the blocking state notification packet 302 from the slave node is received by the receiving unit 703 (S502), and when the blocking state notification packet detecting unit 704 detects the presence or absence of the blocked port, the control unit 700a detects the blocked state. The ring state monitoring database unit 701a is updated based on the transmission source address of the notification packet 302, the presence / absence of a blocked port of port 1 and port 2, and the presence / absence of a failure. The control unit 700a monitors the ring state monitoring database unit 701a when the ring state monitoring database unit 701a is updated, or at regular intervals or irregularly, thereby determining whether or not a failure has occurred on the ring network. The number of blocked ports other than the port where the failure has occurred (hereinafter, the number of blocked ports other than the failure) is obtained (S500).

  When a failure has occurred on the ring network and the number of blocked ports other than the failure is 0, the control unit 700a recognizes that there is no “blocked port misconfiguration” on the ring network. . If a failure has occurred on the ring network and the number of blocked ports other than the failure is 1 or more, a "blocked port misconfiguration" is detected and the blocked network other than the failure on the ring network is blocked. The node to be released from the block and its port are determined so that the number of ports becomes 0 (S504a). Note that the operation for determining the node and the port to be blocked or set when the failure has not occurred in the ring network is the same as that of the first embodiment, and the description thereof will be omitted. Further, since the generation and transmission method of the block control packet 401 is the same processing as in the first embodiment, the description thereof is omitted.

  Next, the operation of generating the slave node blocking state notification packet 302 will be described with reference to FIGS. 13 and 14. Since the operation when receiving the block control packet 401 is the same as that of the first embodiment, the description thereof is omitted.

  For example, when the slave port detects the occurrence of a blocked port of the own node or a failure (path where communication is impossible) in the blocked port detection unit 802a (S602, S605), the control unit 800a generates a blocked state notification packet. The block 806a generates a block state notification packet 302 with the block port and failure information added (S603a), and transmits the packet from the transmitter 805 (S604).

  In this way, the master node 100 monitors the presence / absence of a failure on the ring network and the number of blocked ports, thereby performing “erroneous setting of blocked ports” even on a ring network where a failure has occurred. It can be detected and recovered, and it is possible to avoid disconnection of the communication path due to “erroneous setting of blocked port”.

Embodiment 3 FIG.
Next, a ring network according to the third embodiment will be described. In the ring network according to the second embodiment, as shown in FIG. 15, when a failure 90 has occurred, an erroneously set blocked port 55 exists, and a blocked port is detected due to a failure of the slave node 203. When the 55 blockage state notification packet cannot be generated, there is a problem that the master node 100 cannot detect the erroneous setting of the blockage port 55.

  Therefore, in the ring network according to the third embodiment, in addition to the communication between the slave node having the blocked port and the failure information and the master node 100 in the second embodiment, the master node 100 Then, a monitoring packet requesting whether or not there is a blocked port and a failure is transmitted, and all the slave nodes that have received the monitoring packet transmit a monitoring response packet to the master node 100 and respond. Thereby, since the master node 100 cannot receive the monitoring response packet from the slave node 203, it is possible to provide a network that can grasp that a failure has occurred in the slave node 203.

  Next, based on FIG. 15, an operation for grasping that a failure has occurred in the slave 203 will be described using FIGS. 16, 17, and 18. Here, the packet format of the monitoring response packet 403 is the same as that of the blocking state notification packet 302 shown in FIG. 10 in the second embodiment. In addition, the process in which the master node 100 obtains a blocked port on the ring and the presence or absence of a failure and transmits the blocked control packet 401 is the same process as the flowchart of FIG.

The master node 100 generates a monitoring packet 503 by a monitoring packet generation unit 707 and detects a port at a transmission unit 705 in order to detect the presence or absence of a blocked port on the ring network at regular intervals or irregularly. 1. Transmit from port 2 to all slave nodes in two directions. Similar to the blocking state notification packets 301 and 302 and the blocking control packet 401 in the first and second embodiments, the monitoring packet 503 is relayed without being discarded at the blocking port even if there is a blocking port.

  All slave nodes receive the monitoring packet 503 by the receiving unit 803 (S607), and the monitoring packet detecting unit 804b detects port information (whether the port is blocked or not) (S608).

  Based on the port information obtained by the monitoring packet detection unit 804b, the control unit 800b performs a monitoring response in the monitoring response packet generation unit 806b in the same processing as the blocking state notification packet generation unit 806a of the second embodiment. A packet 403 is generated (S605) and transmitted by the transmission unit 805 (S604b).

  The master node 100 receives the monitoring response packet 403 transmitted by each slave node at the receiving unit 703, and detects the blocked port of the slave node and the presence / absence of a failure at the monitoring response packet detecting unit 704b. The control unit 700b updates the blocked port and the presence / absence of a failure for every slave node in the ring state monitoring database unit 701b.

  Here, since the slave node 203 is out of order, the monitor response packet 403 cannot be generated by the monitor response packet generator 806b, and the master node 100 cannot detect the monitor response packet 403 from the slave node 203. It is possible to grasp that a failure has occurred.

  Thus, the master node 100 can detect the node where the failure has occurred by transmitting the monitoring packet 503 from the master node and confirming the arrival of the monitoring response packet 403 from the slave node, and has high reliability. A network can be provided.

  Here, when no failure has occurred in the ring network, the ring state monitoring database unit 701b manages the MAC addresses and blocked ports of all nodes on the ring. For example, in FIG. 15, in the case where the slave node 203 cannot fail to generate the blocking state notification packet and cannot relay the monitoring response packet, the master node 100 receives the monitoring response packets of the slave nodes 201, 202, and 203. Since it is not possible, it can be understood that the communication path is divided between the slave nodes 201, 202, and 203.

Embodiment 4 FIG.
Next, a ring network according to the fourth embodiment will be described. In the ring networks according to the first to third embodiments, when the port blocking release or blocking setting by the blocking control packet 401 cannot be performed due to a failure or the like, the communication path may remain disconnected. Therefore, in the ring network according to the fourth embodiment, as illustrated in FIG. 19, when the slave node that has received the blocking control packet 401 fails to release the blocking port 51, it notifies the blocking release failure. The master node 100 can detect the failure of the slave node 202 by returning the blocking control response packet 402 to the master node 100. Further, when the master node 100 detects a failure of the slave node 202, it is possible to avoid disconnection of the communication path by releasing the blockage of the blockage port 50. Here, the block control response packet 402 may be notified to the master node 100 only when block release or block setting fails.

Next, based on FIG. 19, the slave node 202 that has received the block control packet 401 for notifying the block port 51 of the block port cancellation using FIG. 20, FIG. 21, FIG. 22, and FIG. A description will be given of an operation for avoiding that the communication path remains disconnected when the communication path cannot be obtained.

  In FIG. 19, the slave node 202 detects the block control packet 401 that notifies the block port 51 of the block port release by the block control packet unit 804, and performs the block port 51 block release by the control unit 800 c. After a certain period, the controller 800c uses the blocked port detector 802c to confirm the success or failure of the blocking release (S608).

  The control unit 800c generates a block control response packet 402 for notifying the success / failure of block release to the master node 100 by using the block control response packet generation unit 808 (S609), and transmits it by the transmission unit 805 (S610).

  The master node 100 that has received the blocking control response packet 402 receives it at the receiving unit 703 (S508), detects whether the blocking control response packet detecting unit 704c succeeds in releasing the blocking, and the control unit 700c monitors the ring state. The success or failure of the closing of the database unit 701c is updated (S503c). Here, the master node 100 indicates the status of the blocked port existing on the ring (for example, the MAC address of the node, the presence / absence of the blocked port, the presence / absence of the port failure, the success / failure result of the blocking control), and the ring status monitoring database unit 701c. It is managed by. The control unit 700c monitors the ring state monitoring database unit 701c when the ring state monitoring database unit 701c is updated, or at regular intervals or irregularly (S500), and the blocked ports 50 other than the port for which the blocking setting has failed. Is determined as a port for releasing the blockage (S504c).

  Thereafter, by releasing the blocking port 50 (S505), it is possible to avoid the communication path from being disconnected.

  As described above, when the port block release or block setting cannot be performed by the block control packet 401 due to a failure or the like, the master node 100 receives a notification of the block release failure from the slave node 202 and receives the block port 51 of the slave node 202. Recognize that cannot be released. Furthermore, by releasing the blockage of the blockage port 50, it is possible to avoid disconnection of the communication path and to realize a highly reliable ring network.

  If the master node 100 does not receive the block control response packet 402 for a certain period of time despite sending the block control packet 401, it is determined that a failure has occurred in the slave node that is the destination of the block control response packet 402. It doesn't matter. In that case, it is possible to avoid disconnection of the communication path by releasing the blocked port 50.

1, 2 Port 50, 51, 52, 53, 54, 55 Blocked port 90 Fault 100 Master node 200, 201, 202, 203, 204 Slave node 301, 302 Blocked state notification packet 401 Blocked control packet 402 Blocked control response packet 403 Monitoring response packet 503 Monitoring packet 700, 700a, 700b, 700c, 800, 800a, 800b, 800c Control unit 701, 701a, 701b, 701c Ring state monitoring database unit 702, 702a, 802, 802a Blocked port detection unit 703, 803 Reception Blocks 704, 704a, 704b Blocking state notification packet detector 704c Blocking control response packet detectors 705, 805 Transmitter 706 Blocking control packet generator 707 Monitor packet generator 802c Blocked port detector 04 closure control packet detecting unit 804b monitors the packet detecting unit 806,806a closed state notifying packet generating unit 806b monitors the response packet generation unit 807 relay unit 808 closed control response packet generator

Claims (4)

A plurality of nodes having a plurality of ports are forwarded, a plurality of the nodes are connected in a ring shape, at least one of the nodes is a master node that monitors a network state, and the other nodes are In a ring network that is a slave node following the instructions from the master node,
The slave node includes a blocked port detection unit that detects presence / absence of a blocked port of the slave node itself, a blocked state notification packet generation unit that generates a blocked state notification packet that notifies the master node of the presence of a blocked port, When the blocked port detection unit detects a blocked port, the blocked unit notification packet generation unit controls to generate a blocked state notification packet, and the blocked state notification generated by the blocked state notification packet generation unit. A transmission unit for transmitting packets,
The master node detects the presence / absence of a blocking state notification packet transmitted from the slave node, a blocking state notification packet detection unit for detecting the blocking state notification packet, and the presence / absence of a blocking port of the master node itself. Blocking port detection unit for each node, ring state monitoring database unit for managing the presence / absence of blocking port for each port for each node, blocking of each node in the blocking port detection unit and blocking state notification packet detection unit of the master node Update the ring state monitoring database unit based on the presence or absence of ports, and based on the information of the ring state monitoring database unit, determine the node to be blocked and its port according to the number of blocked ports existing in the ring, If the port to be blocked is the master node itself, block the corresponding port on the master node itself. When the port to be blocked is the slave node, a control unit for controlling to generate a blocking control packet for notifying the port to be controlled for blocking and the content of the blocking control of the port, and the blocking control packet A block control packet generation unit to generate, and a transmission unit to transmit the block control packet,
The slave node further includes a receiving unit that receives a blocking control packet transmitted from the master node, a blocking control packet detecting unit that detects a blocking control packet addressed to itself, and a blocking control packet that is not addressed to itself. Regardless of the presence or absence of a port, a relay unit that relays a block control packet and a control unit that performs block control of the corresponding port based on the content of the block control added to the block control packet and the port to be blocked A ring network characterized by that. In addition to detecting the presence or absence of its own blocked port in each of the master node and the slave node, including a blocked port detection unit that can detect a failure state,
In addition to notifying the master node of the presence of a blocked port, the slave node includes a blocked state notification packet generation unit that generates a blocked state notification packet that adds a failure state of the port,
In addition to managing the presence / absence of blocked ports for each node, the master node has a ring state monitoring database unit for managing failure states, and identifies and releases erroneously set blocked ports The ring network according to claim 1, wherein   The master node includes a monitoring packet generation unit and a monitoring response packet detection unit, and the slave node includes a monitoring packet detection unit and a monitoring packet response unit, and the master node confirms arrival at the slave node. The ring network according to claim 1 or 2, characterized in that   The slave node includes a block control response packet generation unit that notifies the success or failure of block control, the master node includes a block control response packet detection unit, and the ring state monitoring database unit determines whether or not block control is successful. The ring network according to any one of claims 1 to 3, wherein the ring network is managed, and a port that cannot be blocked is recognized, and the other blocked ports are blocked.