JP2011024000A - Node, and network control method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To detect loop occurrence with sufficient accuracy without needing complicated control. <P>SOLUTION: In the case that an own node operates as a root node in the topology of a communication path, a loop detection initial setting portion 13A initializes a transfer counter value of a BPDU which is transmitted to the other node other than the relevant root node; in the case that the own node operates as the other node, when the BPDU is received from one node adjacent to the own node and is transferred to another adjacent node, a loop detection processing portion 13B detects the loop occurrence in the communication path according to the comparison result of the transfer counter value obtained from the relevant BPDU and a counter reference value; in the case that the transfer counter value obtained from the BPDU is equal to or below the counter reference value, the transfer counter value of the BPDU to be transferred is incremented. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to Ethernet (registered trademark) communication technology, and more particularly, to ring-type Ethernet control technology that realizes data communication by forming a tree topology based on RSTP (Rapid Spanning Tree Protocol).

  In a supervisory control system that monitors and controls building facilities and plant facilities, communication devices having various functions such as information collection functions and control functions are connected as nodes via a communication network, and data communication is performed between these nodes. Each equipment is monitored and controlled by a central monitoring device. In such a monitoring control system, Ethernet is used as a communication network, and data communication in a higher-level application is realized by exchanging communication frames, which are data communication MAC frames, between nodes.

  In the Ethernet, when connecting a plurality of nodes, a star wiring system in which each node is connected to a hub or a switch is fundamental. Such a star wiring system is suitable for a relatively small office environment, but is not necessarily suitable for a large-scale facility such as a building facility or a plant facility. The reason is that in the star wiring method, it is necessary to connect the hub or switch to each node via individual wiring, and when nodes are installed over a wide range, the wiring connecting the nodes is complicated. This is because the burden of wiring work and maintenance increases.

  In such Ethernet, a ring type Ethernet that connects nodes in a ring shape has been proposed (see, for example, Patent Document 1). This ring type Ethernet has STP (Spanning Tree Protocol / IEEE 802.1D) function to avoid communication errors due to ring topology existing in the communication path, and RSTP (Rapid Spanning Tree Protocol / IEEE 802.1w) function improved from this It is possible to realize system redundancy by exchanging control frames, which are MAC frames for redundancy control called BPDU (Bridge Protocol Data Unit), between nodes using the redundant control function.

FIG. 8 is a configuration example of a typical ring Ethernet system. Here, five nodes N1 to N5 are ring-connected to the ring. Normally, in the RSTP redundancy control process mounted on a node, one root node, for example, the node with the smallest MAC address is selected from the nodes connected in the ring. In the example of FIG. 8, the node N1 is selected as the root node.
Thereafter, by exchanging with the control frame between the root node and other nodes, the active communication path of the tree topology is set based on, for example, the cost between the nodes or the priority order of the nodes.

Specifically, using the root node as a reference, the segment connecting the nodes with the highest cost from the root node is selected as an unnecessary route other than the active communication route, and the port of one node connected to this unnecessary route is blocked. By doing so, it is set as a backup communication path at the time of failure.
At this time, as the cost, a parameter proportional to the number of nodes located between the root node and an arbitrary node is used, and blocking is performed between the nodes having the maximum cost.

In the example of FIG. 8, the cost increases by 100 each time one node increases, so the maximum cost 200 from the root node N1 to the node N4 counterclockwise is reached, and the segment connecting the node N3 and the node N4 is It is judged as an unnecessary route. Then, blocking is performed at one of the nodes N3 and N4 connected to the end point of the unnecessary path, for example, the port of the node N4 having the larger MAC address.
For this reason, the original ring consisting of the ring topology is changed to a tree topology consisting of two branch paths from the root node N1 to the nodes N3 and N4. As a result, the occurrence of a data loop is avoided even in a network that physically forms a ring topology.

  After the tree topology is formed in this way, the RSTP redundancy control process proceeds to the health check stage. FIG. 9 is an explanatory diagram showing the health check operation. In the health check stage, health check messages using Forwarding BPDUs are periodically transmitted from the root node N1 in two directions, ie, the node N2 direction and the node N5 direction, in the failure monitoring control frame. When the nodes N2 to N5 normally receive the health check message from the root node N1, the nodes N2 to N5 transfer them to subsequent nodes. Therefore, when the health check message cannot be received normally, it can be confirmed that a failure has occurred in the segment between the adjacent nodes on the root node N1 side.

  In such a case, a reconstruction request is transmitted from the node that has confirmed the occurrence of the failure in the opposite direction to the root node. In response to the reception of the reconstruction request, the blocking node releases the blocking. As a result, the blocked backup communication path is used to reconstruct a new communication path.

JP 2005-109846 A JP 2007-036824 A JP 2007-053421 A JP 2006-217696 A

In such a ring type Ethernet system, there is a case where each node is connected in a ring shape to form a loop due to a malfunction of RSTP redundancy control processing. For example, when the function of the root node is lost and the blocking is released, the communication band between the nodes is rapidly reduced, and finally communication becomes impossible.
Conventionally, from the viewpoint of fail-safety, even when such a loop is formed, a loop detection technique for automatically detecting loop formation and avoiding communication failure has been proposed (for example, Patent Document 2- 4).

These loop detection techniques are roughly classified into a passive detection method for detecting a loop occurrence by monitoring a communication state at a node and an active detection method for separately transmitting / receiving a dedicated detection frame for detecting the loop occurrence. The
Specific examples of the passive detection method include monitoring whether the MAC address storage table of the node is frequently rewritten, monitoring whether the same frame is received multiple times, and monitoring the flow rate of the broadcast frame. There is a way to do it.
As a specific example of the active detection method, a master-slave function is added to each node and a dedicated detection frame is transmitted and received.

However, such a conventional technique has a problem that not only complicated control is required, but also the occurrence of a loop cannot be detected accurately.
In other words, in the passive detection method, when monitoring the communication status at a node, the detection criteria for the monitoring target is ambiguous, and it is necessary to adjust the detection criteria for each application. Difficult to do. Further, in the active detection method, it is necessary to mount a complicated control function on the node, which increases the processing load and cost on the node.

  The present invention is intended to solve such problems, and provides a network control technique capable of accurately detecting a loop occurrence without requiring complicated control for RSTP redundancy control in a ring Ethernet system. The purpose is that.

  In order to achieve such an object, a node according to the present invention is a node used in a ring Ethernet system that realizes data communication between nodes by connecting a plurality of nodes via a ring-shaped communication path. In addition, by exchanging BPDUs with nodes via the communication path, redundancy control for the communication path is performed based on RSTP, and the number of times the BPDU is transferred at each node during transmission of the BPDU is counted. RSTP processing unit that stores and transmits a transfer count value for transmission, and a BPDU transfer counter that transmits to other nodes other than the root node when the own node is operating as a root node in the topology of the communication path Loop detection initial setting unit that initializes the value, and the own node is operating as another node In this case, when receiving a BPDU from one adjacent node adjacent to the own node and transferring it to the other adjacent node, depending on the comparison result between the transfer counter value acquired from the BPDU and the counter reference value, And a loop detection processing unit that increments the transfer counter value of the BPDU to be transferred when the transfer counter value acquired from the BPDU is less than or equal to the counter reference value.

  At this time, when the transfer counter value exceeds the counter reference value, the loop detection processing unit may set the stop of the data communication frame transfer operation to the communication port that has received the BPDU.

  Further, the RSTP processing unit may start a connection establishment process based on RSTP with an adjacent node of the own node in response to the stop of the data communication frame transfer operation in the loop detection processing unit.

  A network control method according to the present invention is a network control method used in a ring Ethernet system that realizes data communication between nodes by connecting a plurality of nodes via a ring-shaped communication path, The RSTP processing unit exchanges BPDUs with nodes via the communication path, thereby performing redundancy control on the communication path based on RSTP, and the number of times the BPDU is transferred at each node at the time of BPDU transmission. When the RSTP processing step that stores and transmits the transfer count value for counting and the loop detection initial setting unit operates as a root node in the topology of the communication path, Loop detection that initializes the transfer counter value of the BPDU to be transmitted When the period setting step and the loop detection processing unit are operating as another node, the BPDU is received when the BPDU is received from one adjacent node adjacent to the own node and transferred to the other adjacent node. In response to the comparison result between the transfer counter value acquired from the counter and the counter reference value, the occurrence of a loop in the communication path is detected, and if the transfer counter value acquired from the BPDU is less than or equal to the counter reference value, A loop detection processing step for incrementing the transfer counter value.

  At this time, if the transfer counter value exceeds the counter reference value, the loop detection processing step may include a step of setting the stop of the data communication frame transfer operation to the communication port that has received the BPDU.

  The RSTP processing step may include a step of starting a connection establishment process based on RSTP with an adjacent node of the own node in response to the stop of the data communication frame transfer operation in the loop detection processing step.

  According to the present invention, even when an infinite loop state occurs in the ring L, the number of times that the BPDU is repeatedly transferred due to the infinite loop state can be confirmed by the transfer counter value stored in the BPDU. For this reason, it is possible to accurately detect the occurrence of a loop without requiring complicated control.

It is a block diagram which shows the principal part structure of the node concerning this Embodiment. It is a frame structure example of BPDU. It is a flowchart which shows a loop detection initial setting process. It is a flowchart which shows a loop detection process. It is explanatory drawing which shows the loop detection operation at the time of ring establishment. It is explanatory drawing which shows the loop detection operation (continuation) at the time of ring establishment. It is explanatory drawing which shows the loop detection operation (continuation) at the time of ring establishment. It is explanatory drawing which shows the loop detection operation at the time of failure occurrence. It is explanatory drawing which shows the loop detection operation at the time of failure occurrence (continuation). It is explanatory drawing which shows the loop detection operation at the time of failure occurrence (continuation). It is explanatory drawing which shows the loop detection operation at the time of failure occurrence (continuation). It is explanatory drawing which shows the loop detection operation | movement at the time of failure recovery. It is explanatory drawing which shows the loop detection operation at the time of failure recovery (continuation). It is a structural example of a typical ring type Ethernet system. It is explanatory drawing which shows a health check operation | movement.

Next, an embodiment of the present invention will be described with reference to the drawings.
[Configuration of the embodiment]
First, a node according to the present embodiment will be described with reference to FIG. FIG. 1 is a block diagram showing a main configuration of a node according to the present embodiment.

  This node N is a data communication node for constructing a ring type Ethernet system. These nodes N are connected by a ring L consisting of a ring-shaped communication path, and a tree topology is constructed by performing a redundancy control process for the ring L based on RSTP (Rapid Spanning Tree Protocol). It has a function to perform data communication.

The node N is provided with a ring connection control unit 10 and an application processing unit 20 as main functional units.
The ring connection control unit 10 has a function of controlling data communication (compliant with Ethernet) with another node N via the ring L.
The application processing unit 20 has a function of executing various host application processes for monitoring and controlling, for example, building facilities and plant facilities by Ethernet communication with other nodes N via the ring connection control unit 10.

The ring connection control unit 10 includes MAC processing units 11 and 12, an RSTP processing unit 13, and a transfer processing unit 14 as main processing units.
The MAC processing unit 11 is connected to one end L1 of the ring L via the ring connection port P1, and transmits / receives various MAC frames to / from each node N. The RSTP processing unit 13 connects the port P1 to the port P1. When blocking is set, with respect to port P1, reception / reception of a communication frame that is a MAC frame for data communication is restricted, and a function for transmitting / receiving only a control frame that is a MAC frame for redundancy control is received from an adjacent node. And a function of outputting a control frame to the RSTP processing unit 13.

  The MAC processing unit 12 is connected to the other end L2 of the ring L via the ring connection port P2, and transmits / receives various MAC frames to / from each node N. The RSTP processing unit 13 connects the port P1 to the port P1. If blocking is set, the port P2 restricts transmission / reception of communication frames, which are data communication MAC frames, and transmits / receives only control frames, which are redundancy control MAC frames, and received from adjacent nodes. And a function of outputting the control frame to the RSTP processing unit 13.

  The RSTP processing unit 13 is connected to each of the MAC processing units 11 and 12 and exchanges a control frame called BPDU (Bridge Protocol Data Unit) with another node via the ring L, thereby making a ring based on RSTP. Functions for performing various types of redundancy control processing such as network topology construction, root node determination, failure monitoring, and failure recovery for L, and transfer for counting the number of times the BPDU is transferred at each node during transmission of the BPDU A function of storing and transmitting the count value.

The RSTP processing unit 13 includes a loop detection initial setting unit 13A and a loop detection processing unit 13B as main processing units.
The loop detection initial setting unit 13A has a function of initializing a BPDU transfer counter value to be transmitted to a node other than the root node when the own node is operating as a root node in the topology of the ring L. ing.

  When the own node is operating as another node, the loop detection processing unit 13B receives the BPDU from one adjacent node adjacent to the own node and transfers the BPDU to the other adjacent node. A function for detecting the occurrence of a loop in the ring L according to the comparison result between the counter value and the counter reference value, and a transfer counter of the BPDU to be transferred when the transfer counter value acquired from the BPDU is less than or equal to the counter reference value A function for incrementing the value, and a function for setting the stop of the transfer operation of the communication frame for data communication to the communication port that has received the BPDU when the transfer counter value exceeds the counter reference value. A function to instruct the RSTP processing unit 13 to start the RSTP connection establishment process as the root node according to the operation stop setting. The has.

  FIG. 2 is an example of a frame structure of a BPDU. In this configuration example of the BPUD frame, a transfer counter value that is incremented every time the BPDU frame is transferred by a node other than the root node is added to the standard BPDU frame configuration. Therefore, for example, by comparing the reference value consisting of the total number of nodes connected to the ring type Ethernet system with this transfer counter value, the BPDU frame has made a round around the ring L, that is, a loop has occurred. Can be detected with high accuracy.

  At this time, if a forward BPDU for a health check message is used as a BPDU frame that is regularly exchanged with other nodes by the RSTP processing unit 13, for example, a transfer counter value can also be exchanged periodically. Can be constantly monitored.

  The transfer processing unit 14 transfers the MAC frame input from the MAC processing unit 11 or 12 to either the MAC processing unit 11 or 12 or the application processing unit 20 based on the destination information included in the MAC frame. And a function of transferring the MAC frame input from the application processing unit 20 to the MAC processing units 11 and 12.

[Operation of this embodiment]
Next, the operation of the node according to this exemplary embodiment will be described with reference to FIGS. FIG. 3 is a flowchart showing loop detection initial setting processing. FIG. 4 is a flowchart showing loop detection processing.

  When power supply to the node N is started, the RSTP processing unit 13 starts RSTP redundancy control, and starts connection establishment processing (handshaking) for a segment between the own node and both adjacent nodes. In the present embodiment, the various redundant control processes in the RSTP processing unit 13 are general well-known redundant control processes, and detailed description thereof is omitted here.

In response to the start of power supply to the node N, the loop detection initial setting unit 13A starts the loop detection process of FIG.
First, the loop detection initial setting unit 13A confirms the RSTP connection establishment status with the adjacent node in the RSTP processing unit 13 stored in a storage unit (not shown) in the own node (step 100), and the own node It waits until a connection is established for a segment with an adjacent node on one of the ports P1, P2 (step 100: NO).

  On the other hand, in the node N, when connection is established for a segment with an adjacent node connected to one of the ports (step 100: YES), the loop detection initial setting unit 13A stores the same as described above. Referring to the section, it is confirmed whether or not the connection is established for the segment between each adjacent node connected to the ports on both sides (step 101).

On the other hand, when the connection is established only in the segment with the adjacent node (step 101: NO), the loop detection initial setting unit 13A completes the connection establishment among the ports P1 and P2 of the own node. The transfer counter value stored in the BPDU transmitted from the existing port is initialized, for example, cleared to zero (step 102), and the process returns to step 100.
The transfer counter value stored in the BPDU transmitted from each port P1, P2 is stored in, for example, the storage unit described above, and the transfer counter value of the corresponding port is stored in the loop detection initial setting unit 13A or the loop detection processing unit 13B. Should be updated. Further, the RSTP processing unit 13 may read the transfer counter value of the corresponding port from the storage unit and store it in the BPDU to be transmitted.

On the other hand, if the connection is established for each segment between the adjacent nodes of both side ports (step 101: YES), the loop detection initial setting unit 13A refers to the storage unit described above, and the own node is the root node. Whether or not (step 103).
Here, when the own node is the root node (step 103: YES), the loop detection initial setting unit 13A initializes the transfer counter value stored in the BPDU transmitted from both ports P1 and P2 of the own node, for example, zero. (Step 104), and the process returns to step 100. If the own node is not the root node (step 103: NO), the process returns to step 100.

  In this way, by the loop detection initial setting process in the loop detection initial setting unit 13A, the transfer counter to be stored in the BPDU transmitted from the port for each port according to the connection establishment status in the segment with the adjacent node. The value is initialized.

On the other hand, when the own node is operating in a node other than the root node, the loop detection processing unit 13B executes the loop detection process of FIG. 4 in response to the reception of the BPDU from one adjacent node.
First, the loop detection processing unit 13B acquires the transfer counter value stored in the received BPDU and compares it with a counter reference value set in advance in the storage unit, thereby detecting the loop occurrence in the ring L. (Step 110)

  At this time, if the transfer counter value is equal to or less than the counter reference value (step 110: NO), it is determined that no loop has occurred (step 111), and the loop generation in the ring L is avoided by referring to the storage unit described above. It is confirmed whether the blocking port for this is set to one of the ports P1 and P2 of the own node (step 112).

  Here, when the blocking port is not set in the own node (step 112: NO), the loop detection processing unit 13B determines the transfer counter of the received original BPDU for the port opposite to the port that received the BPDU. By incrementing the value, that is, by adding the number of times of transfer in the own node once, the transfer counter value stored in the BPDU transmitted from the port is calculated and stored in the storage unit (step 113), and a series of loop detection processes Exit.

  On the other hand, when the blocking port is set in the own node (step 112: YES), the loop detection processing unit 13B stores the port on the side opposite to the port that received the BPDU in the BPDU transmitted from the port. The transfer counter value is initialized, here cleared to zero and stored in the storage unit (step 114), and a series of loop detection processing is terminated.

  In Step 110, when the transfer counter value acquired from the received BPDU exceeds the counter reference value (Step 110: YES), the loop detection processing unit 13B determines that a loop has occurred in the ring L (Step 110). 115), the stop of the transfer operation of the communication frame for data communication is set for either of the MAC processing units 11 and 12 of the port that has received the BPDU among the ports P1 and P2 of the own node (step 116). .

Thereafter, the loop detection processing unit 13B instructs the RSTP processing unit 13 to start redundancy control processing as a root node (step 117), and ends a series of loop detection processing.
As a result, the RSTP processing unit 13 starts the redundancy control processing as the root node in response to the stop of the transfer operation of the communication frame for data communication in the loop detection processing unit 13B. A connection establishment process based on RSTP is started.

[Operation example]
Next, with reference to FIGS. 5A to 7B, an example of a node state display control operation according to this exemplary embodiment will be described. 5A to 5C are explanatory diagrams illustrating a loop detection operation when a ring is established. 6A to 6D are explanatory diagrams illustrating a loop detection operation when a failure occurs. 7A and 7B are explanatory diagrams showing a loop detection operation at the time of failure recovery. It is assumed that a value “7” larger than the number of nodes is set as a counter reference value in each node.

[When ring connection is established]
First, a loop detection operation when a ring connection is established will be described with reference to FIGS. 5A to 5C. Here, the nodes N1 to N5 are connected in a ring shape by a ring L, and in the initial state, power supply to all these nodes N1 to N5 is stopped.

  In such an initial state, as shown in FIG. 5A, when the nodes N1 and N2 are first turned on, the nodes N1 to N5 other than the nodes N1 and N2 are not turned on. Connection establishment is completed only for the segment between N2. Here, for example, when the node N1 is determined as the root node, the node N1 starts to periodically transmit a failure monitoring BPDU to the node N2. At this time, the transfer count value (CNT = 0) initialized by the loop detection initial setting unit 13A is stored in the failure monitoring BPDU.

  Since the node N2 is not the root node, in response to the reception of the BPDU from the root node N1, the node N2 compares the transfer count value “0” notified by the BPDU with the count reference value “7”. Since the transfer count value is equal to or less than the count reference value and no blocking port is set in the own node, the value “1” obtained by incrementing the transfer count value notified by the BPDU is set to the side opposite to the node N1 side. This is calculated as the port transfer count value. At this time, since the connection of the opposite port is not confirmed, the received BPDU is not transferred.

Next, as shown in FIG. 5B, when the nodes N3 and N5 are powered on, connection establishment is completed in the segment between the nodes N2 and N3 and the segment between the nodes N1 and N5.
Thereby, the BPDU received from the node N1 is transferred from the port on the node N3 side of the node N2 to the node N3. At this time, the transfer count value (CNT = 1) calculated as described above is stored in this BPDU.
In addition, the node N1 starts to periodically transmit a failure monitoring BPDU to the node N5. At this time, the transfer count value (CNT = 0) initialized by the loop detection initial setting unit 13A is stored in the failure monitoring BPDU.

Thereafter, as shown in FIG. 5C, when the node N4 is powered on, connection establishment is completed in the segment between the nodes N3 and N4 and the segment between the nodes N4 and N5.
Thereby, the BPDU received from the node N2 is transferred from the node N3 to the node N4. At this time, this BPDU stores a new transfer count value (CNT = 2) obtained by incrementing the transfer count value of the received BPDU at the node N3 in the same manner as described above.
Further, the BPDU received from the node N1 is transferred from the node N5 to the node N4. At this time, this BPDU stores a new transfer count value (CNT = 1) obtained by incrementing the transfer count value of the received BPDU at the node N5 in the same manner as described above.

  Here, in the node N4, when blocking is set for the port on the node N3 side based on RSTP, the transfer count value regarding the opposite port is initialized in response to reception of the BPDU from each port. . Since the blocking port exists in the node N4, the received failure monitoring BPDU is not transferred to the adjacent port.

[When an error occurs]
Next, a loop detection operation when a failure occurs will be described with reference to FIGS. 6A to 6D. Here, it is assumed that the nodes N1 to N5 are connected in a ring shape by the ring L, and all the nodes N1 to N5 are operating normally in the initial state as shown in FIG. 5C described above.

  In such an initial state, as shown in FIG. 6A, it is assumed that a failure has occurred in the node N1 for some reason and the root node function has been lost. At this time, for example, in the node N1, when the BPDU that should be transmitted to both the nodes N2 and N5 is transmitted only to the node N2, the blocking is abnormal because the node N4 does not receive the BPDU from the node N5. It will be released. As a result, the communication frame for data communication falls into an infinite loop state where the communication frame is transferred infinitely between the nodes N1 to N5, and communication may eventually be disabled.

Accordingly, the BPDU transmitted from the node N1 has its transfer count value (CNT = 0) incremented at the node N2 (CNT = 1) and transferred to the node N3.
Similarly, also in the node N3, the BPDU transferred from the node N2 is transferred to the node N4 with its transfer count value (CNT = 1) incremented (CNT = 2).

As shown in FIG. 6B, since blocking is abnormally canceled at the node N4, the transfer count value (CNT = 2) of the BPDU from the node N3 is incremented (CNT = 3). Forwarded to
Similarly, also in the node N5, the BPDU transferred from the node N4 is transferred to the node N1 with its transfer count value (CNT = 3) incremented (CNT = 4).

Thus, the BPDU transferred from the node N4 toward the node N5 is further transferred to the nodes N1 to N5 as shown in FIG. 6C. The transfer count value is also incremented at each node.
Therefore, as shown in FIG. 6D, since the transfer count value (CNT = 8) of the BPDU transferred from the node N4 is larger than the reference count value “7” in the node N5, the node N4 side that has received the BPDU The stop of the data communication frame transfer operation is set for the port. As a result, the infinite loop state is eliminated.

[When recovering from a failure]
Next, with reference to FIGS. 7A and 7B, a loop detection operation at the time of failure recovery will be described. Here, the nodes N1 to N5 are connected in a ring shape by the ring L, and in the initial state, the transfer operation stop is set at the node N4 side port in the node N5 as in FIG. 6B described above. To do.

  In the node N5, as shown in FIG. 7A, after the stop of the data communication frame transfer operation is set for the port on the node N4 side, the redundancy control processing as the root node is started (route ID = 5). The BPDU for the connection establishment request of the segment based on RSTP is transmitted to the nodes N1 and N4 adjacent to the node N5.

  In response to the connection confirmation request BPDU from the node N5, the node N1 (route ID = 1) starts connection establishment processing (handshake) with the node N5 in the same manner as when the power is turned on. As a result, as shown in FIG. 7B, when the node N1 is determined to be the root node, for example, by comparing the route ID and the MAC address of both, the root node function is restored at the node N1, and as shown in FIG. 5C described above. Then, the network topology construction process by the root node N1 is started.

The node N4 also performs connection establishment processing (handshake) with the node N5 in response to the connection confirmation request BPDU from the node N5.
At this time, since the route ID of the node N4 is not changed, the transfer operation stoppage at the node N5 is canceled according to the redundancy control by the node N1, and the node N4 side port of the node N5 among the nodes N4 and N5 is released. A new blocking is set.
As a result, a new network topology is constructed on the ring L, and data communication between the nodes under RSTP redundancy control is resumed.

[Effects of the present embodiment]
As described above, in this embodiment, when the own node is operating as a root node in the topology of the communication path, the loop detection initial setting unit 13A transfers BPDUs to be transmitted to other nodes other than the root node. When the counter value is initialized and the loop detection processing unit 13B is operating as the other node, when the BPDU is received from one adjacent node adjacent to the own node and transferred to the other adjacent node, According to a comparison result between the transfer counter value acquired from the BPDU and the counter reference value, occurrence of a loop in the communication path is detected, and if the transfer counter value acquired from the BPDU is equal to or less than the counter reference value, the transfer is performed. The BPDU transfer counter value is incremented.

  Therefore, even when an infinite loop state occurs in the ring L, the number of times that the BPDU is repeatedly transferred due to the infinite loop state can be confirmed by the transfer counter value stored in the BPDU. For this reason, it is possible to accurately detect the occurrence of a loop without requiring complicated control.

  Further, in the present embodiment, when the transfer counter value exceeds the counter reference value, the loop detection processing unit 13B sets the stop of the data communication frame transfer operation to the communication port that has received the BPDU. As a result, the infinite loop state can be immediately resolved in response to the detection of the occurrence of the loop.

  Further, in the present embodiment, the RSTP processing unit 13 starts connection establishment processing based on RSTP with the adjacent node of the own node in response to the stop of the data communication frame transfer operation in the loop detection processing unit 13B. Thus, it is possible to automatically return from the infinite loop state to the normal state.

[Extended embodiment]
The present invention has been described above with reference to the embodiments, but the present invention is not limited to the above embodiments. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

  N, N1, N2, N3, N4, N5 ... node, 10 ... ring connection control unit, 11, 12 ... MAC processing unit, 13 ... RSTP processing unit, 13A ... loop detection initial setting unit, 13B ... loop detection processing unit, 14: Transfer processing unit, 20: Application processing unit, P1, P2 ... Port.

Claims (6)

A node used in a ring-type Ethernet system that realizes data communication between these nodes by connecting a plurality of nodes via a ring-shaped communication path,
By exchanging BPDU (Bridge Protocol Data Unit) with the node via the communication path, redundancy control for the communication path is performed based on Rapid Spanning Tree Protocol (RSTP), and at the time of transmission of the BPDU An RSTP processing unit that stores and transmits a transfer count value for counting the number of times the BPDU is transferred at each node;
When the own node is operating as a root node in the topology of the communication path, a loop detection initial setting unit that initializes the transfer counter value of the BPDU transmitted to other nodes other than the root node;
When the own node is operating as the other node, when the BPDU is received from one adjacent node adjacent to the own node and transferred to the other adjacent node, the transfer counter value and the counter reference acquired from the BPDU A loop that detects the occurrence of a loop in the communication path according to the comparison result with a value and increments the transfer counter value of the BPDU to be transferred when the transfer counter value acquired from the BPDU is equal to or less than the counter reference value A node comprising: a detection processing unit. The node of claim 1, wherein
The loop detection processing unit, when the transfer counter value exceeds the counter reference value, sets the stop of the data communication frame transfer operation to the communication port that has received the BPDU. The node according to claim 2, wherein
The RSTP processing unit starts a connection establishment process based on RSTP with an adjacent node of the own node in response to a stop of a data communication frame transfer operation in the loop detection processing unit. A network control method used in a ring Ethernet system that realizes data communication between nodes by connecting a plurality of nodes via a ring-shaped communication path,
The RSTP processing unit performs redundancy control on the communication path based on RSTP (Rapid Spanning Tree Protocol) by exchanging BPDU (Bridge Protocol Data Unit) with the node via the communication path. An RSTP processing step of storing and transmitting a transfer count value for counting the number of times the BPDU is transferred at each node at the time of transmission of the BPDU;
Loop detection in which a loop detection initial setting unit initializes the transfer counter value of the BPDU to be transmitted to a node other than the root node when the own node is operating as a root node in the topology of the communication path An initial setting step;
When the own node is operating as the other node, the loop detection processing unit receives the BPDU from one adjacent node adjacent to the own node and transfers it to the other adjacent node. The occurrence of a loop in the communication path is detected according to the comparison result between the transfer counter value and the counter reference value. If the transfer counter value acquired from the BPDU is equal to or less than the counter reference value, the transfer of the BPDU to be transferred And a loop detection processing step for incrementing a counter value. The network control method according to claim 4, wherein
The loop detection processing step includes a step of setting a stop of a data communication frame transfer operation to a communication port that has received the BPDU when the transfer counter value exceeds the counter reference value. Network control method. The network control method according to claim 5, wherein
The RSTP processing step includes a step of starting a connection establishment process based on RSTP with an adjacent node of the own node in response to the stop of the data communication frame transfer operation in the loop detection processing step. Network control method.

Images