WO2008089614A1

WO2008089614A1 - A method, system and device for ring link protection

Info

Publication number: WO2008089614A1
Application number: PCT/CN2007/002398
Authority: WO
Inventors: Haizhou Xiang; Binxuan Li; Chenggui Cai
Original assignee: Huawei Technologies Co., Ltd.
Priority date: 2007-01-17
Filing date: 2007-08-10
Publication date: 2008-07-31
Also published as: CN101001192A; CN101001192B

Abstract

A ring link protection method, which is used in Ethernet ring, is disclosed. The said ring comprises master node and more than one transit nodes. The said master node comprises primary port and secondary port. The said method is as follows: the said master node sends a HELLO message to transit nodes through said primary port and said secondary port; if said primary port and said secondary port do not receive the said HELLO message, which is forwarded by transit nodes, in an interval after sending HELLO message, the said master node unblocks said secondary port. The said method can avoid the communication loop which is created when the masternode uses the backup link.

Description

Method, system and device for ring network link protection

The present invention relates to the field of IP network technologies, and in particular, to a method, system and device for ring link protection. Background technique

The Rapid Ring Protection Protocol (RRPP) is a link layer protocol specifically applied to the Ethernet ring. It can quickly enable the backup link to recover the ring network when a link failure occurs on the Ethernet ring. The communication path between nodes.

An Ethernet ring that uses the RRPP protocol is mainly composed of multiple nodes. One of them is the master node, and the other nodes are transit nodes. The two ports on the ring of the master node are classified into the primary port and the secondary port. In order to avoid loops on the ring network, the secondary port of the primary node is usually in the closed state. The secondary port in the closed state cannot send and receive service packets, but can send and receive protocol packets.

In order to detect a link fault in the ring network, the master node periodically sends a ring protocol (HELLO) message from the master port. In the case of a complete ring, the master node receives the self-send on the secondary port within a certain period. The HELLO packet, so that the primary node considers that the ring network is in a complete state, keeps the secondary port in the closed state to ensure that there is no loop; if the primary node does not receive the HELLO packet sent by itself on the secondary port within a certain period, it is considered When the ring network is in the fault state, the primary node opens the secondary port to send and receive service packets, and sends a message to notify other transit nodes to update the forwarding table. After the transit node updates the forwarding table, the data flow is switched to the backup link.

Once the secondary port of the primary node receives the HELLO packet sent by the primary port, it considers that the ring network is restored to the complete state, immediately shuts down the secondary port, and sends a message to notify other nodes to refresh the forwarding table. After the transmission node updates the forwarding table, the data flow is restarted. Switch back to the original link.

In addition, when a link between a node on a ring network or a directly connected node fails, a down event of the node port is triggered, and the node immediately sends a fault (link_down) message to the master node. When the master node receives the faulty packet, it also considers that the ring network is in the fault state, immediately opens the secondary port, and sends a message to notify other transmitting nodes to update the forwarding table. After the transmitting node updates the forwarding table, the data stream is switched to the backup link.

If the fault is rectified, the port of the faulty node will be up. The faulty node will temporarily block the port. However, the port can also transparently transmit RRPP packets. The HELLO packet sent by the master node from the master port can penetrate the temporary blocked port. Once the secondary port of the primary node receives the HELLO packet sent by the primary port, it considers that the ring network is in a complete state, immediately shuts down the secondary port, and sends a message to notify other nodes to open the temporary blocked port and refresh the forwarding table. After the publication, the data stream is switched back to the original link.

It can be seen that in the ring network using the RRPP protocol, the master node detects the ring network link according to the fact that the secondary port does not receive the protocol packet sent by the master port or receives the fault packet sent by the transit node in a certain period. malfunction.

When a link of the transmitting node fails, for example, the forwarding chip of the transmitting node fails, causing a unidirectional path, that is, a communication disconnection, a reverse path, or a slave from the primary port of the primary node to the secondary port of the primary node. The secondary port of the primary node is disconnected from the primary port of the primary node, and the reverse path. At this time, when the primary node detects a link failure, it opens the secondary port to enable it to send and receive service packets. However, since the path from the secondary port to the primary port or from the primary port to the secondary port is originally a path, a loop occurs when the secondary port is opened, which eventually causes serious consequences of the entire network.

As shown in FIG. 1A, in a normal ring network consisting of four switches: node 1, node 2, node 3, and node 4, RRPP is enabled for each switch, and node 1 is set as a master node. As the primary port of the primary node, S is the secondary port of the primary node, and the secondary port is closed to prevent loops;

As shown in FIG. 1B, it is assumed that the node 4 is faulty, such as a forwarding chip abnormality, etc., causing the link of node 3 to node 4 to be unreachable, and the link of node 4 to node 3 is a path, and the single pass of node 4 to node 3 occurs. phenomenon. When the master node detects that the ring network is in the fault state, it immediately opens the secondary port S. At this time, the secondary port S can send and receive service packets. Therefore, a loop from the secondary port to the primary port appears in the ring network, and the whole may appear. The serious consequences of the net. Summary of the invention

The embodiment of the invention provides a method, a system and a device for protecting a ring network link, which are used to solve the problem in the prior art. In the Ethernet ring network, when the master node detects a link fault, the secondary port is opened, and the communication loop is caused. The problem.

An embodiment of the present invention provides a ring network link protection method, which is applied to an Ethernet ring network, where the ring network includes a primary node and one or more transit nodes, and the primary node includes a primary port and a secondary port, and the method includes the following Steps:

A. The master node sends a protocol packet from the primary port and the secondary port to the transit node;

B. The master port and the secondary port open the secondary port when the protocol sent by the transmitting node is not received within a certain period of time after the protocol is issued.

An embodiment of the present invention provides a ring network link protection system, where the system includes:

a master node device including a primary port and a secondary port, configured to send a protocol packet from the primary port and the secondary port to the transit node device; a certain time after the primary port and the secondary port send a protocol packet When the protocol packet sent by the transit node device is not received, the secondary port is opened; the transit node device is configured to receive the protocol packet sent by the primary port, and send the protocol packet to the secondary port; Receiving a protocol packet sent by the secondary port, and sending the protocol packet to the primary port.

The embodiment of the present invention provides a master node device, which is connected to a transit node device, where the master node device includes a primary port and a secondary port, and the device further includes:

a sending unit, configured to send a protocol packet from the primary port and the secondary port to the transit node device;

The fault detecting unit is configured to open the secondary port when the protocol packet sent by the transit node device is not received within a certain period of time after the sending unit sends the protocol packet from the primary port and the secondary port.

Compared with the prior art, in the embodiment of the present invention, the primary node of the ring network sends a protocol packet from the primary port and the secondary port to the transit node, and the primary port and the secondary port are after the protocol packet is sent. When the protocol packet sent by the transit node is not received within the fixed time, the master node opens the secondary port. Because the primary port and the secondary port do not receive the protocol on time, it means that the ring network has failed but is not single. If the fault occurs, the secondary port is opened, and no loop occurs in the ring network. When one of the primary port and the secondary port does not receive the protocol packet on time, and the other port receives the protocol packet on time, it means A single-pass fault occurs in the ring network. At this time, the secondary port cannot be opened. Otherwise, a loop will occur in the ring network, which may lead to more serious consequences. It can be seen that the present invention solves the problem that the primary node opens the secondary node once the link fault is detected in the prior art, causing a communication loop. DRAWINGS

1A is a schematic diagram of a link failure in an Ethernet ring network in the prior art;

FIG. 1B is a schematic diagram of the prior art in which the secondary port of the primary node is opened after a single-pass failure occurs in the Ethernet ring network;

2 is a flowchart of a method in an embodiment of the present invention;

3A is a schematic diagram of a link failure of an Ethernet ring network in the first embodiment of the present invention; FIG. 3B is a schematic diagram of a single-pass failure of an Ethernet ring network according to Embodiment 1 of the present invention; FIG. 4B is a schematic diagram of a case where a non-single-pass fault occurs in an Ethernet ring network; FIG. 4B is a schematic diagram of a secondary port of a primary node after a non-single-link failure occurs in the Ethernet ring network according to Embodiment 2 of the present invention;

FIG. 5A is a schematic structural diagram of a system according to an embodiment of the present invention; FIG.

5B is a schematic structural diagram of a master node device in a system according to an embodiment of the present invention;

FIG. 6 is a schematic structural diagram of a device according to an embodiment of the present invention. detailed description

A method for protecting a ring network link according to an embodiment of the present invention, in the method, the master node sends protocol packets from the primary port and the secondary port respectively, and the primary port and the secondary port send the protocol packet within a certain period of time. When the protocol packet is not received, the master node will open the secondary port. Otherwise, to avoid loops, keep the secondary port in the closed state. The embodiment of the present invention provides a ring network link protection method, which is applied to an Ethernet ring network using the RRPP protocol, where the ring network is mainly composed of multiple nodes, one of which is a master node, and the other nodes are transit nodes, and the master node The two ports on the ring are divided into a primary port and a secondary port. To avoid loops on the ring network, the secondary port of the primary node is usually in the closed state. The secondary port in the closed state cannot receive or send service packets, but can send and receive protocol packets.

Referring to FIG. 2, the specific implementation process of the method provided by the embodiment of the present invention is as follows:

Step 201: The master node issues an agreement from the primary port and the secondary port respectively.

On the ring network that uses the RRPP protocol, the master node sends protocol packets from the primary port and the secondary port. The master node sends protocol packets from the primary port and the secondary port at regular intervals. When the protocol packet is sent, the timer is started for the primary port and the secondary port. The protocol packet sent from the primary port will reach the secondary port of the primary node through the transit nodes in the ring network when the ring network is complete. After the primary port sends the protocol packet, if it is within a certain period of time, If the secondary port does not receive the protocol packet, the primary node considers that the link is faulty. Similarly, when the ring network is complete, the protocol packets sent from the secondary port will eventually reach the primary port of the primary node through the various transit nodes in the ring network. If the primary port does not receive the protocol within a certain period of time, In the case of the text, the master node also believes that the link has failed.

Step 202: The primary node opens the secondary port when the primary port and the secondary port do not receive the protocol packet within a certain period of time after the protocol is sent.

This step includes the following three situations:

When the primary port does not receive the protocol packet within a certain period of time after the protocol packet is sent, the master node considers that the link is faulty, but does not immediately open the secondary port. The protocol packet is not received within a certain period of time. If yes, it means that there is no one-way path in the ring network. The link fault detected by the master node is not a single-pass fault. At this time, the master node opens the secondary port. It can send and receive service packets, and send messages to other transmitting nodes to update the forwarding table. After the transmitting node updates the forwarding table, the data flow is switched to the backup link; otherwise, it means that there is a direction from the primary port to the secondary port in the ring network. In a one-way path, the link fault detected by the master node is a one-way fault. In order to avoid loops, the secondary port is not opened, but the secondary port is kept closed. In the closed state, only the timer is cleared;

When the secondary port does not receive the protocol packet within a certain period of time after the protocol packet is sent, the master node considers that the link is faulty but does not immediately open the secondary port. Instead, it first determines whether the primary port is also transmitting the protocol packet. The protocol packet is not received within a certain period of time. If yes, it means that there is no one-way path in the ring network. The link fault detected by the master node is not a single-pass fault. At this time, the master node opens the secondary port. The device can send and receive service packets, and send messages to other transmitting nodes to update the forwarding table. After the transmission node updates the forwarding table, the data flow is switched to the backup link; otherwise, the ring network has a direction from the secondary port to the primary port. In a one-way path, the link fault detected by the master node is a one-way fault. In order to avoid loops, the secondary port is not opened, but the secondary port is kept closed, and only the timer is cleared.

When receiving the link fault packet sent by the transit node, the master node does not immediately open the secondary port. Instead, it first determines whether the primary port and the secondary port have not received the protocol packet within a certain period of time after the protocol packet is sent. If yes, it means that there is no unidirectional path in the ring network. The link fault detected by the master node is not a single-pass fault. At this time, the master node opens the secondary port to enable it to send and receive service packets and send packets at the same time. Notifying other transit nodes to update the forwarding table, and the data stream is switched to the backup link after the transit node updates the forwarding table; otherwise, it means that there is a one-way path from the primary port to the secondary port or from the secondary port to the primary port in the ring network. The link fault detected by the master node is a single-pass fault. In order to avoid loops, the secondary port is not opened, but the secondary port is kept closed, and only the timer is cleared.

In this step, the certain time can be set and changed. For example, a timer can be set for the primary port and the secondary port respectively in the primary node, and then the primary port can be returned when the timer of the primary port expires. The protocol packet is not received, and the primary port does not receive the protocol packet within a certain period of time after the protocol packet is sent. Similarly, the secondary port has not received the protocol packet when the timer of the secondary port expires. It is determined that the secondary port does not receive the protocol packet within a certain period of time after the protocol packet is sent; of course, the primary port and the secondary port of the master node can share a timer, in this case, when the timer expires The master port has not received the protocol packet, and it is determined that the master port does not receive the protocol within a certain period of time after the protocol is issued. When the timer expires, the secondary port has not received the protocol packet to determine that the secondary port has not received the protocol packet within a certain period of time after the protocol packet is sent.

Step 203: After the secondary port is opened, when the primary port or the secondary port of the primary node receives the protocol sent by the transmission node, the secondary port is closed;

Here, after the secondary port is opened, when the primary port and the secondary port receive the protocol packet, the ring network is restored to the complete state. To prevent loops, the secondary port needs to be shut down, and a packet is sent to notify other transmitting nodes to refresh the forwarding table. After the transit node updates the forwarding table, the data stream is switched back to the original link. When one port of the primary port or the secondary port receives the protocol packet, only one port receives the protocol packet. At this time, the ring network does not return to the complete state, but a communication loop occurs, in order to avoid loops. More serious consequences, such as broadcast storms, also require the secondary port to be closed and send a message to inform other transmitting nodes to refresh the forwarding table.

The method provided by the present invention is specifically described below by using specific embodiments:

Embodiment 1:

Referring to FIG. 3A, in the embodiment, a switch is composed of four switches: node 1, node 2, node 3, and node 4, and RRPP is enabled on each switch, and node 1 is set as the master. Node, P is the primary port of the master node, S is the secondary port of the master node, and the secondary port is closed to prevent loops;

As shown in Fig. 3B, the support node 4 fails, causing the link from node 3 to node 4 to be unreachable, and the link from node 4 to node 3 is the path. When the master node detects that the ring network is in the fault state, in order to distinguish whether the fault is a single-pass fault, it is first determined whether the master port and the secondary port have not received the protocol packet within a certain period of time. Since the link from node 4 to node 3 is the path, the unidirectional path of node 4-> node 3-> node 2-> node 1 appears in the ring network. At this time, the master port can receive the protocol packet sent by the slave port. Therefore, the link fault detected by the master node is a single-pass fault, and in order to avoid the communication loop in the direction of node 4->node 3->node 2->node 1->node 4, the secondary port is not opened, and Keep the secondary port off and only clear the timer.

Embodiment 2:

Referring to FIG. 3A, in this embodiment, four nodes are still connected by node 1, node 2, node 3, and node 4. In the normal ring network, each switch is enabled with RRPP, and node 1 is set as the master node, P is the master port of the master node, and S is the secondary port of the master node. The secondary port is closed. To prevent loops;

As shown in Fig. 4A, it is assumed that node 3 and node 4 fail, causing the link from node 3 to node 4 and node 4 to node 3 to fail. When the master node detects that the ring network is in the fault state, in order to distinguish whether the fault is a single-pass fault, it is first determined whether the master port and the secondary port have not received the protocol packet within a certain period of time. Since the link from node 3 to node 4 and node 4 to node 3 are unreachable, there is no unidirectional path in the ring network. At this time, neither the primary port nor the secondary port can receive the protocol packet, so the link detected by the master node The fault is not a single-pass fault, so the master node opens its secondary port to enable it to send and receive service packets, and sends a message to inform other transport nodes to update the forwarding table. After the transport node updates the forwarding table, the data flow is switched to the backup link. . At this time, a communication path for service message transmission is established between the four nodes, as shown in Fig. 4B.

Referring to FIG. 5A, an embodiment of the present invention further provides a ring network link protection system, where the system includes: a master node device 501 including a primary port P and a secondary port S, respectively, for respectively from the primary port and the secondary The port sends a protocol packet to the transit node device; when the primary port and the secondary port do not receive the protocol packet sent by the transit node device within a certain period of time after the protocol packet is sent, the secondary port is opened; The node device 502 is configured to receive the protocol packet sent by the primary port, and send the protocol packet to the secondary port; receive the protocol packet sent by the secondary port, and send the protocol packet to the primary port. Text.

Referring to FIG. 5B, the master node device 501 further includes a sending unit 5011 and a fault detecting unit 5012, where the sending unit 5011 is configured to send a protocol packet from the primary port and the secondary port to the transmitting node device, respectively; 5012. The method is: when the sending unit does not receive the protocol packet sent by the transit node device within a certain period of time after the sending unit sends the protocol packet from the primary port and the secondary port, the secondary port is opened.

The sending unit 5011 is configured to send a protocol packet from the primary port and the secondary port to the transmitting node device at regular intervals.

The fault detecting unit 5012 includes a first processing unit 50121, a second processing unit 50122, and a The third processing unit 50123 is configured to: when the primary port does not receive the protocol packet sent by the transit node device within a certain period of time after the primary port sends the protocol packet, determine whether the secondary port is The protocol message sent by the transit node device is not received within a certain period of time after the protocol packet is sent, and if so, the secondary port is opened; the second processing unit 50122 is configured to send the protocol packet after the secondary port sends the protocol packet. If the protocol packet sent by the transit node device is not received within a certain period of time, it is determined whether the master port does not receive the protocol packet sent by the transit node device within a certain period of time after the protocol packet is sent, and if yes, Opening the secondary port; the third processing unit 50123 is configured to: when receiving the fault packet sent by the transit node device, determine whether the primary port and the secondary port are not within a certain period of time after sending the protocol packet Receiving a protocol packet sent by the transit node device, and if yes, opening the secondary port.

The master node device 501 further includes a fault recovery unit 5013, configured to close the secondary port when the primary port or/and the secondary port receives a protocol packet sent by the transit node device.

Referring to FIG. 6, an embodiment of the present invention further provides a master node device, where the device includes a primary port P and a secondary port S, which can be applied to a ring link protection system, and is connected to a transmission node device, and the device further includes sending The unit 601 and the fault detecting unit 602, wherein the sending unit 601 is configured to send a protocol packet from the primary port and the secondary port to the transmitting node device, respectively; and the fault detecting unit 602 is configured to send the sending unit from the primary When the port and the secondary port do not receive the protocol packet sent by the transit node device within a certain period of time after the protocol packet is sent, the secondary port is opened.

The sending unit 601 is configured to send a protocol packet from the primary port and the secondary port to the transmitting node device at regular intervals.

The fault detection unit 602 includes a first processing unit, a second processing unit 6022, and a third processing unit 6023. The first processing unit 6021 is configured to not receive the protocol packet after the primary port sends the protocol packet. When the protocol packet sent by the node device is transmitted, it is determined whether the secondary port does not receive the protocol packet sent by the transit node device within a certain period of time after the protocol packet is sent, and if yes, the secondary port is opened; The second processing unit 6022 is configured to: when the protocol port sent by the transit node device is not received within a certain period of time after the slave port sends the protocol packet, determine whether the master port is certain after the protocol packet is sent. The protocol sent by the transit node device was not received within the time limit. And the third processing unit 6023 is configured to determine, when receiving the fault message sent by the transit node device, whether the primary port and the secondary port are in an agreement. The protocol packet sent by the transit node device is not received within a certain period of time after the packet, and if so, the secondary port is opened.

The device further includes a fault recovery unit 603, configured to close the secondary port when the primary port or/and the secondary port receives a protocol packet sent by the transmitting node device.

In the present invention, the master node may not receive the protocol packet within a certain period of time after the protocol packet is sent by the primary port or the secondary port, or may detect the link fault by receiving the fault packet sent by the transit node. A link failure from the primary port to the secondary port can be detected, and a link failure from the secondary port to the primary port can also be detected. However, after the master node detects a link failure, the backup link is enabled only when it is determined that the link failure is not a single-pass failure, and the secondary port is opened, because opening the secondary port in the single-pass failure causes communication on the ring network. Loops, while communication loops can cause serious adverse consequences such as broadcast storms. Here, if the primary port and the secondary port do not receive the protocol packet sent by the transit node within a certain period of time after the protocol packet is sent, the link fault is not a single-pass fault, and the secondary port can be opened; if the primary port And one of the ports on the secondary port does not receive the protocol packet sent by the transit node within a certain period of time after the protocol packet is sent, and the other port receives the protocol sent by the transit node within a certain period of time after the protocol packet is sent. The packet indicates that the link fault is a single-pass fault and does not open the secondary port.

It can be seen that, with the present invention, it can be ensured that in the case of a single-pass failure of the Ethernet ring network link, the primary node does not enable the backup link and opens the secondary port, thereby ensuring that communication does not occur in the ring network. The loop, in turn, avoids serious adverse consequences such as broadcast storms caused by communication loops.

It should be noted that when the present invention is sold or used as a stand-alone software product, the software corresponding to the invention can be stored in a computer readable storage medium, and the computer can implement the functions that the software can perform.

It is apparent that those skilled in the art can make various modifications and variations to the invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention are claimed in the present invention The invention is also intended to cover such modifications and variations within the scope of the invention.

Claims

Rights request

A ring network link protection method, which is applied to an Ethernet ring network, the ring network includes a master node and one or more transit nodes, and the master node includes a primary port and a secondary port, wherein the method includes The following steps:

B. The primary port and the secondary port open the secondary port when the protocol sent by the transmitting node is not received within a certain period of time after the protocol is sent.

2. The method according to claim 1, wherein in step A, the master node periodically issues a protocol temple message from the primary port and the secondary port.

3. The method of claim 1, wherein step B comprises:

When the primary port does not receive the protocol packet sent by the transit node within a certain period of time after the protocol packet is sent, the master node determines whether the secondary port does not receive the protocol packet within a certain period of time after the protocol packet is sent. The protocol packet sent by the transit node, and if so, the secondary port is opened.

4. The method of claim 1, wherein step B comprises:

When the slave port does not receive the protocol packet sent by the transit node within a certain period of time after the slave port sends the protocol packet, the master node determines whether the master port does not receive the protocol packet within a certain period of time after the protocol packet is sent. The protocol packet sent by the transit node, and if so, the secondary port is opened.

5. The method of claim 1, wherein step B comprises:

When the master node receives the fault packet sent by the transit node, it determines whether the master port and the slave port do not receive the protocol packet sent by the transit node within a certain period of time after the protocol packet is sent, if Yes, the secondary port is opened.

The method of claim 1, wherein after the primary node opens the secondary port in step B, the method further includes:

The master node closes the secondary port when the primary port or/and the secondary port receives a protocol packet sent by the transit node.

7. A system for ring link protection, characterized in that the system comprises: a master node device including a primary port and a secondary port, configured to send a protocol packet from the primary port and the secondary port to the transit node device; a certain time after the primary port and the secondary port send a protocol packet When the protocol packet sent by the transit node device is not received, the secondary port is opened; the transit node device is configured to receive the protocol packet sent by the primary port, and send the protocol packet to the secondary port; Receiving a protocol packet sent by the secondary port, and sending the protocol packet to the primary port.

The system according to claim 7, wherein the master node device further includes: a sending unit, configured to send a protocol packet from the primary port and the secondary port to the transmitting node device;

The system according to claim 8, wherein the sending unit is configured to: send protocol packets from the primary port and the secondary port to the transmitting node device at regular intervals.

The system according to claim 8, wherein the fault detecting unit comprises: a first processing unit, configured to not receive the transmitting node device within a certain time after the primary port sends the protocol packet When the protocol packet is received, it is determined whether the secondary port does not receive the protocol packet sent by the transit node device within a certain period of time after the protocol packet is sent, and if yes, the secondary port is opened.

The system according to claim 8, wherein the fault detecting unit comprises: a second processing unit, configured to not receive the transmitting node device within a certain time after the slave port sends the protocol packet When the protocol packet is received, it is determined whether the primary port does not receive the protocol packet sent by the transit node device within a certain period of time after the protocol packet is sent, and if yes, the secondary port is opened.

The system according to claim 8, wherein the fault detecting unit comprises: a third processing unit, configured to: when receiving the fault message sent by the transmitting node device, determine the location Whether the primary port and the secondary port do not receive the protocol packet sent by the transit node device within a certain period of time after the protocol packet is sent, and if yes, the secondary port is opened.

The system of claim 8, wherein the master node device further comprises:

The fault recovery unit is configured to close the secondary port when the primary port or/and the secondary port receives the protocol text sent by the transmission node device.

A master node device, which is connected to a transit node device, where the master node device includes a primary port and a secondary port, and the device further includes:

The device according to claim 14, wherein the sending unit is configured to: send a protocol message from the primary port and the secondary port to a transit node device at regular time intervals.

The device according to claim 14, wherein the fault detecting unit comprises: a first processing unit, configured to not receive the transmitting node device within a certain time after the primary port sends the protocol packet When the protocol packet is received, it is determined whether the secondary port does not receive the protocol packet sent by the transit node device within a certain period of time after the protocol packet is sent, and if yes, the secondary port is opened.

The device according to claim 14, wherein the fault detecting unit comprises: a second processing unit, configured to not receive the transmitting node device within a certain time after the slave port sends the protocol packet When the protocol packet is received, it is determined whether the primary port does not receive the protocol packet sent by the transit node device within a certain period of time after the protocol packet is sent, and if yes, the secondary port is opened.

The device according to claim 14, wherein the fault detecting unit comprises: a third processing unit, configured to: when receiving the fault message sent by the transit node device, determine whether the primary port and the secondary port have not received the transmission node device within a certain time after sending the protocol packet Protocol message, if yes, open the secondary port.

The device of claim 14, further comprising: a failure recovery unit, configured to: when the primary port or/and the secondary port receives a protocol text sent by the transmission node device, Close the secondary port.