WO2011035674A1

WO2011035674A1 - Method for accelerating data switching during ethernet ring network failure

Info

Publication number: WO2011035674A1
Application number: PCT/CN2010/076324
Authority: WO
Inventors: 何三波
Original assignee: 迈普通信技术股份有限公司
Priority date: 2009-09-27
Filing date: 2010-08-25
Publication date: 2011-03-31
Also published as: CN101651596B; CN101651596A

Abstract

A method for accelerating data switching during an Ethernet ring network failure includes: when the physical link between a master node and a transmission node fails, the transmission node generates a LINK_DOWN message and a RING_DOWN_FLUSH_FDB message, sends the LINK_DOWN message to the master node, sends the RING_DOWN_FLUSH_FDB message to the other transmission nodes, and refreshes its own MAC list items; the master node generates a RING_DOWN_FLUSH_FDB message, sends the RING_DOWN_FLUSH_FDB message to all of the transmission nodes, and refreshes its own MAC list items; the other transmission nodes refresh MAC list items after receiving the first RING_DOWN_FLUSH_FDB message; when the physical link between two transmission nodes fails, these two transmission nodes generate LINK_DOWN messages and RING_DOWN_FLUSH_FDB messages respectively, send the LINK_DOWN messages to the master node, send the RING_DOWN_FLUSH_FDB messages to the other transmission nodes, and refresh their own MAC list items; the master node sets the state of the spanning tree of its secondary port to be a forwarding state after receiving the first LINK_DOWN message, generates a RING_DOWN_FLUSH_FDB message which is sent respectively from the primary port and secondary port to all of the transmission nodes, and refreshes its own MAC list items; the other transmission nodes refresh MAC list items after receiving the first RING_DOWN_FLUSH_FDB message.

Description

Method for data acceleration switching when Ethernet ring network fails

Technical field

The present invention relates to Ethernet technology and Ethernet Automatic Protection (EAPS, Ethernet Automatic Protection) Switching) protocol, especially relates to the method of data acceleration switching when the Ethernet ring network fails.

Background of the invention

In the traditional EAPS technology, a typical EAPS consists of one master node and several transit nodes. Figure 1 is in the prior art Schematic diagram of EAPS. As shown in Figure 1, EAPS includes a primary node and multiple transit nodes, each of which includes a primary port P and a secondary port S, when EAPS Normally, the master node sets the spanning tree state of its secondary port S to the blocking state. In the current standard, there are two methods for judging the fault of the Ethernet ring network. The first method includes: the master node periodically slaves its own primary port P sends a fault detection (HELLO) message. If the fault detection packet is received from its own secondary port S, it indicates that EAPS is normal. If the fault detection packet is not received within the predetermined time, it is considered EAPS has failed. The second method involves monitoring the status of the link directly connected to it by each of the transmitting nodes. If a transit node discovers its EAPS port (P port or S If the physical signal of the port is lost, the ring network fault (LINK_DOWN) packet is sent to the master node. After receiving the LINK_DOWN packet of the transit node, the master node considers EAPS. A failure has occurred. In addition, a method commonly used by related equipment manufacturers to determine an Ethernet ring failure includes: if the primary node detects its primary port P and secondary port S If the physical signal on any of the connected links is lost, the primary node considers EAPS to be faulty.

When the master node detects EAPS through any of the above three methods When a fault occurs, the data switching operation that is usually performed is as follows: the master node sets the spanning tree status of its secondary port S to the forwarding state, and generates a ring refresh (RING_DOWN_FLUSH_FDB) message. And sending the RING_DOWN_FLUSH_FDB message from the primary port P and the secondary port S to all the transmitting nodes respectively, and each transmitting node receives the After the RING_DOWN_FLUSH_FDB packet is refreshed, its Media Access Control (MAC) entry is refreshed to implement data switching.

The prior art also discloses an Ethernet ring network protection switching method. In the disclosed technical solution, a transmission node pair LINK_DOWN is added. After the LINK_DOWN packet is received, the transmitting node needs to deliver the LINK_DOWN packet to the control plane, and then refresh the MAC address. Entry. This method can reduce the failure of the transmitting node to refresh the MAC in time because the RING_DOWN_FLUSH_FDB packet sent by the primary node is lost. The entry causes the data to be interrupted, but this method increases the processing of the LINK_DOWN packet by the transmitting node, thus increasing the complexity of the protocol.

In view of the above shortcomings of the prior art, especially for the data switching performed by the transmission node after the failure occurs, the complexity of the protocol is increased, and it is necessary to propose a The method of data acceleration switching in the event of EAPS failure to solve all problems in the prior art.

Summary of the invention

In view of this, embodiments of the present invention provide an EAPS The method of data acceleration switching in the case of failure can improve the data switching speed when the ring network fails without increasing the burden of the transmission node and the complexity of the protocol.

Embodiments of the present invention also provide an EAPS The system with data acceleration switching in the event of failure can improve the data switching speed when the ring network fails without increasing the burden of the transmission node and the complexity of the protocol.

To solve the technical problem, the technical solution of the embodiment of the present invention is implemented as follows:

A method for data acceleration switching when an Ethernet ring network fails, comprising:

When a physical line between the primary node and the transit node in the Ethernet ring network fails, the transmitting node that detects the fault generates a ring network fault ( LINK_DOWN) the message and the ring network refresh (RING_DOWN_FLUSH_FDB) message, and send the LINK_DOWN message to the master node, The RING_DOWN_FLUSH_FDB message is sent to other transit nodes and refreshes its own media access control (MAC) entry; the master node generates RING_DOWN_FLUSH_FDB message, the RING_DOWN_FLUSH_FDB message is sent to all transit nodes in the Ethernet ring network, and its MAC address is refreshed. After the first RING_DOWN_FLUSH_FDB packet is received, the other transit nodes refresh their respective MAC entries.

When the physical line between two transit nodes in the Ethernet ring network fails, the two transport nodes respectively generate LINK_DOWN Send the LINK_DOWN packet to the master node and send the RING_DOWN_FLUSH_FDB message to the RING_DOWN_FLUSH_FDB packet. The packet is sent to the other transit node and the MAC address entry is refreshed. After receiving the first LINK_DOWN packet, the master node sets the spanning tree status of the secondary port to the forwardable state. The RING_DOWN_FLUSH_FDB packet sends the RING_DOWN_FLUSH_FDB packet to all transit nodes from the primary port and the secondary port, and refreshes its own MAC address. Table entry; after receiving the first RING_DOWN_FLUSH_FDB packet, the other transit node refreshes its MAC entry.

A system for data acceleration switching in an Ethernet ring network fault, comprising a primary node and a plurality of transmission nodes located in an Ethernet ring network;

Any one of the plurality of transmission nodes is configured to generate a ring network failure when a physical line between the primary node and the transmission node fails ( LINK_DOWN) the message and the ring refresh (RING_DOWN_FLUSH_FDB) message, and send the LINK_DOWN message to the master node, The RING_DOWN_FLUSH_FDB message is sent to other transit nodes except the transit node, and its own media access control (MAC) entry is refreshed;

The master node is configured to generate a RING_DOWN_FLUSH_FDB message, and the Sending a RING_DOWN_FLUSH_FDB message to the multiple transit nodes and refreshing its own MAC entry;

The other transit nodes except the transit node are used to refresh the respective RING_DOWN_FLUSH_FDB messages after receiving the first RING_DOWN_FLUSH_FDB message. MAC entry; or

Any two adjacent transmission nodes of the plurality of transmission nodes are configured to generate separately when a physical line between the two transmission nodes fails The LINK_DOWN packet and the RING_DOWN_FLUSH_FDB packet are sent to the primary node by the LINK_DOWN packet. The RING_DOWN_FLUSH_FDB message is sent to other transit nodes except itself, and its own MAC entry is refreshed.

After receiving the first LINK_DOWN packet, the master node sets the spanning tree status of the secondary port to a forwardable state and generates RING_DOWN_FLUSH_FDB message, sending RING_DOWN_FLUSH_FDB messages from the primary port and the secondary port to the multiple transit nodes, respectively, and refreshing its own MAC Entry

Other transit nodes than any two adjacent transit nodes are used to receive the first RING_DOWN_FLUSH_FDB After the packet is received, its MAC entry is refreshed.

It can be seen from the above technical solutions that the embodiments provided by the present invention are provided. The method and system for data acceleration switching in the case of an Ethernet ring network failure can improve the data switching speed when the ring network is faulty and reduce the ring network of the master node without substantially increasing the burden of the transmission node and the complexity of the protocol. Refreshed packet loss caused by The probability that a MAC entry cannot be refreshed in time.

BRIEF DESCRIPTION OF THE DRAWINGS

1 is a schematic structural view of an EAPS in the prior art;

2 is a schematic diagram of an Ethernet ring network failure in the first embodiment;

3 is a schematic diagram of an Ethernet ring network failure in Embodiment 2;

FIG. 4 is a schematic diagram of an Ethernet ring network failure in the third embodiment.

Mode for carrying out the invention

In order to make the objects, technical solutions and advantages of the present invention more clear, the following The embodiments of the present invention are described in detail.

The method for data acceleration switching in the Ethernet ring network failure provided by the embodiment of the present invention includes: when the Ethernet ring network When the physical line between the primary node and the transit node fails, the transmitting node that detects the fault generates a ring network fault (LINK_DOWN) message and a ring network refresh ( RING_DOWN_FLUSH_FDB ) message, and send the LINK_DOWN message to the master node, RING_DOWN_FLUSH_FDB The message is sent to other transit nodes, and the respective media access control (MAC) entries are refreshed; the master node generates a RING_DOWN_FLUSH_FDB message, which will be described. The RING_DOWN_FLUSH_FDB message is sent to all transit nodes in the Ethernet ring network and its own MAC entry is refreshed; the other transit nodes receive the first one. After the RING_DOWN_FLUSH_FDB packet, the respective MAC entries are refreshed; when the physical line between the two transit nodes in the Ethernet ring network fails, the two transit nodes are respectively generated. LINK_DOWN message and RING_DOWN_FLUSH_FDB message, send the LINK_DOWN message to the master node, RING_DOWN_FLUSH_FDB message is sent to other transit nodes and its own MAC entry is refreshed; the primary node receives the first LINK_DOWN After the packet is sent, the spanning tree status of the secondary port is set to the forwardable state, and the RING_DOWN_FLUSH_FDB packet is generated, which will be sent from the primary port and the secondary port respectively. RING_DOWN_FLUSH_FDB message is sent to all transit nodes and its own MAC entry is refreshed; other transit nodes receive the first RING_DOWN_FLUSH_FDB After the packet is received, its MAC entry is refreshed.

Embodiment 1

2 is a schematic diagram of an Ethernet ring network failure in the first embodiment. as shown in picture 2 The Ethernet ring network in this embodiment includes a primary node and five transmission nodes, wherein the transmission nodes 1 , 2 , 3 , 4 , and 5 are sequentially connected, and the primary node P of the primary node and the transmission node 1 Connected, the secondary port S is connected to the transmitting node 5. Only five transmission nodes are shown in Figure 2, but those skilled in the art will appreciate that the number of transmission nodes can be set as needed during networking.

In the present embodiment, when the physical line between the transmission node 2 and the transmission node 3 fails, the transmission node 2 and the transmission node 3 It can detect that the physical line is faulty, and the transit node 2 and the transit node 3 perform the following operations respectively: generating a LINK_DOWN message and a RING_DOWN_FLUSH_FDB The packet is sent to the master node, and the RING_DOWN_FLUSH_FDB packet is sent to other transit nodes, and its MAC address entry is refreshed.

The primary node receives the first LINK_DOWN After the packet is modified, the spanning tree status of the secondary port is changed to the forwarding state. The primary port and the secondary port are sent to the transit nodes 1 , 2 , 3 , 4 , and 5 on the Ethernet ring network . RING_DOWN_FLUSH_FDB packet and refresh the MAC entry.

After receiving the LINK_DOWN packet sent to the master node, the transit node does not process the packet and forwards it directly to the next node.

Because the primary node receives only the first LINK_DOWN packet (the first LINK_DOWN packet may be the transit node 2) Issued and forwarded by transport node 1, or forwarded by transport node 3 and forwarded by transport nodes 4 and 5, and processed accordingly, RING_DOWN_FLUSH_FDB message, and transport node 1 receives the RING_DOWN_FLUSH_FDB message directly from transport node 2, so Under normal circumstances, the transit node 1 should receive the RING_DOWN_FLUSH_FDB message sent by the transit node 2 first, and then receive the reply from the master node. RING_DOWN_FLUSH_FDB message; likewise, transport nodes 4 and 5 receive RING_DOWN_FLUSH_FDB directly from transport node 3. The message, therefore, the transmitting nodes 4 and 5 should receive the RING_DOWN_FLUSH_FDB message of the transmitting node 3 first, and then receive the master node's RING_DOWN_FLUSH_FDB message. Therefore, a transit node that is not connected to the failed physical line is always received first from the transmitting node connected to the failed physical line. RING_DOWN_FLUSH_FDB message, after receiving RING_DOWN_FLUSH_FDB from the master node Message. In this way, the transport node that is not connected to the failed physical line refreshes the MAC entry earlier. Through practical application, it can be found that the transmission node that is not connected to the failed physical line refreshes the MAC. The action of the entry and the action of the master node changing the spanning tree state of the secondary port to the forwarding state occur almost simultaneously, so that the method provided in this embodiment can reduce the data interruption time and accelerate the data switching in the Ethernet ring network failure.

This embodiment is a transmission node 2 and a transmission node 3 The physical circuit failure between the two is described as an example, but those skilled in the art can understand that the processing principle when the physical circuit between any two transmission nodes fails is the same as the principle of the above method, and the same can be realized. Fast data switching when the ring network fails.

Example 2

FIG. 3 is a schematic diagram of an Ethernet ring network failure in the second embodiment. As shown in Figure 3 The Ethernet ring network in this embodiment includes a primary node and five transmission nodes, wherein the transmission nodes 1 , 2 , 3 , 4 , and 5 are sequentially connected, and the primary node P of the primary node and the transmission node 1 Connected, the secondary port S is connected to the transmitting node 5. Only five transmission nodes are shown in Figure 3, but those skilled in the art will appreciate that the number of transmission nodes can be set as needed during networking.

In this embodiment, when the physical line between the primary node and the transit node 1 fails, the transport node 1 If the physical line is faulty, the LINK_DOWN packet and the RING_DOWN_FLUSH_FDB packet are generated, and then the LINK_DOWN packet is sent to the master node. The RING_DOWN_FLUSH_FDB message is sent to transport nodes 2, 3, 4, and 5 and its own MAC entry is refreshed. When the primary node and the transit node 1 When the physical line fails, the primary node can also detect that the physical line has failed, but for the primary node, whether it detects its own connection to its primary port P If the physical line fails, it is confirmed that the physical line connected to its primary port P is faulty after receiving the LINK_DOWN message sent by the transmitting node 1. The primary node must first set its secondary port S. The spanning tree state is a forwardable state, and a RING_DOWN_FLUSH_FDB message is generated, which is then sent from the secondary port S to the transmitting nodes 5, 4, 3, 2, and 1 And refresh the MAC table entry. The transmission node 1 connected to the failed physical line also sends RING_DOWN_FLUSH to other transmission nodes after detecting the failure of the physical line. The message, so some of the transmitting nodes will first receive the RING_DOWN_FLUSH message sent by the transmitting node 1, so that these transmitting nodes refresh the MAC. The action of the entry is earlier than the RING_DOWN_FLUSH message sent by the primary node in the prior art. The action of the entry, and the method provided in this embodiment can also reduce the MAC of some transit nodes caused by the loss of the RING_DOWN_FLUSH_FDB message sent by the primary node. The probability that the entry cannot be refreshed in time.

Example three

FIG. 4 is a schematic diagram of an Ethernet ring network failure in the third embodiment. As shown in Figure 4 The Ethernet ring network in this embodiment includes a primary node and five transmission nodes, wherein the transmission nodes 1 , 2 , 3 , 4 , and 5 are sequentially connected, and the primary node P of the primary node and the transmission node 1 Connected, the secondary port S is connected to the transmitting node 5. Only five transmission nodes are shown in FIG. 4, but those skilled in the art can understand that the number of transmission nodes can be set as needed when networking.

In the present embodiment, when the physical line between the primary node and the transmission node 5 fails, the transmission node 5 If the physical line is faulty, the LINK_DOWN packet and the RING_DOWN_FLUSH_FDB packet are generated, and then the LINK_DOWN packet is sent to the master node. The RING_DOWN_FLUSH_FDB message is sent to transport nodes 4, 3, 2, and 1 and its own MAC entry is refreshed. When the primary node and the transit node 5 When the physical line fails, the master node can also detect that the physical line has failed, but for the master node, whether it detects its connection with its secondary port S If the physical line fails, it is confirmed that the physical line connected to its secondary port S is faulty after receiving the LINK_DOWN message sent by the transmitting node 5, and the primary node must generate RING_DOWN_FLUSH_FDB message, sending RING_DOWN_FLUSH_FDB message from primary port P to transport node 1, 2, 3, 4 and 5, and refresh the MAC table entry. The transmitting node 5 connected to the failed physical line also sends RING_DOWN_FLUSH to other transmitting nodes after detecting a physical line failure. Packet, so some of the transmitting nodes will first receive the RING_DOWN_FLUSH message sent by the transmitting node 5, so that these transmitting nodes refresh the MAC. The action of the entry is earlier than the action of refreshing the MAC entry after receiving the RING_DOWN_FLUSH message sent by the master node in the prior art, and using the method provided in this embodiment. The probability that the MAC entries of some transit nodes cannot be refreshed in time due to the loss of the RING_DOWN_FLUSH_FDB packet sent by the primary node can be reduced.

The embodiment of the invention further provides a system for data acceleration switching when an Ethernet ring network fails. The system is included in the Ethernet ring network A master node and multiple transport nodes, which can be connected as shown in Figures 1, 2, 3, and 4.

In this embodiment, any one of the multiple transmission nodes is configured to generate when a physical line between the primary node and the transit node fails. The LINK_DOWN packet and the RING_DOWN_FLUSH_FDB packet are sent to the primary node by the LINK_DOWN packet. The RING_DOWN_FLUSH_FDB message is sent to other transit nodes except the transit node, and its own MAC entry is refreshed;

The master node is configured to detect, when a physical line between the primary node and the transit node fails, or send the transmission node that detects the fault After the LINK_DOWN packet is generated, the RING_DOWN_FLUSH_FDB packet is generated, and the generated RING_DOWN_FLUSH_FDB Send the packet to multiple transit nodes and refresh its own MAC entry.

In this embodiment, after receiving the LINK_DOWN packet, the transit node does not process, but directly LINK_DOWN The message is forwarded to the next node.

In this embodiment, when the physical line between the primary port and the transit node of the primary node fails, the primary node sets the spanning tree state of its secondary port to a forwardable state.

The above is only the preferred embodiment of the present invention and is not intended to limit the scope of the present invention. Any modifications, equivalents, improvements, etc. made within the spirit and scope of the present invention are intended to be included within the scope of the present invention.

Claims

1 . A method for data acceleration switching when an Ethernet ring network fails, wherein the method comprises:

When the physical line between the primary node and the transit node in the Ethernet ring network fails, the transmitting node that detects the fault generates a ring network fault (LINK_DOWN) The packet and the ring network refresh ( RING_DOWN_FLUSH_FDB ) message, and send the LINK_DOWN message to the master node, The RING_DOWN_FLUSH_FDB message is sent to other transit nodes and refreshes its own media access control (MAC) entry; the master node generates The RING_DOWN_FLUSH_FDB packet sends the RING_DOWN_FLUSH_FDB packet to all transit nodes in the Ethernet ring network and refreshes its own MAC address. After the first RING_DOWN_FLUSH_FDB packet is received, the other transit nodes refresh their respective MAC entries.

When the physical line between two transit nodes in the Ethernet ring network fails, the two transport nodes respectively generate a LINK_DOWN message and RING_DOWN_FLUSH_FDB packet, send LINK_DOWN packet to the master node, and RING_DOWN_FLUSH_FDB The packet is sent to the other transit node and the MAC address entry is refreshed. After receiving the first LINK_DOWN packet, the master node sets the spanning tree status of the secondary port to the forwardable state. The RING_DOWN_FLUSH_FDB packet sends RING_DOWN_FLUSH_FDB packets to all transit nodes from the primary port and the secondary port, and refreshes its own MAC address. Table entry; after receiving the first RING_DOWN_FLUSH_FDB packet, the other transit node refreshes its MAC entry.
2. The method for data acceleration switching in an Ethernet ring network fault according to claim 1, wherein the LINK_DOWN is received. The transit node of the packet directly forwards the LINK_DOWN packet to the next node.
3, according to claim 1 The method for data acceleration switching in an Ethernet ring network fault is characterized in that a physical line between a primary node and a transit node in an Ethernet ring network fails, and the physical line is located at a primary port of the primary node and Between the transit nodes, the master node sets the spanning tree state of its secondary port to a forwardable state.
The method for data acceleration switching in an Ethernet ring network fault according to claim 1 or 3, wherein the master node generates The RING_DOWN_FLUSH_FDB message includes: the primary node detects that a physical line between the primary node and the transit node fails, or receives the transmission node that detects the fault. After the LINK_DOWN packet, the RING_DOWN_FLUSH_FDB packet is generated.
5, A system for data acceleration switching in an Ethernet ring network fault, characterized in that the system comprises: a master node and a plurality of transit nodes located in an Ethernet ring network;

Any one of the plurality of transmission nodes is configured to generate a ring network failure (LINK_DOWN) when a physical line between the primary node and the transmission node fails The packet and the ring refresh ( RING_DOWN_FLUSH_FDB ) message, and the LINK_DOWN packet is sent to the primary node, and the The RING_DOWN_FLUSH_FDB message is sent to other transit nodes except the transit node, and its own media access control (MAC) entry is refreshed;

The master node is configured to generate a RING_DOWN_FLUSH_FDB message, and the RING_DOWN_FLUSH_FDB Sending the packet to the multiple transit nodes and refreshing its own MAC entry;

The other transmitting nodes except the transmitting node are used to refresh the respective MAC after receiving the first RING_DOWN_FLUSH_FDB message. Entry; or

Any two adjacent transmission nodes of the plurality of transmission nodes are configured to generate LINK_DOWN when a physical line between the two transmission nodes fails Send the LINK_DOWN packet to the master node and send the RING_DOWN_FLUSH_FDB message to the RING_DOWN_FLUSH_FDB packet. Send the packet to other transit nodes except itself and refresh its own MAC entry.

After receiving the first LINK_DOWN packet, the master node sets the spanning tree status of the secondary port to a forwardable state and generates RING_DOWN_FLUSH_FDB message, sending RING_DOWN_FLUSH_FDB messages from the primary port and the secondary port to the multiple transit nodes, respectively, and refreshing its own MAC Entry

The other transmitting nodes except the two adjacent transmitting nodes are used to refresh the MAC after receiving the first RING_DOWN_FLUSH_FDB message. Entry.
The system for data acceleration switching in an Ethernet ring network fault according to claim 5, wherein the transmission node is configured to receive LINK_DOWN The LINK_DOWN packet is directly forwarded to the next node after the packet.
7. According to claim 5 The system for data acceleration switching in the case of an Ethernet ring network fault is characterized in that the master node is configured to fail when a physical line between a master node and a transit node in an Ethernet ring network is located, and the physical line is located at the The spanning tree state of the secondary port of the primary node and the transit node is set to a forwardable state.
8. According to claim 5 or 7 The system for data acceleration switching in the case of an Ethernet ring network fault, wherein the master node is configured to: when detecting that a physical line between the master node and the transit node fails, or receive the detected fault Transmitted by the transmitting node After the LINK_DOWN packet, the RING_DOWN_FLUSH_FDB packet is generated.