WO2010031295A1

WO2010031295A1 - Control method for ethernet failure recovery

Info

Publication number: WO2010031295A1
Application number: PCT/CN2009/073452
Authority: WO
Inventors: 宋晓丽
Original assignee: 中兴通讯股份有限公司
Priority date: 2008-09-22
Filing date: 2009-08-24
Publication date: 2010-03-25
Also published as: CN101686167A

Abstract

A control method for Ethernet ring failure recovery is provided. The method comprises that: when the adjacent transmission nodes of the failed link detect that the failed link recovers to be normal, they migrate the ports corresponding to themselves to a pre-forwarding state in order to make themselves allow to forward the control protocol messages of the Ethernet ring and forbid to forward the data messages of the protection virtual local area network VLAN. The failed link recovery switching could be completed in reason by using the invention and the instant loop which is made at the time of the link recovery could be avoided.

Description

Control method for Ethernet fault recovery

Technical field

The present invention relates to an Ethernet fault recovery technology, and more particularly to a control method for Ethernet ring network fault recovery.

Background technique

At present, the Ethernet loop structure is widely used in both the metropolitan area network and the enterprise network to improve the reliability of the network. One ring physically corresponds to a ring-connected Ethernet topology, which is a set of Ethernet connected to each other. Switch collection. The Ethernet loop protection protocol defines the roles of each node on the loop. The master node is the main decision and control node on the ring. Each ring must have one and only one master node. The ring has other than the master node. Nodes are all transit nodes. Each node uses a virtual local area network (VLAN) to transmit protocol packets and cooperate with each other to protect the loops. The protected packets are transmitted through the protected VLANs. The data packets and control packets are isolated from each other through different VLANs. When a link is disconnected, the loop protection mechanism is enabled to resume communication between nodes on the ring network.

Each of the transit nodes on the loop monitors the link status of the port in real time. When a link in the ring network fails, the nodes at both ends of the link detect the port DOWN. The corresponding control protocol packet should be sent from another faultless port. The master node is notified of the change. After receiving the packet, the master node opens the slave port data forwarding function, and notifies each of the transit nodes on the ring to refresh their respective media access control (MAC, Media Access Control) addresses. In order to prevent the failure notification packet from the faulty node from being sent to the master node, the master node has an alternate detection mechanism. If the HELLO packet is not received within the specified time, the link network is considered to have a link fault. The HELLO packet is A health test is a slogan.

Under this mechanism, when a link on the ring network fails, the slave port of the master node is opened, and after the fault is recovered, if there is no corresponding protection mechanism, the slave port of the master node and the faulty link are restored. The ports are connected at the same time to form a broadcast loop for data packets. Summary of the invention

The technical problem to be solved by the present invention is to provide a control method for the recovery of the fault of the Ethernet ring network, which solves the problem that the slave port of the master node and the port after the fault link is restored are connected to each other to form a data packet after the fault recovery of a link of the ring network is completed. The problem with the broadcast loop.

In order to solve the above problem, the present invention provides a control method for Ethernet ring network failure recovery, the method comprising:

A. When the adjacent transit node of the faulty link detects that the faulty link is normal, the port corresponding to the faulty link is migrated to the pre-forward state, so that the switch can forward the Ethernet ring control protocol packet and prohibit the forwarding protection. Data packet of the virtual LAN VLAN.

Further, after the step of the transit node migrating itself to the pre-forward state, the method further includes:

The transmitting node receives the packet, and determines that the received packet is prohibited from being forwarded when the VLAN TAG of the received packet is inconsistent with the control protocol VLAN identifier preset by the loop protection protocol in the pre-forwarding state.

The transmitting node receives the packet, and when the received packet is determined to be a packet that does not identify the U TAG, the forwarded message is prohibited from being forwarded.

Preferably, after the step of the transport node migrating itself to the pre-forward state, the method further includes:

The receiving node receives the packet, and determines that the VLAN TAG of the received packet is consistent with the preset control protocol VLAN identifier in the pre-forwarding state, and forwards the packet according to the destination address of the received packet. Or send the message to the CPU for processing.

further,

The packet received by the transit node is sent by the primary node in the ring network on the primary port; and after the step B, the method further includes the step C:

C. After the master node receives the packet forwarded by the transport node from its own slave port, Determining the link failure recovery, blocking the data packet forwarding function of the slave port protection VLAN that is opened when the link is faulty, and sending a control protocol packet from the master port to notify all the transmission node links on the Ethernet ring to recover. And notify all the transmitting nodes on the Ethernet ring to refresh their respective media access control MAC address tables.

Preferably, the method further comprises:

After receiving the control protocol packet sent by the master node, the transit node migrates the port corresponding to the forwarding node to the forwarding state.

Compared with the prior art, the application of the present invention enables the node to set the port corresponding to the faulty link to the pre-forward state when detecting the link failure recovery, that is, blocking the protection of the VLAN data packet, but the control protocol packet can be Therefore, the HELLO packet used for ring health detection can be returned to the slave port of the master node through the port in the pre-forward state. The master node re-blocks the slave port after receiving the HELLO packet from the slave port, and uses the corresponding control protocol. The transmitting node on the ring informs the ring node to refresh the MAC address. After receiving the packet, the neighboring link of the faulty link ends the pre-forwarding state, migrates to the forwarding state, and updates its own MAC address. It can be seen that the method can reasonably complete the faulty link recovery switching and avoid the instantaneous loop caused by the link recovery. BRIEF abstract

Figure 1 is a schematic diagram of the fault-free state of the Ethernet ring network;

Figure 2 is a schematic diagram of the fault condition of the Ethernet ring link;

Figure 3a and Figure 3b are schematic diagrams of two phases of fault recovery of the ring network link;

4 is a flow chart of a control method for Ethernet ring network failure recovery in the present invention.

Preferred embodiment of the invention

The specific implementation of the present invention will be further described in detail below with reference to the accompanying drawings.

Figure 1 is a schematic diagram of the fault-free state of the Ethernet ring network, where S2 is configured as the primary node of the Ethernet loop, and the remaining nodes are all transit nodes. Under normal circumstances, in order to prevent data packets from forming a broadcast loop, the master node should block its data packet forwarding function from the protected VLAN of port 22, so that the data in the protected VLAN cannot pass through the port. For regular network health checks, the primary node will cycle from it The master port 21 sends a HELLO message, passes through each transport node in turn, and finally returns from the slave port of the master node to the master node. In Figure 1, the primary node S2 sends a HELLO packet periodically. Because the loop is faultless, the HELLO packet passes through the loop to the slave port of the master node and returns to the master node.

Figure 2 is a schematic diagram of the link failure of the Ethernet ring link. The transit nodes on the Ethernet ring monitor the link status of the port in real time. When a link in the ring network fails, the nodes at both ends of the link detect the port DOWN. The corresponding control protocol message should be sent from another faultless port immediately to notify the master node of this change. In Figure 2, the link between the nodes S5 and S6 is faulty. After detecting that the corresponding link of the port is faulty, the nodes S5 and S6 block the corresponding port of the faulty link and send corresponding protocol packets from the other port to notify the master node. After the link is faulty, the master node considers that the loop is faulty, and then opens the data packet forwarding function of the protected VLAN of the slave port of the master node, and notifies each transit node on the ring to refresh its MAC address table.

Figure 3 is a schematic diagram of the recovery of the Ethernet ring link failure, where

Figure 3a shows that when a neighboring node detects a link failure recovery, it immediately blocks the data packet forwarding function of the protected VLAN of the slave port of the restored master node and enters the pre-forward state. Here, the pre-forwarding state is a state in which the control protocol packet can be forwarded and the data packet of the protected VLAN cannot be forwarded, so that the HELLO packet sent by the master node can be returned to the slave port of the master node through the loop; The status of a data packet that can forward control protocol packets and forward protected VLANs.

Figure 3b shows the master node receiving the HELLO packet sent by the slave port. The link is recovered, the data packet forwarding function of the protection VLAN of the slave port of the master node is re-blocked, and the control protocol packet is sent from the master port. The transit node on the notification ring recovers and refreshes its respective MAC address. The port in the pre-forward state receives the packet and sets its pre-forward state to the forwarding state.

The method of the present invention can set the adjacent port of the faulty link to the pre-forward state when the corresponding link of the faulty port is detected on the transit node, and the faulty link is obtained after receiving the control protocol packet sent by the master node. The state of the adjacent port moves to the forwarding state. The control protocol message sent by the master node is used to notify the transit node of the address ring to refresh and the corresponding port state switch.

The main idea of the present invention is: when the transit node detects the fault recovery of the adjacent link, the corresponding packet detection mechanism is started, and the VLAN TAG is detected for all the received packets, for example, the VLAN TAG and the loop of the packet are determined. The control protocol VLAN ID set by the protection protocol is inconsistent or the report If the packet is a U TAG (unidentified) packet, the packet is forbidden to be forwarded. If the VLAN TAG of the packet is consistent with the VLAN ID of the control protocol, the packet is forwarded according to the destination address of the packet. Or send the message to the CPU for processing.

As shown in FIG. 4, an Ethernet ring network fault recovery control method includes the following steps: Step 41: The process starts, the Ethernet ring network transmission node detects the adjacent link fault recovery; Step 42: The transit node starts packet detection. The mechanism is: detecting the received packet, determining whether the received packet is a TAG packet, if yes, executing step 43, otherwise, performing step 44;

Step 43: Detecting the VLAN TAG of the received packet, determining whether the VLAN identifier is the same as the control protocol VLAN identifier preset by the ring network protection protocol, and if yes, executing step 45, otherwise performing step 44;

Step 44: Disabling the forwarding of the packet on the corresponding port, ending the detection process of the packet, and performing the detection of the next new text;

If the packet detection function is that the packet is a U TAG packet or the packet VLAN ID is different from the preset control protocol VLAN ID, the packet is forbidden to be forwarded on the corresponding port.

Step 45: Forward the packet according to the destination address of the packet or send the packet to the CPU for processing; at this time, the VLAN identifier of the received packet is consistent with the preset control protocol VLAN identifier, according to the packet The destination address forwards the packet or sends the packet to the CPU for processing (the destination address is local).

Step 46: The master node sends a HELLO protocol packet on the primary port. If the HELLO protocol packet passes the faulty link neighboring node to the master node at the slave port, the master node determines that the link fault is recovered, and re-blocks the slave port. Protect the data forwarding function of the VLAN, and send the control protocol packet from the primary port to notify the transit node of the ring to recover and refresh the media access control MAC address table of each transport node;

The HELLO packet belongs to one type of protocol control packet, so its VLAN TAG must also be the same as the set control VLAN flag.

Step 47: The transit node receives the control protocol packet sent by the master node, and learns that the master node is blocked. The data packet forwarding function of the slave port disables the packet detection mechanism that implements the pre-forward state, ends the pre-forwarding state, and the port of the transit node enters the forwarding state, and the process ends.

The above is only a preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope of the present invention. All should be covered by the scope of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Industrial applicability

The invention can enable the node to set the port corresponding to the faulty link to the pre-forward state when detecting the link failure recovery, that is, blocking the protection of the VLAN data packet, but the control protocol packet can pass, so that the ring network health detection is used. The HELLO packet can be returned to the slave port of the master node through the port in the pre-forward state. After receiving the HELLO packet from the slave port, the master node re-blocks the slave port and uses the corresponding control protocol packet to notify the ring node of the ring to refresh the MAC address. Address, the adjacent link of the faulty link receives the packet, ends the pre-forwarding state, migrates to the forwarding state, and updates its own MAC address. It can be seen that the method can reasonably complete the faulty link recovery switching and avoid the instantaneous loop caused by the link recovery, so it has strong industrial applicability.

Claims

Claim

1. A control method for recovering an Ethernet ring network, the method comprising:

2. The control method according to claim 1, wherein after the step of the transport node migrating itself to the pre-forward state, the method further comprises:

3. The control method according to claim 1, wherein after the step of the transport node migrating itself to the pre-forward state, the method further comprises:

4. The control method according to claim 1, wherein after the step of the transport node migrating itself to the pre-forward state, the method further comprises:

5. The control method according to claim 4, wherein

After the receiving, by the master node, the packet forwarded by the transit node, the master node determines that the link fault is recovered, and the data packet of the slave port protection VLAN that is opened when the link fault is blocked is forwarded. And transmitting a control protocol message from the primary port to notify all transport node links on the Ethernet ring to recover and notify all transport nodes on the Ethernet ring to refresh their respective media access control MAC address tables.

6. The control method according to claim 5, wherein after the step C, the method further comprises: