WO2009146580A1

WO2009146580A1 - Method for updating address fowarding table

Info

Publication number: WO2009146580A1
Application number: PCT/CN2008/002135
Authority: WO
Inventors: 吴少勇
Original assignee: 中兴通讯股份有限公司
Priority date: 2008-06-03
Filing date: 2008-12-30
Publication date: 2009-12-10
Also published as: CN101599892A

Abstract

A method for updating address forwarding table is provided, which includes that nodes in a second domain update address forwarding tables when link switch happens in a first domain of Ethernet, wherein the first domain is a double-return protection domain, a secondary ring protection domain or a primary ring protection domain and the second domain is the domain out of the first domain, or the partial domain out of the first domain. The method for updating address forwarding table used in Ethernet protection improves the switch quality and speed in Ethernet protection by preventing that the address forwarding table of each node is updated inadequately when a link of part domain in the Ethernet protection is switched, which mistakes the address forwarding tables after switch and leads the instability of convergence.

Description

Address forwarding table update method

Technical field

The present invention relates to the field of data communications, and in particular, to an address forwarding table update method. Background technique

In the practical application of Ethernet, various Ethernet protection technologies are widely used to implement redundant backup between the primary path (or link) and the alternate path. When the primary path and the alternate path are in a normal state, the protected data forwarding function of the alternate path is blocked, and the protection data between the networks is transmitted on the primary path to prevent the protected data from being repeatedly received and forming a broadcast storm; when the primary path occurs In case of failure, the protection data forwarding function of the alternate path is enabled, and the protection data between the transmission networks on the alternate path is switched to improve the anti-fault capability of the Ethernet, and meet the high real-time of the convergence time less than 50 milliseconds (ms) at the time of handover. Sexual requirements.

As shown in Figure 1, the dual-homed connection is used (that is, a network area has two links respectively connected to the outer nodes of the area, and one link backs up another link). In the topology diagram, nodes S1 and S2 are Ethernet switches, nodes R1 and R2 are Ethernet switches or other devices with Ethernet switching functions, network A is connected to nodes R1 and R2, and network B is connected to node S1. C is connected to node S2; node R1 is connected to node S1, node S1 is connected to node S2, and node S2 is connected to node R2. Network A can communicate with Network B or with Network C. Network A and network B can communicate through two physical paths, that is, network B<-> node Sl<-> node Rl<-> network A, network B<-> node Sl<-> node S2<-> Node 1 2<—>Network A; Similarly, there are two physical path networks between Network A and Network C. C<-> Node S2<-> Node 81<-> Node 1 1<一> Network Eight, Network C< —> Node S2<—> Node R2<—> Network A.

When the Ethernet dual-homed connection protection is applied, the system defines the Ethernet dual-homed connection protection domain. The Ethernet dual-homed connection protection domain can also be called the Ethernet dual-homed connection fast protection switching group. Group or abbreviated as a protection group. As shown in FIG. 2, the nodes included in the protection domain have S1 and S2, and the included link has a link between the node S1 and the node R1, and a link between the node S1 and the node S2. The link between node S2 and node R2, which also specifies the scope of protection data. Wai. The protection domain defines a protection link, a node to which the protection link belongs (also called a protection link control node), and a first port and a second port of the protection link control node, where the second port of the protection link control node is defined. The direct link is the protection link. As shown in Figure 2, node S2 is the protection link control node, port 22 of node S2 is the first port, port 21 is the second port, and the protection link control node S2 is the first. The two-port 21 direct link (the link connecting the node S2 and the node R2) is a protection link. In the normal state of the network shown in FIG. 2, the control node S2 blocks the protection data forwarding function of the second port 21 to prevent The protection data is transmitted on the protection link to prevent the protection data from being repeatedly forwarded and the broadcast storm is formed. The communication path between the network B and the network A is: network B<-> node S 1 <-> node 1 1 <one> network, The communication paths of networks C and A are: network C<-> node S2<-> node Sl<-> node Rl<-> network A.

When the link in the Ethernet dual-homed protection domain fails, if the faulty link is not a protection link, the node in the protection domain connected to the protection link sends a link fault to other nodes connected to it in the protection domain. After the control node receives the fault information, the protection data forwarding function of the second port is opened, and each node in the protection domain updates the address forwarding table, and the data between the networks is transmitted according to the new path. As shown in Figure 3, the link between nodes S1 and R1 is faulty. Node S1 blocks the protection data forwarding function of port 11 and notifies other nodes that the link has failed. After receiving the fault notification, control node S2 opens the port. The protection data forwarding function of 21, in addition, each node in the domain updates the address forwarding table, and the new communication paths of network B and A are: network B<-> node Sl<-> node S2<-> node 1 2<-> network A The communication paths of network C and A are: network C<-> node S2<-> node R2<-> network A.

When the faulty link in the Ethernet dual-homed protection domain is restored, the recovery switch is performed, and the network transmission returns to the normal transmission path. Because the path changes, the node also needs to update the address forwarding table.

When the Ethernet dual-homed connection protection is switched, the address forwarding table needs to be updated. The node in the Ethernet dual-homing protection domain can update the address forwarding table according to the handover event. After the node in the domain updates the address forwarding table, the node in the domain can broadcast the protection data, so that the communication data reaches the destination device. As shown in Figure 3, after the Ethernet dual-homing connection performs the protection switching of the fault, both nodes S1 and S2 update the address forwarding table, and there is no outbound port information about the network A on the node, so the protection data sent by the network B to the network A is After the broadcast of node S1 and node S2, it can finally arrive. Network A, at the same time, nodes S1, S2 and R2 also learn the address of network B, but since nodes R1 and R2 outside the protection domain do not update the address forwarding table, the address forwarding table is also the entry before the path switching, that is, the wrong address entry. Before the nodes R1 and R2 learn the new egress port that arrives at the network B path, the protection data sent by the network A to the network B is still forwarded according to the wrong address forwarding table, that is, from the node R1 to the egress port of the node S1, the data is sent. In fact, due to the link failure, network B cannot be reached. Only after waiting for the correct address of the network B to exit the port, the path sent by network A to network B is correct, and network B can be reached. Therefore, network A sends The path to network B is switched, depending on the traffic sent by network B to network A. The switching time is often unstable, sometimes even more than 50ms. The communication between network A and B exists when the Ethernet dual-homed connection is restored. A similar problem, when the Ethernet dual-homed connection is used for fault protection or recovery protection switching, the communication between network A and network C also has similar problems.

In addition, for example, the protection technology of the Ethernet loop, as shown in FIG. 4 is a multi-ring topology in the loop protection, and the nodes S1, S2, S3, and S4 form a primary ring (the primary ring is also called a ring), and the node S3 S5, S6, and S4 form a slave ring (also known as a subring from the ring). In the normal case, a link is blocked and the data forwarding function is blocked. The link is called the ring protection link of the primary ring, that is, the link between node S1 and node S4 in Figure 4; In the case where there is also a link blocked, the link is called a ring protection link of the slave ring, that is, the link between node S5 and node S6 in FIG. The primary ring and the secondary ring block the ring protection link under normal conditions, preventing the protection data from forming a broadcast storm. When the ring node detects that there is a link fault in the slave protection domain, if the fault link is not the ring protection link of the slave ring, the slave ring opens the protection data forwarding function of the ring protection link belonging to the slave ring. And updating the address forwarding table from each node on the ring, the communication of protection data between the nodes can be transmitted on the ring protection link of the slave ring, and a new transmission path is formed between the nodes, as shown in FIG. 5, node S5 The link between the node and the node S3 is faulty. After detecting the failure of the link corresponding to the port 51, the node S5 blocks the protection data forwarding function of the port 51, and sends the fault state protocol frame outward. The node S6 receives the fault state protocol. After the frame, the protected data forwarding function of the port 62 is opened, and both the node S5 and the node S6 update the address forwarding table, and a new transmission path is formed between the network A and the network B, that is, the network B<-> node S5<-> node S6 <—> Node S4<—> Node 81<—> Network Eight. After the node S5 and the node S6 update the address forwarding table, the protection data sent by the network B to the network A is broadcasted by the node S5 and the node S6, and finally reaches the network A, and each node learns the address of the network B. However, since the node S1 does not update the address forwarding table, the address forwarding table is also an entry before the path switching, that is, The wrong address entry, before the node SI learns the new egress port that arrives at the network B path, the protection data sent by the network A to the network B is still forwarded according to the wrong address forwarding table, that is, from the egress port 12 of the node S1. The data is actually unable to reach network B due to link failure and port blocking. Only after waiting for the correct address of the network B to exit the port, the path sent by network A to network B is correct, and network B can be reached. The path switching sent by network A to network B depends on the traffic sent by network B to network A. The switching time is often unstable, sometimes even more than 50 ms. Between the network A and B during the protection switching of the Ethernet loop for recovery. There are similar problems with communication.

Therefore, in Ethernet protection, there are often path switches in local areas (such as the above-mentioned Ethernet dual-homed protection domain or slave ring protection domain in Ethernet loop protection), but only the address translation of nodes in the local area is updated. Published, when a network connected to another area (such as the above-mentioned Ethernet dual-homed domain through the dual-homed network node area or the primary ring protection domain in the Ethernet ring) sends communication data to the network connected to the local area, The switching path is unstable due to an error in the address forwarding table of other areas, and the switching time is unstable.

Summary of the invention

The technical problem to be solved by the present invention is to provide an address forwarding table update method. When local area path switching in Ethernet protection, the address forwarding table of each node in the network is updated, and the switching quality of the Ethernet protection is improved. speed.

In order to distinguish between the local area and the non-local area of the address forwarding table when the local area path is switched, the local area is defined as the first area, and the link forwarding table needs to be updated when the local area link is switched. The non-first area is defined as the second area. The primary ring can also be called a ring, and the primary ring protection area can also be called a ring protection field. The secondary ring can also be called a sub-ring. The secondary ring protection area can also be called a sub-ring protection domain.

In order to solve the above technical problem, the present invention provides a method for updating a node address forwarding table, including: when a link switch occurs in a first area of an Ethernet, a node updates an address forwarding table in a second area; A dual-homing protection area, a slave protection area or a primary ring protection area in the network, the second area being an area outside the first area, or a partial area outside the first area. Further, when the link switching occurs in the first area, the node in the first area sends an update address forwarding table message to notify other nodes to perform an update operation of the address forwarding table, and the node in the second area receives the update. After the address forwarding table message, its address forwarding table is updated according to this message.

Further, when the system is initialized, each port of all nodes of the second area is configured in a virtual local area network where the update address forwarding table message is located.

Further, the update address forwarding table message is a status notification frame; the method specifically includes: when a link failure occurs between nodes in the first area, a node located in the first area and connected to the faulty link is Adding state information indicating that the current state of the first area is a fault state to a status notification frame, and periodically transmitting or transmitting the status notification frame to the node connected to the faultless link for the first time; When the faulty link between the nodes in the area returns to normal, the node in the first area adds state information indicating that the current state of the first area is the recovery state, and periodically sends the status information to the node connected thereto. The status notification frame is sent only once or sent several times; after receiving the status notification frame, the node in the first area forwards the status notification frame to a node connected thereto; the node of the second area receives After the status notification frame, it is determined that the status information included in the recorded previous status notification frame and the status information in the current status notification frame are The same, if the same, to keep its address forwarding table unchanged; If not, update their address forwarding table.

Further, when a link failure occurs between nodes in the first area, a node that is located in the first area and is connected to the faulty link, or a faulty link between the nodes in the first area returns to normal, is located at the When the node in an area sends a status notification frame, when it is determined that there is a path switch in the first area, the update indication information is added in the status notification frame; when it is determined that there is no path switch in the first area, the status notification frame is Adding no update indication information; after receiving the status notification frame, the node of the second area determines that the status information included in the recorded previous status notification frame is different from the status information in the current status notification frame, Whether the update notification information or the update indication information is included in the status notification frame, and if yes, when the update indication information is included, the address forwarding table is updated, and when the indication includes no update indication information, the address forwarding table is kept unchanged; If the status notification frame does not include the update indication information or does not update the indication information, the address forwarding is updated. table.

Further, the update address forwarding table message is a status notification frame; the method specifically includes: when a link failure occurs between nodes in the first area or when the faulty link returns to normal, the first The node in the area sends a status notification frame; after receiving the status notification frame, the node in the first area forwards the status notification frame to a node connected thereto; the node in the second area receives the status notification frame. After that, update its address forwarding table.

Further, the update address forwarding table message is a protocol frame of the update address forwarding table constructed by the node in the first area, and the method specifically includes: when a link failure occurs between nodes in the first area or fails When the link is restored, the node in the first area is configured to notify the protocol frame of the update address forwarding table, and sends the protocol frame. After receiving the protocol frame, the node in the first area forwards the node to the node connected thereto. The node of the second area receives the protocol frame and updates its address forwarding table.

Further, the update address forwarding table message is a protocol frame of the update address forwarding table constructed by the node in the first area, where the method specifically includes: when a link failure occurs between nodes in the first area, A node in the first area and connected to the faulty link constructs a protocol frame of the address forwarding table, and adds state information indicating that the current state of the first area is a fault state to the protocol frame, and the faulty link is connected thereto The connected node periodically sends or sends the protocol frame only once or several times; when the inter-node failure link in the first area returns to normal, the node located in the first area constructs a protocol for updating the address forwarding table. And adding, in the protocol frame, state information indicating that the current state of the first area is a recovery state, and periodically transmitting or transmitting the protocol frame to the node connected thereto; the first area; After receiving the protocol frame, the node in the inner node forwards the protocol frame to a node connected thereto; the node in the second area receives the protocol frame After that, it is judged whether the status information included in the previous protocol frame of the record is the same as the status information in the current protocol frame. If they are the same, the address forwarding table is kept unchanged; if not, the address forwarding table is updated.

Further, when a link failure occurs between nodes in the first area, a node that is located in the first area and is connected to the faulty link, or a faulty link between the nodes in the first area returns to normal, is located at the When a node in an area sends a protocol frame, when it is determined that there is a path switch in the first area, the update indication information is added to the protocol frame; when it is determined that there is no path switch in the first area, the protocol is Adding no update indication information to the frame; after receiving the protocol frame, the node of the second area determines that the status information included in the recorded previous protocol frame is different from the status information in the current protocol frame, Whether the update indication information or the update indication information is included in the protocol frame, and if yes, when the determination includes the update indication information, updating the address forwarding table, including not updating When the information is indicated, the address forwarding table is kept unchanged; if the update indication information is not included or the indication information is not updated, the address forwarding table is updated.

Further, after receiving the update address forwarding table message, the node of the second area forwards the update address forwarding table message to other nodes in the second area connected thereto; or the node of the second area according to the When the update address forwarding table message determines that the address forwarding table does not need to be updated, the update address forwarding table message is not forwarded to other nodes in the second area connected thereto.

The method for updating the address table of the Ethernet protection address of the present invention can prevent the address forwarding table update of each node from being incomplete when the local area path is switched in the Ethernet protection, so that the address forwarding table after the switch is incorrect, resulting in unstable convergence. Improves the quality and speed of Ethernet protection switching.

BRIEF abstract

1 is an Ethernet topology diagram of a prior art dual-homed connection protection technology;

2 is a schematic diagram of a network in a prior state in which a dual-homed connection protection technology is used in a prior art;

3 is a schematic diagram of a network in the prior art in which a dual-homed connection protection technology is used and a link in a protection domain fails;

4 is a schematic diagram of a network in a prior art in which Ethernet loop protection technology is used;

FIG. 5 is a schematic diagram of a network in which a dual-homed connection protection technology is used in the prior art and a link in a protection domain is faulty;

6 is a flowchart of a method for updating an address forwarding table of the present invention;

7 is a flowchart of a method for updating an address forwarding table according to Embodiment 1;

8 is an Ethernet topology diagram of the dual-homing connection protection technology in the first embodiment;

9 is a network diagram of the first embodiment in which the dual-homing protection technology is used and is in a normal state;

FIG. 10 is a network in a fault state in the first embodiment using dual-homing connection protection technology Schematic diagram

11 is a schematic diagram of a network in a fault state in which the Ethernet loop protection technology is used in the second embodiment;

12 is a network diagram of the Ethernet loop protection technology and the dual-homing protection technology in the third embodiment;

13 is a schematic diagram of a network in which the network using the Ethernet loop protection technology and the dual-homing protection technology in the third embodiment is in a normal state;

FIG. 14 is a schematic diagram of a network in which the network using the Ethernet loop protection technology and the dual-homing protection technology in the third embodiment is in a fault state. Preferred embodiment of the invention

As shown in FIG. 6, the address update method proposed by the present invention includes the following steps:

Step 601: When a link switch occurs in the first area of the Ethernet, the second area node needs to update the address forwarding table, and the node in the first area sends an update address forwarding table message to notify other nodes to perform an update operation of the address forwarding table.

Step 602: After receiving the update address forwarding table message, the node in the second area updates its address forwarding table according to the message.

Before the step 601, the node in the second area needs to be configured to receive the update address forwarding table message sent by the node in the first area. Each port of all nodes of the second area is configured in the virtual local area network where the update address forwarding table message is located.

In the embodiments described below, the primary ring may also be referred to as a ring, and the primary ring protection region may also be referred to as a ring protection domain; the secondary ring may also be referred to as a sub-ring, and the secondary ring protection region may also be referred to as a sub-ring protection domain. . The first area is a dual-homed protection area in the Ethernet, a secondary protection area or a primary protection area, and the second area is an area outside the first area. In other embodiments, the first area may also be an area composed of nodes set in the network; the second area may also be a partial area outside the first area. Embodiment 1

As shown in FIG. 7, the address update method specifically includes the following steps:

Step 701: Configure, on a node other than the protection area, protocol frame information that can receive the update address forwarding table.

Each port of the node is configured in a virtual local area network in which the protocol frame of the update address forwarding table is notified, so that the protocol frame can be received from each port, and the protocol frame can be forwarded through the port.

As shown in Figure 8, the dual-homed network structure includes a node S1 and a node S2. Ports 31 and 32 of node R1 are configured in a virtual local area network in which the protocol frame of the update address forwarding table is notified, so that the node can receive the notification update address protocol frame through these ports, and the protocol frame received by the node R1 from the port 31 can be Forwarding to port 32, the same protocol frame received from port 32 can be forwarded to port 31;

Similarly, the ports 41 and 42 of the node R2 are configured in the virtual local area network in which the protocol frame of the update address forwarding table is notified, so that the node can receive the notification update address protocol frame through these ports, and facilitate the protocol received by the node from the port 41. The frame can be forwarded to port 42, and the protocol frame also received from port 42 can be forwarded to port 41.

Step 702: After the path switching occurs in the Ethernet dual-homed protection domain, the node in the protection domain periodically sends a protocol frame to the neighboring node to notify other nodes to update the address forwarding table. As shown in FIG. 9, the node S1 in the protection domain is protected. After the port 11 fails, the node S1 detects the port fault and blocks the protected data forwarding function of the port 11. The node S1 includes the status information indicating that the current state of the protection domain is the fault state in the status notification frame sent to the node S2. The node S1 periodically transmits the above status notification frame to S2.

After the fault is eliminated, the node S2 writes the state of the current protection domain to the recovery state in the state information in the state notification frame transmitted to the node S1, and periodically transmits the state notification frame to the node S1. In other embodiments, after the fault is eliminated, the node S2 writes the state of the current protection domain to the recovery state in the state information in the state notification frame sent to the node S1, and sends the state notification frame to the node S1 instead of the cycle. Sexually sent.

When the node in the protection domain sends a status notification frame, it determines whether there is a path switch in the protection domain, such as If there is a path switch, the update indication information is added in the status notification frame; if there is no path switch, the non-update indication information is added in the status notification frame.

After receiving the status notification frame, the control node S2 determines that the status information is in the fault state, and then opens the protection data forwarding function of the corresponding port of the protection link, and each node in the domain updates its address to publish. The node S2 sends the status notification frame to the node R2 outside the protection domain, and the status notification frame is the protocol frame of the Ethernet dual-homed connection protection domain to notify the out-of-domain node to update the address forwarding table. In addition, when a recovery handover event occurs in the Ethernet dual-homed connection protection domain, the status notification frame sent to the extra-domain node is a protocol frame for notifying the extra-domain node to update the address forwarding table.

Step 703: The extra-domain node receives the protocol frame that notifies the update address forwarding table.

Step 704: The extra-domain node determines whether the status information included in the recorded previous protocol frame is the same as the status information in the current protocol frame. If yes, go to step 708, otherwise, proceed to the next step; Step 705, determine whether the protocol frame is Contains the update indication information or does not update the indication information, and if so, proceeds to the next step, otherwise, proceeds to step 707;

Step 706, determining whether the protocol frame is refreshed, if yes, performing the next step, otherwise, proceeding to step 708;

If the update indication information is included in the judgment protocol frame, indicating that the update address forwarding table is indicated, and the protocol frame includes not updating the indication information, indicating that the address forwarding table is not updated.

Step 707, updating the address forwarding table; the process ends;

Step 708: Keep the address forwarding table unchanged; the process ends.

As shown in FIG. 10, after receiving the protocol frame for notifying the address update, the node R2 needs to forward the protocol frame to the node R1 through the port, so that after the node R1 receives the packet, the address forwarding table is performed according to the received protocol frame. Operation, or, after determining that the address forwarding table is not needed to be updated, the node R2 may no longer forward the protocol frame to the node S1 connected thereto.

After each node in the network updates the address forwarding table, it re-learns the destination address and establishes a new path in the network. That is, the network B<-> node Sl<-> node S2<-> node R2<-> node Rl<->network A, network C <-> node S2<-> node R2<-> node Rl<-> network A. In another variation of this embodiment, in step 702 of the foregoing process, the node S1 adds the status information only in the status notification frame, does not add the update indication information or does not update the indication information; then in step 704, the protection domain is When the node determines that the status information included in the previous protocol frame of the record is different from the status information in the current protocol frame, the process proceeds directly to step 707.

Embodiment 2:

As shown in Figure 11 (same structure as the network diagram in Figure 4), the slave ring of the Ethernet loop protection system (the slave ring is also called the sub-ring) has a failover event, and the primary ring is notified. The primary ring can also be called a ring. The node performs an update of the address forwarding table.

All the ports of the primary ring are configured in the virtual local area network in which the protocol frame of the update address forwarding table is notified, so that the primary ring node can receive the status notification frame sent from the ring node, and can forward the received message to the connected node. Status notification frame.

When the link between the node S3 and the node S5 fails, the node S5 detects that the corresponding link of the port 51 is faulty, and periodically sends a status notification frame to the node S6, where the status notification frame contains the fault status information, indicating that the status is When the ring is in a fault state, the node S6 opens the port 62 after receiving the status notification frame, and sends the fault status frame to the primary ring node S4. The update address forwarding table operation is performed from each node in the ring. After receiving the status notification frame, the primary ring node S4 corresponds to the primary ring protection domain, and the frame is a notification update address protocol frame.

After receiving the status notification frame, the primary ring node S4 determines whether the status information included in the previous frame of the record is the same as the status information in the current frame. If the same, the address forwarding table is not updated; if not, the address forwarding table is updated. After determining that the two are the same, the primary ring node S4 can no longer forward the status notification frame to the node S1 connected thereto, and when the two are different, the status notification frame is forwarded to the node S1 connected thereto. After the entire ring switch is completed, a new transmission path is established between the networks, as shown in FIG.

Embodiment 3:

As shown in FIG. 12, the network A and the network B are connected by an Ethernet loop and a dual-homing topology. The node S1, the node S2, and the node S3 form a ring network protection domain, and are protected by a loop protection technology. When the link state in the loop protection domain is normal, the node S2 blocks the protection of the port 22. The data forwarding function prevents a network storm from occurring in the closed loop in the network. The link between the node S2 and the node S1 is a ring protection link in the ring protection domain. Node S4 is dual-homed to node S1 and node S2. Node S4 forms a dual-homed protection domain and is protected by dual-homing protection. Under normal conditions, node S4 blocks the protected data forwarding function of port 42. The communication path between network B and network A is: network B<-> node S4<-> node S1 <-> node S3<-> network A. As shown in FIG. 13, when the link between the node S4 and the node S1 fails, the node S4 detects the link fault, blocks the protected data forwarding function of the port 41, and opens the protected data forwarding function of the port 42 to update the address. The forwarding table is also sent out of the status notification frame.

As shown in FIG. 14, applying the method of the present invention, each port of the node S1, the node S2, and the node S3 in the ring protection domain is configured in a virtual local area network in which the protocol frame of the update address forwarding table is notified, so that it can be received from each port. This protocol frame, and this protocol frame can be forwarded out through the port. After receiving the dual-homing status notification frame, the nodes Sl, S2, and S3 perform judgment and perform an update address forwarding operation, and further forward the protocol frame. After each node updates the address forwarding table, the protection data between network A and network B finds a new path by broadcasting and relearning the address, namely: network B<-> node S4<-> node S2<-> node S3 <-> Network A, and the protection data sent by Network A to Network B and the protection data sent by Network B to Network A have reached a switching speed of less than 50ms, meeting the requirements of carrier-class Ethernet protection.

Embodiment 4:

In the fourth embodiment, different from the sending of the status notification frame in the first, second, and third embodiments, the address update protocol frame is sent to the nodes of other areas once or several times, and the nodes in other areas are notified to update the address forwarding table, which is not in the fourth embodiment. Status information is added to the address update protocol frame. After receiving the address update protocol frame from the protection domain or from the ring node, the frame is forwarded to the node connected to it. After receiving the address update protocol frame from the protected domain or from the node outside the ring, update its address forwarding table.

Embodiment 5:

Configuring a port for receiving and transmitting protection data in a protection domain or a node from an area other than the ring in a virtual local area network in which the protocol frame of the update address forwarding table is notified, so that each port can be accessed from each port This protocol frame is received and can be forwarded out through the port.

When a path switch event occurs in the protection domain or from the ring, a protocol frame is notified to update the address forwarding table (the protocol frame can be empty except for the frame header), and is sent to the node connected to it. After receiving the above protocol frame in the protection domain or from the node in the ring, the frame is forwarded to the node connected to it. After the protection domain is received or received from a node outside the ring, its address forwarding table is updated.

As in the method described in Embodiments 1, 2, and 3, the node in the protection domain or connected to the faulty link in the ring may also add the state information of the link in the constructed protocol frame, and protect the domain outside or outside the ring. After receiving the protocol frame, the node updates its address forwarding table by judging that the status information is different from the stored previous status information. When the same, the address forwarding table remains unchanged. The node in the protection domain or connected to the faulty link from the ring may also add update indication information or no update indication information in the protocol frame, and the protection indication information or the update indication information may be obtained after the protection domain or the node outside the ring receives the protocol frame or The instruction information is not updated, and the address forwarding table is operated accordingly.

In summary, the above method can prevent the address forwarding table update of each node from being incomplete when the local area path is switched in the Ethernet protection, so that the address forwarding table after the switch is incorrect, resulting in unstable convergence, and improving the Ethernet protection. Switch quality and speed.

The method of the present invention can be applied to a variety of Ethernet protection technologies.

The invention may be embodied in a variety of other embodiments without departing from the spirit and scope of the invention.

Industrial applicability

Claims

Claim

A method for updating a node address forwarding table, characterized in that

When a link switch occurs in the first area of the Ethernet, the node in the second area updates the address forwarding table; the first area is a dual-homing protection area, a secondary ring protection area, or a primary ring protection area in the Ethernet, where the The two area is an area outside the first area, or a partial area outside the first area.

2. The method of claim 1 wherein:

When the link switching occurs in the first area, the node in the first area sends an update address forwarding table message to notify other nodes to perform an update operation of the address forwarding table, and the node in the second area receives the update address forwarding table. After the message, its address forwarding table is updated based on this message.

3. The method of claim 2, wherein

When the system is initialized, each port of all nodes of the second area is configured in a virtual local area network where the update address forwarding table message is located.

4. The method of claim 3, wherein

The update address forwarding table message is a status notification frame;

The method specifically includes:

When a link fault occurs between the nodes in the first area, the node located in the first area and connected to the faulty link adds state information indicating that the current state of the first area is a fault state in the status notification frame, The status notification frame is periodically sent or sent only once or sent by a node connected to the faultless link;

When the inter-node failure link in the first area returns to normal, the node located in the first area adds state information indicating that the current state of the first area is the recovery state to the status notification frame, and the node period connected thereto Sexually transmitting or transmitting the status notification frame only once or several times;

After receiving the status notification frame, the node in the first area forwards the notification frame to the node connected thereto;

After receiving the status notification frame, the node of the second area determines whether the status information included in the recorded previous status notification frame is the same as the status information in the current status notification frame. If they are the same, Keep its address forwarding table unchanged; if it is different, update its address forwarding table.

5. The method of claim 4, wherein

The node located in the first area and connected to the faulty link when the link fails in the first area, or the faulty link between the nodes in the first area returns to normal, is located in the first area When the node transmits the status notification frame, when it is determined that there is a path switch in the first area, the update indication information is added to the status notification frame; when it is determined that there is no path switch in the first area, the node does not add the status notification frame. Update instructions;

After receiving the status notification frame, the node of the second area determines whether the status information included in the recorded previous status notification frame is different from the status information in the current status notification frame, and determines whether the status notification frame is Include the update indication information or the update indication information, and if yes, continue to determine the update indication information, update the address forwarding table, and when the judgment includes the non-update indication information, keep the address forwarding table unchanged; if the status notification frame does not The address forwarding table is updated if the indication information is updated or the indication information is not updated.

6. The method of claim 3, wherein

The update address forwarding table message is a status notification frame. The method specifically includes:

When a link failure occurs between nodes in the first area or when the faulty link returns to normal, the node located in the first area sends a status notification frame;

After receiving the status notification frame, the node in the first area forwards the status notification frame to a node connected thereto;

After receiving the status notification frame, the node of the second area updates its address forwarding table.

7. The method of claim 3, wherein

The update address forwarding table message is a protocol frame of the update address forwarding table constructed by the node in the first area, and the method specifically includes:

When a link failure occurs between the nodes in the first area or when the faulty link returns to normal, the node structure located in the first area notifies the protocol frame of the update address forwarding table, and sends the protocol frame; the first area After receiving the protocol frame, the node within the node forwards to the node connected thereto; The node of the second area receives the protocol frame and updates its address forwarding table.

8. The method of claim 3, wherein

When a link failure occurs between the nodes in the first area, the node located in the first area and connected to the faulty link constructs a protocol frame of the address forwarding table, and the first frame is added to the protocol frame. The current state is the state information of the fault state, and is periodically sent or sent only once or sent several times to the node connected to the faultless link;

When the inter-node failure link in the first area returns to normal, the node located in the first area constructs a protocol frame of the update address forwarding table, and adds the current status of the first area to the protocol frame. The status information of the status, periodically sent to the node connected thereto or sent only once or sent the protocol frame several times;

After receiving the protocol frame, the node in the first area forwards the protocol frame to a node connected thereto;

After receiving the protocol frame, the node of the second area determines whether the status information included in the recorded previous protocol frame is the same as the status information in the current protocol frame, and if the same, keeps the address forwarding table unchanged; Different, update its address forwarding table.

9. The method of claim 8 wherein:

The node located in the first area and connected to the faulty link when the link fails in the first area, or the faulty link between the nodes in the first area returns to normal, is located in the first area Adding update indication information to the protocol frame when determining that there is a path switch in the first area when transmitting the protocol frame; and determining that there is no path switch in the first area, adding, in the protocol frame, Do not update the indication information;

After receiving the protocol frame, the node of the second area determines whether the status information included in the recorded previous protocol frame is different from the status information in the current protocol frame, and determines whether the protocol frame is included. Updating the indication information or not updating the indication information. If yes, when the determination includes the update indication information, updating the address forwarding table, including not updating the indication information, keeping the address forwarding table unchanged; If the update indication information is not included or the indication information is not updated, the address forwarding table is updated.

The method according to any one of claims 4 to 9, wherein after the node of the second area receives the update address forwarding table message, the node in the second area connected thereto The node forwards the update address forwarding table message;

Or the node of the second area does not need to update the address forwarding table according to the update address forwarding table message, and does not forward the update address forwarding table message to other nodes in the second area connected thereto.