WO2010102479A1

WO2010102479A1 - Method and system for updating media access control

Info

Publication number: WO2010102479A1
Application number: PCT/CN2009/073773
Authority: WO
Inventors: 宋晓丽
Original assignee: 中兴通讯股份有限公司
Priority date: 2009-03-12
Filing date: 2009-09-07
Publication date: 2010-09-16
Also published as: CN101834771B; CN101834771A

Abstract

A method for updating media access control (MAC) address is provided by the invention. The method includes: acquiring the target MAC address to be updated according to the port corresponding to the failure link between the nodes (101); by matching the source MAC address stored by other nodes with the target MAC address, selecting the MAC address to update when the source MAC address matches the target MAC address (102). A system for updating media access control address is also provided by the invention, wherein, an acquiring unit is used for acquiring the target MAC address to be updated according to the port corresponding to the failure link between the nodes; a matching unit is used for matching the source MAC address of other nodes with the target MAC address; an updating unit is used for selecting the MAC address to update when the source MAC address matches the target MAC address. Adopting the method and system of the invention, it can be realized that MAC address is updated optionally.

Description

Method and system for refreshing media access control

The present invention relates to a medium access control (Media Access Control) (MAC) address refreshing technology, and more particularly to a method and system for refreshing a MAC address in an Ethernet ring protection switching. BACKGROUND At present, an Ethernet loop structure is widely used in a metropolitan area network or an enterprise network, and the structure can improve the reliability of the network. When a link on the ring network is disconnected, the backup link can be enabled to restore each node on the ring network. Communication between. FIG. 1 is a schematic structural diagram of an Ethernet loop. As shown in FIG. 1, the Ethernet ring network is composed of a node 21 to a node 25, a user 11 to a user 12, a link between the user 11 and the node 21, and five nodes. The link between the link and the link between the node 24 and the user 12. The link between the node 21 and the node 22 is a protection link (Ring Protection Link, the cylinder is called RPL), the node 21 is the master node, and the other nodes are all transit nodes. Normally, the node 21 blocks its corresponding RPL port (ie, port 31) to prevent the generation of a loop, and the traffic forwarded by the packet between the user 11 and the user 12 through the node 21, the node 25, the node 24, and The communication path formed by the links between them is transmitted. When a node or link of a normal communication path fails, as shown in FIG. 2, for example, the link between the node 25 and the node 24 fails, and after detecting the failure, the node 25 and the node 24 immediately block the node 25. And the faulty port of the node 24, and the corresponding fault detection protocol message is periodically sent out through the non-faulty port to notify the other nodes on the ring of the fault condition, and after receiving the fault detection protocol message, the node 21 starts the protection switching mechanism. Opening the previously blocked port 31, and switching the traffic between the user 11 and the user 12 to the RPL. At this time, the traffic between the user 11 and the user 12 passes through a new communication path, that is, the node 21, the node 22, the node 23, Nodes 24 and the links between them are transmitted. Moreover, in order to implement fast protection switching, the other nodes on the ring receive the MAC address refresh after receiving the fault detection protocol message. The existing MAC address is refreshed by directly deleting the forwarding address database ( Filter

Database, the cartridge is called FDB. All the entries in the list. This refresh method is a "one size fits all" refresh method, which is unreasonable. In fact, when a node or link on a loop fails, not all MAC addresses need to be refreshed. With the refresh method of the existing MAC address, since many entries that do not need to be refreshed are deleted, the convergence time of the protection switching is longer. At the same time, this over-refreshing directly deletes all entries in the FDB table without selecting, so that the packets between users are transferred. The lack of forwarding path is based on the lack of forwarding path. It does not know how to forward the packet. The traffic that is forwarded by all the users on each node is flooded on the ports of the loop. In severe cases, it will lead to congestion and have to discard some business. In summary, in order to avoid traffic flooding caused by too long convergence time and excessive refresh, a new refresh method is urgently needed, and an effective solution has not been proposed for this new refresh method. SUMMARY OF THE INVENTION In view of this, the main purpose of the present invention is to provide a method and a system for refreshing a MAC address, which can implement selective and reasonable refreshing, which can avoid not only excessive convergence time, but also avoid traffic flooding caused by excessive refresh. . In order to achieve the above object, according to an aspect of the present invention, a method for refreshing a medium access control address is provided. The method for refreshing the media access control address according to the present invention includes: acquiring a target media access control MAC address to be refreshed according to a port corresponding to the fault link between the nodes; and by using a source MAC address of the other node and the target MAC The address is matched, and the MAC address matching the source MAC address and the target MAC address is selected to be refreshed. The obtaining the target MAC address is specifically: querying in the FDB list of the forwarding address database of the node, querying and obtaining the corresponding target MAC address according to the port number corresponding to the faulty link; The MAC address creates an address deletion list. Before the matching, the method further includes: the other node receiving the address deletion list; querying in the FDB list of other nodes, if the corresponding MAC address cannot be queried according to the port number corresponding to the other node receiving the address deletion list, The refresh of the MAC address is not performed; otherwise, the matching is performed according to the queried MAC address. The performing the matching is specifically: parsing the target MAC address from the address deletion list, matching the queried MAC address with a target MAC address, and if yes, performing a MAC address refresh; otherwise The MAC address is not refreshed; the MAC address is refreshed by: deleting the MAC address matching the target MAC address, and implementing refreshing. Before acquiring the target MAC address, the method further includes: returning, to all nodes including the node and the other node, all MAC addresses in the current FDB list of each node by using a node MAC address of each node Class, and the node MAC address is used as an index of the current FDB list. The method further includes: first forwarding, by the node, the protocol message including the address deletion list to the other node, and simultaneously performing MAC address refreshing on each of the other nodes. In order to achieve the above object, according to another aspect of the present invention, a refresh system for a medium access control address is provided. The system for refreshing the media access control address according to the present invention includes: an obtaining unit, a matching unit, and a refreshing unit; wherein the acquiring unit is configured to acquire a target MAC address to be refreshed according to a port corresponding to the faulty link between the nodes; a unit, configured to match a source MAC address of the other node with the target MAC address, and a refresh unit, configured to select to refresh the MAC address that matches the source MAC address and the target MAC address. The system further includes: a creating unit, configured to create an address deletion list according to the target MAC address. The system further includes: a forwarding unit, configured to forward, by the node, a protocol message including the address deletion list to the other node. The matching unit is further configured to parse the target MAC address from the address deletion list; ^^ according to the MAC address and the target MAC address queried by the port number corresponding to the other node receiving the address deletion list. Matching, if matched, the matching unit notifies the refresh unit to perform refresh of the MAC address; otherwise, the matching unit does not notify the refresh unit to perform refresh of the MAC address; the refresh unit is further configured to match the target MAC address The MAC address is deleted and implemented to be refreshed. The invention obtains a target MAC address to be refreshed according to a port corresponding to the fault link between the nodes; and selects a MAC that matches the source MAC address and the target MAC address by matching the source MAC address of the other node with the target MAC address. The address is refreshed. The present invention is different from the "one size fits all" and unselected refresh mechanism adopted by the existing MAC address refreshing method, and obtains the target MAC address to be refreshed according to the port corresponding to the faulty link between the nodes; and then uses the target MAC address. The address is used as a basis for selectively and MAC address refreshing, and only refreshing the MAC address matching the target MAC address is more reasonable than the existing MAC address refreshing method. By adopting the invention, the traffic of the traffic caused by the over-refresh can be prevented from flooding on the Ethernet loop while effectively reducing the protection switching time. BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are set to illustrate,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, In the drawings: FIG. 1 is a schematic structural diagram of an Ethernet loop in a normal dog state; FIG. 2 is a schematic structural diagram of an Ethernet loop in a protection switching state; FIG. 3 is a media access control according to an embodiment of the present invention; FIG. 4 is a schematic diagram of an FDB list corresponding to each node in an Ethernet loop in a normal state according to an example of the present invention; FIG. 5 is an Ethernet loop in a fault state according to an example of the present invention; FIG. 6 is a schematic diagram of a MAC address refresh list corresponding to each node in an Ethernet loop according to an example of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention provides a media access control in consideration of the problems existing in the prior art. Address refreshing scheme, the basic idea of the scheme is: obtaining a target MAC address to be refreshed according to a port corresponding to a faulty link between nodes; and then performing MAC address refreshing selectively and ambiguously using the target MAC address The basis is that only the MAC address matching the target MAC address is refreshed, which is more reasonable than the existing MAC address refresh method. The implementation of the technical solution will be further described in detail below with reference to the accompanying drawings. Method Embodiment FIG. 3 is a schematic flowchart of a method for refreshing a media access control address according to an embodiment of the present invention. As shown in FIG. 3, the method includes the following steps: Step 101: According to a port corresponding to a fault link between nodes , get the target MAC address to be refreshed. Step 102: Select, by matching the source MAC address saved by the other node with the target MAC address, to refresh the MAC address matching the source MAC address and the target MAC address. Here, the other nodes refer to: nodes on the Ethernet loop other than the nodes involved in step 101. The node involved in step 101 refers to: Two nodes connecting the faulty link. Here, the source MAC address may be: a MAC address pre-stored in the FDB list among other nodes on the Ethernet ring. The target MAC address to be refreshed in step 101 may be: deleting the MAC address in the list according to the address created by the port corresponding to the faulty link between the nodes, and the MAC address is selectively performing MAC on other nodes. When the address is refreshed, the MAC address on which the purpose of selective and accurate refresh is achieved is specifically matched only when the MAC address in the FDB list of the other node matches the MAC address in the address deletion list. The MAC address is refreshed, that is, the entry of the entry in the FDB list of the other node where the matched MAC address is located is deleted. For the technical solution of the step 101 to the step 102, the specific processing procedure of the step 101 is: querying in the FDB list saved by the node, and querying and obtaining the corresponding target MAC address according to the port number corresponding to the faulty link. After that, the address deletion list (Delete Address List) is created according to the target MAC address. After the DAL is created, the step 101a may be further included before the matching in step 102 is performed. The specific processing procedure of step 101a is: receiving DAL by other nodes; querying in the FDB list saved by other nodes, if receiving DAL according to other nodes The port number cannot be queried. If the MAC address is not used, the MAC address is not refreshed; otherwise, step 102 is performed, and the matching of the MAC address is performed according to the queried MAC address. Here, the other node receives the port number corresponding to the DAL, which is the same as the port number corresponding to the fault link between the nodes mentioned in the specific processing of step 101 above. Because the two are the same, the target MAC address obtained by the port number corresponding to the faulty link between the nodes can be matched to the MAC address of the other node, and the other can be selectively performed. The MAC address of the node is refreshed. The specific process of performing MAC address matching in step 102 is: parsing the target MAC address from the DAL, matching the queried MAC address with the target MAC address, and if yes, performing MAC address refresh; otherwise, not performing The MAC address is refreshed. The specific process of performing the MAC address refreshing in step 102 is as follows: In the FDB list saved by other nodes, the MAC address matching the target MAC address is deleted, and refreshing is implemented. It should be noted that the FDB list saved by the node and the type of the FDB list saved by other nodes include two cases. The list type of the first case includes two contents, specifically: the MAC address of the node and the access under the node. User MAC address, and port number. In this way, the destination MAC address is obtained based on the port number, and then the DAL contains all the MAC addresses when the DAL is created according to the target MAC address. That is to say, the target MAC address at this time includes both the MAC address of the node and the user MAC address accessed by the node. Then, when the matching and MAC address refresh are performed based on the target MAC address in the DAL, the DAL is severely redundant due to the large number of target MAC addresses, and therefore, the MAC address refresh efficiency is affected. In order to speed up the refreshing efficiency of the MAC address and avoid the DAL redundancy caused by the large number of the target MAC addresses, the categorization step may be added before the target MAC address is obtained in step 101. The specific processing procedure of the categorization step is: For all nodes, including other nodes, all the MAC addresses in the current FDB list saved by each node are classified by the node MAC address of each node in all nodes, and the node MAC address is used as the current FDB list. index. Therefore, in the first case, the FDB list saved by the node and the type of the FDB list saved by other nodes are all FDB lists indexed by the node MAC address after categorization; the target MAC address included in the DAL is the node MAC The address is: The list type of the second case includes three items, specifically: The node MAC address as the list index is the first item in the list, the MAC address of the node and the user MAC address accessed under the node. It is the second item in the list, and the third item in the list is the port number. And, create DAL, match, and refresh Both are based on the list index corresponding to the port number. That is to say, the target MAC address at this time only includes the MAC address of the node as the list index. Then, when the matching and MAC address refresh are performed based on the target MAC address in the DAL, since the target MAC address only includes the MAC address of the node, Different from the first case, the number of target MAC addresses is greatly reduced, thereby avoiding the redundancy of the DAL and greatly improving the refresh efficiency of the MAC address. It should be noted that the method further includes: first forwarding the protocol message including the DAL from the node to the other node, and then simultaneously performing MAC address refreshing on each node in the other node. In summary, the present invention selectively and explicitly performs MAC address deletion to implement MAC address refresh, including the following contents: First, the faulty link neighbor node detects a fault, and then according to the phase The DAL is created on the port corresponding to the faulty link between the neighboring nodes. The DAL is created as follows: In the FDB list saved on the adjacent node of the faulty link, find the MAC address of the ring node associated with the port corresponding to the faulty link. The entry of the entry, according to the MAC address of the node, clears the corresponding MAC address in the FDB list on the neighboring node of the faulty link, and creates a DAL according to the MAC address of the node; subsequently, the protocol message carrying the DAL is along the non-failed port. Broadcast or multicast on the ring. After receiving the protocol message containing the DAL, the other nodes on the ring except the neighboring nodes of the faulty link look up the MAC address of the ring node corresponding to the protocol message receiving port in the FDB list. That is, the port MAC address is received, and the MAC address of the receiving port is matched with the MAC address in the DAL. If the matching is successful, the M in the FDB list indexed by the MAC address is deleted. The AC address group is used to implement the refresh of the MAC address. At the same time, the node forwards the protocol message carrying the DAL between other nodes on the ring before the matching process, that is, the protocol message carrying the DAL can be as fast as possible. The multicast is forwarded on all ring nodes by multicast or broadcast, and the FDB list refresh operation is performed almost simultaneously at each node. The method provided by the embodiment of the present invention includes three technical points: First, each node of the Ethernet loop pairs all MAC addresses in the current FDB, that is, the MAC address of the node including each node on the ring and the user MAC address accessed by each node. All MAC addresses within are categorized and categorized by the node MAC address of each node on the ring. In this way, when the subsequent D AL is created and the MAC address is deleted based on the MAC address of the node, the MAC address refresh efficiency is greatly improved. Second, when a loop fault occurs, the neighboring node of the faulty link detects the fault condition, immediately blocks its corresponding faulty port/blocking port, and searches for the corresponding ring in the FDB table according to the port corresponding to the faulty link. Node MAC address, create DAL. In this way, the subsequent deletion of the corresponding MAC address group in the FDB is performed according to the DAL, and the protocol message carrying the DAL is periodically broadcast/multicasted on the ring through the non-faulty port to notify other nodes on the ring of the failure. Happening. Third, after receiving the protocol message carrying the DAL, the master node on the ring initiates a protection switching mechanism to open the previously blocked RPL port, switches the traffic to the new communication path, and carries the DAL on the new communication path. The protocol message is forwarded between the nodes on the ring. At the same time, the MAC address is updated according to the DAL. After receiving the protocol message carrying the DAL, the other nodes on the ring forward the packet and record the port number of the protocol message carrying the DAL. According to the port number, find the corresponding MAC address of the ring node in the FDB list of the respective port, that is, the MAC address of the receiving port, and the MAC address of the receiving port is matched with the MAC address in the DAL. If the matching is successful, then The MAC address group indexed by the MAC address is deleted in the FDB list. The refresh method provided by the embodiment of the present invention can not only perform selective, fast, and accurate MAC address refreshing, but also can effectively shorten the convergence time, and prevent the traffic data caused by excessive refreshing from flooding in the network. . In addition, the refresh method can also solve the address refresh problem of multi-point fault and non-inversion mode in the Ethernet ring network. The reason is that, based on the existing Ethernet ring address MAC address refresh mechanism, to prevent duplicate refresh, the node is set to perform MAC address refresh only once in one state. If it is in the protection switching state, it can only be refreshed once. This will lead to two problems. First, when a node has multiple faults, if the faulty node recovers first, the network topology changes, but all the nodes on the loop are in the protection switching state. Perform a MAC address refresh. Second, in non-reversal mode, if a link in the ring network fails, all nodes enter the protection switching state. Because the ring network works in non-reversal mode, all nodes cannot be protected even after the fault is recovered. Switching state, so when a new fault occurs, although the network topology changes, all nodes cannot perform MAC address refresh. However, with the refresh method provided by the embodiment of the present invention, since the possibility of repeated refresh is very small, the principle that the state is only refreshed once is not set, and the above-mentioned multi-point fault and non-reverse mode do not exist. The MAC address under the refresh problem. The example consists of an Ethernet consisting of four nodes, users accessing each node, and links between them. Among them, each node is represented by A, B, C and D respectively; the users accessed under node A are represented by AX, AY and AZ respectively; the users accessed under node B are represented by BX; the users accessed by node C are respectively CX and CY are indicated. Each port number in the figure indicates the port number corresponding to the link accessed by each node; node A is the master node, the link between node A and node B is the RPL link, and a represents the RPL port of node A. Normal ^! In the dog state, node A first blocks the RPL port, and in the faulty state, the protection switching mechanism is started. At jtb, node A opens the previously blocked RPL port. FIG. 4 is a diagram showing an FDB column corresponding to each node in an Ethernet loop in a normal state according to an example of the present invention. Schematic diagram of the table. FIG. 5 is a schematic diagram showing the creation of a MAC address deletion list in an Ethernet ring in a fault state according to an example of the present invention. FIG. 6 is a schematic diagram of a MAC address refresh list corresponding to each node in an Ethernet loop according to an example of the present invention. In this example, the specific processing procedure of the MAC address refresh method is: Normal ^! In the dog state, node A blocks its corresponding RPL port to prevent loop formation. The FDB list maintained by each node on the ring is shown in Figure 4. Each node on the ring pairs all user MAC addresses and ring node MAC addresses. The node MAC on the ring is classified for the index. In the fault state, when the link on the ring fails, for example, the link between node C and node D fails. First, node C and node D detect the fault and block their corresponding fault ports respectively. The port corresponding to the path finds the MAC address of the corresponding ring node and creates a DAL. The created DAL and the FDB list maintained by each node on the ring are shown in Figure 5. Moreover, the DAL in FIG. 5 includes three contents, DA (ie, destination address) as the first item in the DAL; SA (ie, source address) as the second item in the DAL; the specific value of the DAL as the DAL The third item, the specific value of the DAL refers to the target MAC address to be refreshed. For the node C, the port corresponding to the faulty link is port 2, then the port number is 2, and the FDB list corresponding to the node C is known. The destination MAC address is D, and the D is the MAC address. The indexed MAC address group, thus the DAL created for node C contains D. For node D, the port corresponding to the faulty link is port 1, then the port number is 1, and the FDB list corresponding to node D is known. The destination MAC address is A, B, and C, and is cleared by A. B and C are MAC address groups indexed by the MAC address. Thus, the DAL created for node D contains A, B, and C. Then, according to the DAL created for the node C and the DAL nodes created for the node D, the corresponding MAC addresses in the FDB list are respectively cleared, and at the same time, the protocol message for the DAL created for the node C is created and created for the node D. The protocol message of the DAL is periodically broadcasted on the ring along the non-faulty port, and the fault condition is notified to other nodes on the ring. After receiving the protocol message, the node A initiates the protection switching mechanism to open the previously blocked RPL port, and switches the traffic to the RPL link. After receiving the protocol message carrying the DAL list, each node on the ring combines the receiving of the protocol message. The port compares the MAC address of the ring node corresponding to the receiving port in the FDB list of the node with the MAC address in the DAL. If the port matches the MAC address of the corresponding ring, the MAC address group of the FDB list is deleted. As shown in FIG. 6, for example, the node A receives the protocol message of the DAL created by the node D from the node D through the port 1, and searches for the MAC address of the ring node corresponding to the port 1 in the FDB list. Empty, no need to refresh, and node A receives the protocol of DAL created by node C for node C through port 2. The message finds the MAC address of the ring node corresponding to port 2 in its own FDB list, and the search result includes three addresses B, C, and D, and matches the search result with the DAL sent by node C, because node C sends it. The DAL contains the MAC address D. Therefore, the entry in the FDB list that matches the MAC address D is deleted. That is, the port 2 corresponds to the entry with the MAC address being D. Similarly, the node B takes the node on the ring corresponding to the port 1. The MAC address A matches the DAL created by the node D, and the MAC addresses C and D corresponding to the port 2 match the DAL created by the node C, and in the FDB list, delete the DAL created by the node D and the DAL created by the node C. All the matching entries, that is, the MAC address group indexed by A and the MAC address group indexed by D. According to an embodiment of the present invention, there is also provided a computer readable medium having stored thereon computer executable instructions for causing a computer or processor to perform, for example, when executed by a computer or processor The processing of step S101 and step S102 shown in Fig. 3, preferably, one or more of the above embodiments may be performed. System Embodiment The embodiment of the present invention provides a MAC address refreshing system, where the system includes: an obtaining unit, a matching unit, and a refreshing unit. The obtaining unit is configured to obtain a target MAC address to be refreshed according to the port corresponding to the faulty link between the nodes. A matching unit, configured to match source MAC addresses of other nodes with target MAC addresses. The refresh unit is configured to select to refresh the MAC address that matches the source MAC address and the destination MAC address. Here, the system may further include: a creating unit, configured to create an address deletion list with the target MAC address. Preferably, the system may further include: a forwarding unit, configured to forward the protocol message including the address deletion list to the other node by the node. The matching unit is further configured to parse the target MAC address from the address deletion list; in the FDB list saved by the other node, the MAC address and the target MAC address that are queried according to the port number corresponding to the address deletion list received by the other node are performed. Matching, if matched, the matching unit notifies the refresh unit to perform refresh of the MAC address; otherwise, the matching unit does not notify the refresh unit to perform refresh of the MAC address. The refresh unit is further configured to delete the MAC address matching the target MAC address in the FDB list saved by the other node, and implement refreshing. In summary, the medium access control address refreshing scheme provided by the embodiment of the present invention is more than the existing one. The MAC address refresh scheme is more reasonable, and it can effectively reduce the protection switching time and prevent the traffic caused by over-refreshing from flooding on the Ethernet loop. In addition, the implementation of the present invention does not modify the system architecture and the current processing flow, is easy to implement, facilitates promotion in the technical field, and has strong industrial applicability. The above is only the preferred embodiment of the present invention and is not intended to limit the scope of the present invention.

Claims

Claim

A method for refreshing a media access control address, the method comprising:

Obtaining a target media access control MAC address to be refreshed according to a port corresponding to the faulty link between the nodes;

The MAC address matching the source MAC address and the target MAC address is selected to be refreshed by matching the source MAC address of the other node with the target MAC address.

2. The method according to claim 1, wherein the obtaining the target MAC address is: querying in a FDB list of a forwarding address database of a node, querying according to a port number corresponding to the faulty link, and obtaining a corresponding The target MAC address; after that, the address deletion list is created according to the target MAC address.

The method according to claim 2, wherein before the matching, the method further comprises: the other node receiving the address deletion list; querying in an FDB list of other nodes, if the address deletion list is received according to other nodes If the corresponding port number cannot be queried to the corresponding MAC address, the MAC address is not refreshed; otherwise, the matching is performed according to the queried MAC address.

The method according to claim 3, wherein the performing the matching is specifically: parsing the target MAC address from the address deletion list, and performing the queried MAC address and a target MAC address. Match, if it matches, perform MAC address refresh; otherwise, MAC address refresh is not performed;

The refreshing of the MAC address is specifically: deleting a MAC address matching the target MAC address, and implementing refreshing.

The method according to any one of claims 1 to 4, further comprising: before acquiring the target MAC address, further comprising: for each node including the node and the other node, The node MAC address of the node classifies all MAC addresses in the current FDB list of each node separately, and uses the node MAC address as an index of the current FDB list.

The method according to claim 2, further comprising: first forwarding, by the node, the protocol message including the address deletion list to the other node, and then each of the other nodes The node simultaneously implements the refresh of the MAC address.

A system for refreshing a media access control address, the system comprising: an acquiring unit, a matching unit, and a refreshing unit;

An obtaining unit, configured to obtain a target MAC address to be refreshed, according to a port corresponding to the faulty link between the nodes;

a matching unit, configured to match source MAC addresses of other nodes with the target MAC address;

And a refreshing unit, configured to select to refresh the MAC address matching the source MAC address and the target MAC address.

8. The system according to claim 7, wherein the system further comprises: a creating unit, configured to create an address deletion list according to the target MAC address.

9. The system according to claim 8, wherein the system further comprises: a forwarding unit, configured to forward, by the node, a protocol message including the address deletion list to the other node.

10. The system according to claim 8 or 9, wherein the matching unit is further configured to parse the target MAC address from the address deletion list; and receive an address deletion list according to other nodes. The MAC address queried by the corresponding port number is matched with the target MAC address. If the matching, the matching unit notifies the refresh unit to perform the refresh of the MAC address; otherwise, the matching unit does not notify the refresh unit to perform the refresh of the MAC address;

The refreshing unit is further configured to delete the MAC address that matches the target MAC address, and implement refreshing.