CN101895454A

CN101895454A - Multi-layer subring-based address flush method

Info

Publication number: CN101895454A
Application number: CN2010102304003A
Authority: CN
Inventors: 王斌
Original assignee: Nanjing Post and Telecommunication University
Current assignee: Nanjing Post and Telecommunication University; Nanjing University of Posts and Telecommunications
Priority date: 2010-07-19
Filing date: 2010-07-19
Publication date: 2010-11-24

Abstract

The invention discloses a multi-layer subring-based address flush method, which comprises that: after receiving a nearest protocol message on a port of an upper-layer subring and performing flush on an address forwarding table, an interconnecting node records the port number of the port where the protocol message is received; if a lower-layer subring of the interconnecting node needs to transmit a Flush message on a control channel of the upper-layer subring of the interconnecting node due to the change of topology, the interconnecting node only transmits the Flush message to another unrecorded port of the upper-layer subring; after receiving the Flush message, the on-ring port of the node on the subring is checked whether to start a flush timer (FT) or not; when the FT in the port is overtime, if a variable 1 is equal to 0 and a variable 2 is not equal to 11, the node with the port performs the flush on the address forwarding table; and if the variable 1 is not equal to 0, the node with the port performs the flush on the address forwarding table. The method has the advantages of solving the problem on the address flush of the multi-layer subring and improving network performance.

Description

A kind of address flush method based on multi-layer subring

Technical field

The present invention relates to data communication field, relate more specifically to a kind of address flush method based on multi-layer subring.

Background technology

In the practical application of Ethernet, extensively adopted various resist technologies, realize the redundancy backup between primary path and the backup path.When primary path and backup path when all being intact, block the protected data forwarding capability of backup path, the protected data between the network is transmitted on primary path; When primary path goes wrong; open the protected data forwarding capability of backup path; protected data between the network switches on the backup path to be transmitted; realize preventing under the network normal condition that protected data is repeated to receive and form broadcast storm; when breaking down, enables the primary path of network backup path transmission protected data; improve the fault-resistant ability of Ethernet, and satisfy convergence time when switching less than the high real-time requirement of 50ms.

Many ring protections of Ethernet technology for example, as shown in Figure 1, node S1 to S6 is an Ethernet switch, and network B is connected with node S2, and network A is connected with node S5.Communicate between network A and the network B.Four physical pathways are arranged between network A and the network B, that is: network A ← → node S5 ← → node S3 ← → node S2 ← → network B, network A ← → node S5 ← → node S3 ← → node S4 ← → node S ← → node S2 ← → network B, network A ← → node S5 ← → node S6 ← → node S4 ← → node S3 ← → node S2 ← → network B, network A ← → node S5 ← → node S6 ← → node S4 ← → node S1 ← → node S2 ← → network B.

Resist technology for the many rings of Ethernet; the standard of formulating in the world (as ITU G.8032) is thought should comprise ring and subring in the protecting network of too many ring; promptly; ring (Ring) is a complete ether ring; subring (Sub-Ring) is a kind of ether ring that links to each other with other ring or network by interconnecting nodes (InterconnectionNode), and interconnecting nodes (InterconnectionNode) is the common node that belongs to two or more ether rings simultaneously.Shown in Fig. 2 a, comprise a ring and a subring among the figure, Ring1 is a ring, Ring2 is subring.The node that Ring1 comprises has S1, S2, S3 and S4, and the link that comprises has:＜S1, S2 〉,＜S2, S3 〉,＜S3, S4〉and＜S4, S1 〉; The node that Ring 2 comprises has S3, S5, S6 and S4, and the link that comprises has:＜S3, S5 〉,＜S5, S6〉and＜S6, S4 〉.Need benly to be＜S3 S4 link belongs to Ring1 and do not belong to Ring2.In looped network; under the trouble-free situation of looped network; in the ring; need one section link the data message forwarding is in blocked state to prevent into ring; this section link is commonly referred to as the ring protection link and (or often blocks link; Deng), the switching in primary path and protection path in participating in encircling by this section ring protection link.The node that has the ring protection link is called ring protection controlling links node here.Shown in Fig. 2 a, in Ring 1, node S1 is a ring protection controlling links node, with the 11 direct port connection links of node S1 be the ring protection link of Ring1.In Ring2, node S6 is a ring protection controlling links node, with the 62 direct port connection links of node S6 be the ring protection link of Ring2.Under normal circumstances, the ring protection controlling links node of Ring1 and Ring2 blocks their link to each other with ring protection link forwardings of data message (the professional data message of civilian middle finger protection) of port, prevents that protected data is repeated to transmit and the formation broadcast storm.

When the link in the ether multi-ring network when all being intact, the ring protection controlling links node of ring and subring obstruction is from the protected data forwarding capability of port.Shown in Fig. 2 a, node S1 has blocked the protected data forwarding capability of port w, and node S6 has blocked the protected data forwarding capability of port w, and the communication path of network B and A is: network B ← → node S2 ← → S3 ← → S5 ← → network A.

When the link of ether multi-ring network breaks down; if faulty link is not the ring protection link; then ring protection controlling links node is opened the protected data forwarding capability of ring protection link adjacent port, and each node also wants refresh address to transmit, and communication is according to new path transmission between the network.Shown in Fig. 2 b; fault has taken place in node S2 on the ring Ring1 and the link between the S3; after node S2 detects link failure; the data forwarding function of blocking the adjacent port of faulty link; notify other node links that fault has taken place; after node S1 receives signalling trouble; open the protected data forwarding capability of port w; each node on the Ring1 also wants refresh address to transmit in addition, and the communication path that network B and A are new is: network B ← → node S2 ← → node S1 ← → node S4 ← → node S3 ← → node S5 ← → network A.

When the link-recovery in the ether multi-ring network, recover to switch, the transmission path when Network Transmission returns to normal condition, because path changing, node also needs to carry out refreshing of addresses forwarding table.

When the ether multi-ring network being safeguarded and protect switching; need to propagate a large amount of control messages; these control messages are to propagate in the protection control channel automatically; automatically the protection control channel has two classes; one class is corresponding to ring; the control channel that is called ring, another kind of is corresponding to subring, is called the control channel of subring.The control channel of ring is configured in the ring.The control channel of subring has two kinds of configuration modes, and a kind of is the configuration mode that does not have tunnel, that is, the control channel of subring only is configured in the subring, shown in Fig. 3 a.The control channel configuration of another kind of subring comprises part and the tunnel (Virtual Channel) in the subring.Tunnel is to be configured in other network between the interconnection point or other ring provide (comprising other subring) subring of transmission channel for the subring protocol massages control channel.Shown in Fig. 3 b, the control channel of subring Ring2 not only is configured on the subring Ring2, but also is configured on the ring Ring1, is configured in the part of ring on the Ring1 and is the tunnel that the protocol massages for subring Ring2 provides.Because the existence of tunnel, the protocol massages of subring can arrive any one node in the subring.

Below, we discuss the address flush problem of subring in the ether multi-ring network.

For example in Fig. 4, when ether multi-ring network fault-free, the communication path of network B and A is: network B ← → node S2 ← → S3 ← → S5 ← → network A.When the link occurs fault of subring Ring2; shown in Fig. 4 a; link＜S3 of Ring2; S5〉fault appearred; after node S5 detected port w respective links and breaks down, refresh address was transmitted, the protected data forwarding capability of port blocked w; open the protected data forwarding capability of port e, and outwards send the malfunction frame.After node S6 received the malfunction frame, refresh address was transmitted, and formed new transmission path between network A and the network B.After node S5 and node S6 refresh address were transmitted, network A was issued the broadcasting of the protected data of network B by node S5 and node S6, finally can arrive network B, simultaneously each node address of learning network A.But, before network A is issued the network B data, if network B sends data to network A, a large amount of packet loss phenomenons will take place so.This is because node S2 does not also have refresh address to transmit; clauses and subclauses before still switch in the path in the addresses forwarding table; i.e. Cuo Wu address entry; network B is sent to the protected data of network A and still transmits according to the addresses forwarding table of mistake; promptly the outbound port e from node S2 sends; these data are in fact because link failure and port block can not arrive network A; after having only the switches learn of wait to arrive the address outbound port of correct network A; just can reach network B; therefore the flow that whether has network A to be sent to network B is depended in the network B path that is sent to network A switching time, and this time sometimes even surpass 50ms.When the protection switching that Ethernet Circle recovers, also there is similar problem in the communication between network A and the B.

From top analysis as can be seen, when the link occurs fault of subring, subring need send protocol massages by interconnecting nodes other network beyond this subring and notify the node refresh address of other network to transmit.

G.8032v2, existing ITU-T to the solution of this problem is:

Base case: when the subring topology changes, after and if only if interconnecting nodes is received the protocol massages that has address flush information and is refreshed self addresses forwarding table from subring, this interconnecting nodes is just constructed new address flush protocol massages (Flush message), and on the continuous control channel that 3 times this protocol massages is sent to other ring between the interconnecting nodes or subring, after node in other ring or the subring is received this protocol frame, refresh addresses forwarding table separately.

Though existing subring address flush base case can guarantee other ring between the subring interconnecting nodes or the node in the subring and obtain address flush, but diffusion has but taken place in the address flush message of subring interconnecting nodes structure, this will cause that all the node refresh addresses on the tunnel transmit, cause the broadcast storm that moment takes place in ring that the ether multi-ring network is all and the subring, influenced the overall performance of ether multi-ring network greatly.Through heated discussion, thinking that the ring of subring refreshes outward must be through optimizing in the ITU-T meeting G.8032.The final result who discusses of meeting is: the ring of subring refreshes the address of node of (at data and message transmission) on the unobstructed path that only needs to refresh between the subring interconnecting nodes outward and transmits.Simultaneously, there is the participant to propose a very novel scheme.

Shown in Fig. 5 a, Ring1 and Sub-ring2 constitute the ether multi-ring network.Ring1 is a complete closed loop, and the node that comprises has S1～S6, and the link between them.Sub-ring2 is a subring, and the node that comprises has S6, M1～S5 and S3, and the link that comprises has:＜S6, M1 〉,＜M1, M2 〉,＜M2, M3 〉,＜M3, M4〉and＜M5, S3 〉.The Control Node of Ring1 is S1, and what link to each other with the w port is the ring protection link.The Control Node of Sub-ring2 is M1, and what link to each other with the e port is the ring protection link.Node S6 and S3 are the interconnecting nodes that two rings intersect.The criterion that ring upper port numbering is followed is: with counterclockwise flow path direction is benchmark, and the port that enters node is the e port, and the port that leaves node is the w port.

In Fig. 5 b, link＜M3 of Sub-ring2, M4〉fault taken place, node M 3 and M4 block the forwarding capability of the data message of the port adjacent with faulty link, refresh addresses forwarding table separately, and outwards periodically send SF message (warning message) along their intact ring upper port separately.Come from the SF protocol massages that M3 sends when S6 receives first, refresh address is transmitted, and constructs the Flush message simultaneously and spread propagation on the control channel of Ring1; Come from the SF protocol massages that M4 sends when S3 receives first, refresh address is transmitted, and constructs the Flush message simultaneously and spread propagation on the control channel of Ring1.

According to traditional scheme, all nodes on the Ring1 all can be transmitted by refresh address after receiving the Flush message, cause instantaneous broadcast storm.In order to improve this scheme, the ITU-T expert that attends a meeting has G.8032 proposed a kind of new prioritization scheme, is that the refresh path of Flush message in Ring1 is: S6 ← → S5 ← → S4 ← → S3.Concrete scheme is:

Prioritization scheme:

When the node on the ring is received the Flush message from its a ring upper port, check whether its another ring upper port has started FT timer (Flush Timer).If started, this node refresh address is transmitted, otherwise starts the FT timer at the port that receives this message.

The core concept of the prioritization scheme of front is to have utilized on the Ring1 closed loop, and the node on the unobstructed path (node S5 and S4) can be received the Flush message from its two ring upper port.This scheme can refresh problem with the diffusion that low-down cost has solved the Flush message well.But this prioritization scheme only can be used to have the scene of individual layer subring, is inapplicable to the scene of multi-layer subring.Below we just set off a discussion at this problem.

Shown in Fig. 6 a: this scene is the scene of a multi-layer subring, comprises Ring1, Sub-ring2 and Sub-ring3.Wherein, the configuration of Ring1 and Sub-ring2 is with Fig. 5 a, and the node that Sub-ring3 comprises has M2, N1～N3 and M4, and the link that comprises has＜M1, N1 〉,＜N1, N2 〉,＜N2, N3〉and＜N3, M4 〉.N1 is the Control Node of Sub-ring3.

Shown in Fig. 6 b, the M3 node of Sub-ring2 starts manual switch (MS) at its e port, and this node periodically outwards sends the MS message along its two ring upper port, and Control Node M1 opens the data forwarding function of ring protection link.After a period of time, the ether multi-ring network enters stable state.But subsequently Sub-ring3＜N2, N3〉fault taken place in link, node N2 and N3 block the forwarding capability of the data message of the port adjacent with faulty link, refresh addresses forwarding table separately, and outwards periodically send SF message (warning message) along their intact ring upper port separately.When M2 and M4 received the SF protocol massages first, refresh address was transmitted, and constructs the Flush message simultaneously and spread propagation on the control channel of Sub-ring2.According to above-mentioned scheme, the M3 node can be received Flush message (the non-fault choke point of the subring control channel of no tunnel is not blocked protocol massages) from its two ring upper port, thereby M3 node refresh address is transmitted.And for S6, M1, M5 and S3 node, because they can only receive the Flush message from a port, therefore can not transmit by refresh address.Obviously, in Fig. 6 b, the ring of subring Sub-ring3 refreshes outward the address of node of (the e port of node M 3 blocks this path) on the unobstructed path not to be transmitted and refreshes.

Summary of the invention

The present invention seeks to provides a kind of address flush method based on multi-layer subring at the defective that exists in the background technology, improves network performance.

The present invention adopts following technical scheme for achieving the above object:

A kind of address flush method based on multi-layer subring of the present invention is as follows:

Interconnecting nodes after receiving on the port of upper strata subring that nearest protocol massages and refresh address are transmitted, the port numbers of this protocol massages port of interconnecting nodes recorder;

When if lower floor's subring of this interconnecting nodes need send the Flush message because of topology changes on the control channel of the upper strata of this contact that interconnects subring, described interconnecting nodes only sends the Flush message to the port of another upper strata subring that is not recorded;

After if the ring upper port of node is received the Flush message in the described subring, check whether this port starts refresh timer FT (Flush Timer): if be not activated FT, described port starts the FT timer, and the Net_ID in the Flush message and PID are recorded in respectively in variable 1 and the variable 2 promptly: variable 1 ← Net_ID; Variable 2 ← PID; If started the FT timer and the FT timer does not have overtime, Net_ID in the value of variable 1 and the Flush message carries out XOR so, PID in the value of variable 2 and the Flush message carries out XOR, then the result is deposited in respectively in variable 1 and the variable 2 promptly:

Behind the FT timer expiry in the port, if the value of variable 1 be 0 and the value of variable 2 be not 11, the node refresh address that has this port is transmitted; If the value of variable 1 is not 0, the node refresh address that has this port is transmitted;

Wherein:

Net_ID be trigger interconnecting nodes structure Flush message subring control channel be designated special multicast mac address or control VLAN ID;

PID has 2 bits, and the port of interconnecting nodes in subring that indicates structure Flush message is east port or western port: 10 are east port, and 01 is western port.

Preferably, the method for described port numbering is as follows: be benchmark with the counter clockwise direction, the port that enters interconnecting nodes is east port, and the port that leaves interconnecting nodes is western port.

Beneficial effect:

Method of the present invention can solve the address flush problem of ITU-T multi-layer subring G.8032v2, thereby guarantees that the ether multi-ring network can restrain in 50ms behind the link switchover, has improved network performance greatly.

Description of drawings

Fig. 1 is the topological diagram of ether multi-ring network;

Fig. 2 a is the communication path topology figure when link is intact in the ether multi-ring network;

Fig. 2 b is the communication path topology figure during link failure in the ether multi-ring network;

Fig. 3 a is the schematic diagram of the subring control VLAN of no tunnel;

Fig. 3 b is the schematic diagram that the subring control VLAN of tunnel is arranged;

Fig. 4 breaks down for subring need send the instance graph of refresh address message to environment-development;

Fig. 5 a is the schematic diagram of individual layer subring;

Fig. 5 b is that the diffusion of Flush message in ring refreshes in the traditional scheme;

Fig. 6 a is the schematic diagram of multi-layer subring;

Fig. 6 b is the problem that prioritization scheme runs in multi-layer subring;

Fig. 7 is a particular flow sheet of the invention process;

Fig. 8 is embodiments of the invention 1;

Fig. 9 is embodiments of the invention 2.

Embodiment

A kind of address flush method of the present invention based on multi-layer subring, interconnecting nodes is received nearest protocol massages (as: SF on the port of upper strata subring, NR (RB) MS and FS) and after refresh address transmits, the port numbers of this protocol massages port of interconnecting nodes recorder.

When if lower floor's subring of this interconnecting nodes need send the Flush message because of topology changes on the control channel of the upper strata of this contact that interconnects subring, described interconnecting nodes only sends the Flush message to the port of another upper strata subring that is not recorded.

If after the ring upper port of node is received the Flush message in the described subring, check whether this port starts refresh timer (FT:Flush Timer).If be not activated FT, described port starts the FT timer, and the Net_ID in the Flush message and PID is recorded in variable 1 and the variable 2 (that is: variable 1 ← Net_ID respectively; Variable 2 ← PID).If started the FT timer and the FT timer does not have overtime, Net_ID in the value of variable 1 and the Flush message carries out XOR so, PID in the value of variable 2 and the Flush message carries out XOR, then the result is deposited in respectively in variable 1 and the variable 2 (that is:

).

Behind the FT timer expiry in the port, if the value of variable 1 be 0 and the value of variable 2 be not 11, the node refresh address that has this port is transmitted.If the value of variable 1 is not 0, the node refresh address that has this port is transmitted.

It needs to be noted:

1.Net_ID be the sign (special multicast mac address or control VLAN_ID) of control channel that triggers the subring of interconnecting nodes structure Flush message.

2.PID 2 bits are arranged, and the port of interconnecting nodes in subring that indicates structure Flush message is east port or western port.10: east port, 01: western port (criterion of port numbering: with the counter clockwise direction is benchmark, and the port that enters interconnecting nodes is east port, and the port that leaves interconnecting nodes is western port).

Concrete steps of the present invention are described below (as shown in Figure 7):

Step 700, the node that has port block in the subring periodically sends protocol frame along the subring upper port.

Step 701, after the node in the subring was received protocol massages, handle in two kinds of situation: situation 1. nodes were interconnecting nodes of lower floor's subring, change step 702 over to.Situation 2. nodes are not the interconnecting nodes of lower floor's subring, change step 707 over to.

Step 702, whether protocol massages is the Flush message.If change step 708 over to, otherwise change 703 over to.

Step 703 judges whether it is up-to-date address flush message, if change step 704 over to

Step 704, the interconnecting nodes refresh address is transmitted, and the port of this protocol massages of recorder (purpose is meant choke point on the pine torch ring is in which side of this interconnecting nodes), the port numbers of the original record of deletion.

Step 705, this interconnecting nodes judge whether lower floor's subring change in topology takes place.If change in topology is arranged, change step 706 over to.

Step 706, the Flush message that the interconnecting nodes structure is new sends the Flush message then on the control channel of described subring, and the direction of transmission is to send (that is, deviate from the choke point and send the Flush message) along Unrecorded another ring upper port of this interconnecting nodes.

Step 707 for situation 2, judges whether protocol massages is the Flush message, if change 708 over to.

Step 708 checks to receive whether the port of this protocol massages starts refresh timer FT, changes step 709 over to if start, otherwise changes step 710 over to.

Step 709,

Step 710 starts refresh timer FT, variable 1 ← Net_ID, variable 2 ← PID;

Step 711 judges whether the FT timer is overtime, if the overtime step 712 that changes over to.

Step 712, whether the value of decision variable 1 is 0.If be not 0 to change step 713 over to, otherwise change step 714 over to.

Step 713, the node refresh address that has described FT timer is transmitted.

Step 714, whether the value of decision variable 2 is 11.11 do not change step 715 over to if be not.

Step 715, the node refresh address that has described FT timer is transmitted.

In order to further specify the solution of the present invention, we further set forth method of the present invention by following specific embodiment.Below embodiment is described in detail, but not as a limitation of the invention.

Embodiment one, between by lower floor's subring levels subring interconnecting nodes clog-free point:

Fig. 8 be the present invention program at the utilization schematic diagram that refreshes under the passage non-failure conditions, the passage that refreshes of the subring among the figure is realized by VLAN, makes a concrete analysis of as follows:

As shown in Figure 8, Ring1, Sub-ring2 and Sub-ring3 constitute the ether multi-ring network.Ring1 is a complete closed loop, and the node that comprises has S1～S6, and the link between them.Sub-ring2 is a subring, and the node that comprises has S6, M1～S5 and S3, and the link that comprises has:＜S6, M1 〉,＜M1, M2 〉,＜M2, M3 〉,＜M3, M4〉and＜M5, S3 〉.The Control Node of Ring1 is S1, and what link to each other with the w port is the ring protection link.The Control Node of Sub-ring2 is M1, and what link to each other with the e port is the ring protection link.The node that Sub-ring3 comprises has M2, N1～N3 and M4, and the link that comprises has＜M1, N1 〉,＜N1, N2 〉,＜N2, N3〉and＜N3, M4 〉.N1 is the Control Node of Sub-ring3.

The Control Node M1 of Sub-ring2 periodically sends NR (RB) message along the ring upper port, and after interconnecting nodes M2 and M4 received NR (RB) message, the port of this message received in record respectively.

When Sub-ring3＜N2, N3〉fault taken place in link, node N2 and N3 block the forwarding capability of the data message of the port adjacent with faulty link, refresh addresses forwarding table separately, and outwards periodically send SF message (warning message) along their intact ring upper port separately.When M2 (M4) receives the SF protocol massages first, refresh address is transmitted, construct Flush1 (Flush2) message simultaneously and propagate on the control channel of Sub-ring2, the direction of propagation all is to propagate (that is the direction that, deviates from the choke point is propagated) along their e ports on Sub-ring2.The Flush1 message comprises＜Sub-ring3_ID, and 10〉information, the Flush2 message comprises＜Sub-ring3_ID, and 01〉information.

When M3, M4, M5 and S3 node are received Flush1 or Flush2 message first, receiving startup timer FT on the ring upper port of Flush message, and doing variable 1 ← Net_ID and variable 2 ← PID computing.When timer in operation and when receiving the Flush message,

Computing.Behind the FT of M3 and M4 timer expiry, the value of their variable 1 is not 0, and node M 3 and M4 refresh address are transmitted.Behind the FT of M5 and S3 timer expiry, the value of their variable 1 be 0 and the value of variable 2 be 11, node M 5 and S3 not refresh address transmit.

Therefore, the path that the address is refreshed outside the ring of Sub-ring3 is: M2 ← → M3 ← → M4.

By last surface analysis as can be known, the present invention is under this scene, satisfy the requirement of " ring of subring refreshes the address of node that needs only refresh (at data and message transmission) on unobstructed path between the subring interconnecting nodes outward and transmits ", do not caused the phenomenon that refreshes diffusion.

Embodiment two, between by lower floor's subring levels subring interconnecting nodes, the choke point is arranged

Shown in Fig. 9 a, the M3 node of Sub-ring2 starts manual switch (MS) at its e port, and this node periodically outwards sends the MS message along its two ring upper port, and Control Node M1 opens the data forwarding function of ring protection link.After a period of time, the ether multi-ring network enters stable state, and after interconnecting nodes M2 received MS, the port numbers e of MS protocol massages received in record, and after interconnecting nodes M4 received MS, the port numbers w of MS protocol massages received in record.Sub-ring3＜N2, N3〉fault taken place in link over time, node N2 and N3 block the forwarding capability of the data message of the port adjacent with faulty link, refresh addresses forwarding table separately, and outwards periodically send SF message (warning message) along their intact ring upper port separately.When M2 and M4 receive the SF protocol massages first, refresh address is transmitted, construct the Flush message simultaneously and propagate on the control channel of Sub-ring2, that is: node M 2 outwards sends Flush 1 message along the w port, and node M 4 outwards sends Flush 2 messages along the e port.The Flush1 message comprises＜Sub-ring3_ID, and 10〉information, the Flush2 message comprises＜Sub-ring3_ID, and 01〉information.

When node S6, M1, M5 and the S3 on the Sub-ring2 receives the Flush message first, receiving startup timer FT on the ring upper port of Flush message, and doing variable 1 ← Net_ID and variable 2 ← PID computing.When timer in operation and when receiving the Flush message, do Computing.Behind the FT of node S6, M1, M5 and S3 timer expiry, the value of their variable 1 is not 0, and node S6, M1, M5 and S3 refresh address are transmitted.

Wherein, after S6 and S3 refresh address are transmitted, also can produce new Flush protocol massages on the control channel of ring Ring1, that is: node S6 outwards sends the Flush3 message along w and e port, and node S3 outwards sends the Flush4 message along w and e port.According to traditional prioritization scheme, node S5 on the Ring1 and S4 can receive the Flush message from their two ports, and node S5 and S4 can refresh their addresses forwarding table.

By last surface analysis as can be known: the outer address of the ring of Sub-ring3 refresh path is: M2 ← → M1 ← → S6 ← → S5 ← → S4 ← → S3 ← → M5 ← → M4.Therefore, the present invention is under this scene, satisfy the requirement of " ring of subring refreshes the address of node that needs only refresh (at data and message transmission) on unobstructed path between the subring interconnecting nodes outward and transmits ", do not caused the phenomenon that refreshes diffusion.

Certainly; the present invention also can have other various embodiments; under the situation that does not deviate from spirit of the present invention and essence thereof; those of ordinary skill in the art can make various corresponding changes and distortion according to the present invention, but these corresponding changes and distortion all should belong to the protection range of the appended claim of the present invention.

Claims

1. address flush method based on multi-layer subring is characterized in that described method is as follows:

Wherein:

Net_ID be trigger interconnecting nodes structure Flush message subring control channel be designated special multicast mac address or control VLANID;

2. a kind of address flush method according to claim 1 based on multi-layer subring, the method that it is characterized in that described port numbering is as follows: be benchmark with the counter clockwise direction, the port that enters interconnecting nodes is east port, and the port that leaves interconnecting nodes is western port.