CN101656664B - Ethernet ring network system and border node thereof and Ethernet ring network realization method - Google Patents

Ethernet ring network system and border node thereof and Ethernet ring network realization method Download PDF

Info

Publication number
CN101656664B
CN101656664B CN2009100928230A CN200910092823A CN101656664B CN 101656664 B CN101656664 B CN 101656664B CN 2009100928230 A CN2009100928230 A CN 2009100928230A CN 200910092823 A CN200910092823 A CN 200910092823A CN 101656664 B CN101656664 B CN 101656664B
Authority
CN
China
Prior art keywords
subring
port
message
hello message
fault timer
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN2009100928230A
Other languages
Chinese (zh)
Other versions
CN101656664A (en
Inventor
杨碧辉
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
New H3C Technologies Co Ltd
Original Assignee
Hangzhou H3C Technologies Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hangzhou H3C Technologies Co Ltd filed Critical Hangzhou H3C Technologies Co Ltd
Priority to CN2009100928230A priority Critical patent/CN101656664B/en
Publication of CN101656664A publication Critical patent/CN101656664A/en
Application granted granted Critical
Publication of CN101656664B publication Critical patent/CN101656664B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Small-Scale Networks (AREA)

Abstract

The invention discloses an Ethernet ring network system and a border node thereof and an Ethernet ring network realization method; in the invention, the concept of an auxiliary border node and an SRPT detection mechanism is cancelled, and the concept of sub-ring priority is introduced. A HELLO message transmitted by a host node of a sub ring with low priority can be transmitted in a sub ring with high priority, and when the transmission is carried out in the sub ring with high priority, the HELLO message transmitted by the host node of the sub ring with low priority is regarded as data message by the sub ring with high priority, and the HELLO message is blocked or transmitted according to the data message. The HELLO message transmitted by the host node of the sub ring with high priority can not be transmitted in the sub ring with low priority. By adopting the method, a plurality of the sub rings are supported to be accessed to the host node from different positions, thereby improving the networking flexibility and ensuring the maximum communication of the Ethernet ring network to the maximum extent.

Description

Ethernet ring network system and fringe node thereof and ethernet ring network implementation method
Technical field
The present invention relates to the ethernet ring network technology, be specifically related to a kind of Ethernet ring network system, a kind of fringe node and a kind of Ethernet ring network realization method that is applied to this Ethernet ring network system.
Background technology
Fig. 1 is a kind of networking sketch map of typical ethernet ring network.Ethernet ring network comprises main ring and subring, and subring can be for one or more.When Ethernet ring is complete, data loopback may occur, thereby cause broadcast storm.For fear of the generation of this situation, each company has all released the agreement of avoiding the ethernet ring network data loopback at present.For example, rapid ring protection protocol (RRPP, Rapid Ring ProtectionProtocol).
When the device start RRPP agreement in the ethernet ring network as shown in Figure 1, Fig. 1 is a kind of RRPP ring net system, and this RRPP ring net system comprises two crossing rings, and the equipment of S1~S4 is formed main ring, and the equipment of S3~S6 is formed subring.Each ring is gone up must have a host node, and can only have one, and host node is the promoter of the active detecting mechanism of looped network state, also is to detect the policymaker who encircles the executable operations after the fault.Every other node on the ring except that host node is a transmission node.In two ports of host node and transmission node access ring, one is master port, and another is secondary port.As shown in Figure 1, the host node of main ring is S1, and the host node of subring is S6, and S2, S3 and S4 are the transmission nodes of main ring, and S5 is the transmission node of subring.
Two intersection points that main ring and subring are intersected are called Edge node and assistant edge node respectively, and Edge node and assistant edge node needs configuration in pairs in the prior art.As shown in Figure 1, S4 is a fringe node, and S3 is the assistant edge node.The port that Edge node and assistant edge node inserts subring is called edge port, and two ports that insert main ring are called public port, on the fringe node on public port and the assistant edge node link between the public port be called as common link.As shown in Figure 1, port one that S3 is last and port 2 are public port, and port 3 is an edge port, and the link between S3 and the S4 is a common link.
Host node active detecting looped network state in the ring.Specifically; Host node cycle property ground sends (HELLO) message of shaking hands from its master port; This HELLO message is propagated through the transmission node on the main ring successively; When the secondary port of host node is received the HELLO message of host node oneself transmission in (this stipulated time is by the definition of Fail timer) at the appointed time, just think that host node is in good working condition (CompleteState), promptly the looped network state is complete.For the data message on the anti-stop ring forms the broadcasting loop, the secondary port of host node blocks its secondary port when receiving the HELLO message that oneself sends at the appointed time, and this obstruction is meant the obstruction to the data message.As shown in Figure 1, the symbol among the figure on the device port "
Figure G2009100928230D00021
" is represented this port get clogged (BLOCK).After host node blocked secondary port, network topology changed, and for fear of the directed mistake of message, host node also need notify all transmission nodes on the ring to refresh MAC address entries and ARP/ND list item.
If the secondary port of host node is not received hello packet at the appointed time, just thinking has link to be in malfunction on the looped network, and at this moment, host node is decontroled the obstruction of secondary port to the data message, does not interrupt to guarantee the communication on the looped network.Behind link failure recovery, host node is received the HELLO message from secondary port once more, thinks that then looped network returns to good working condition, and block secondary port this moment again, prevents to form data loopback.Main ring host node and subring host node blocked the sketch map of its secondary port when Fig. 1 showed the equal fault-free of main ring and subring.
Node in the main ring always at the port Link State of monitoring oneself, will take immediate steps in case find port DOWN, and the mechanism of this policing port Link State is called quick LINK DOWN informing mechanism.Specifically, after the master port of host node in the main ring detected port DOWN, the direct perception link failure of host node was decontroled secondary port immediately.When master port on the transmission node in the main ring or secondary port generation link DOWN, this transmission node will be that secondary port or the master port of UP sends LINK-DOWN message notifying host node from the state of non-working port pairing.After host node is received the LINK-DOWN message, decontrol secondary port.Compare with the mechanism that adopts HELLO message detection ring net state, the LINKDOWN informing mechanism can perceive link failure apace.Adopt the mechanism of HELLO message detection ring net state to be applied to main ring and subring, LINK DOWN informing mechanism only is applied to main ring.
For the situation of many main rings, prior art has also been introduced subring protocol massages passage (SRPT) checking mechanism.SRPT just is meant the passage of subring protocol massages in main ring.The RRPP agreement is regarded main ring as a logical node of subring in theory; The protocol massages of subring can pass through the main ring transparent transmission; Be that the subring protocol massages can be broadcast in the main ring by Edge node and assistant edge node; As shown in Figure 1, fringe node S3 receives the subring protocol massages that S6 sends, and it is broadcasted away through port one and port 2.Except that the following EDGE-HELLO that makes referrals to, main ring is regarded as data message with the subring protocol massages, and the subring protocol massages can not be through the port that blocks.
Each subring has 2 SPRT, in Fig. 1, is respectively S3-S4 and S3-S2-S1-S2.When main ring was complete, the secondary port of its host node was in blocked state, has only S3-S4 to lead to.When the main ring fault, if fault occurs on the S3-S2-S1-S2, then S3-S4 leads to, and S3-S2-S1-S2 leads to if fault occurs on the S3-S4.Therefore, at any time, among 2 SRPT of subring, have only one to lead to, subring host node S6 can keep the obstruction of secondary port so, has so just avoided the formation data loopback.If 2 SRPT of subring all interrupt; Then the subring host node can not receive the HELLO message that oneself sends; So the Fail timer expiry, the subring host node is decontroled secondary port, and the S4 in the subring, S5 and S6 still can communicate like this; Thereby make subring can obtain maximum communication path, and can not cause loop.
But in reality, adopting more dual-homed ethernet ring network shown in Fig. 2, two dual-homed sub-loop are subring 1 and subring 2 interconnects by Edge node and assistant edge node, itself have just formed a loop.In 2 SRPT of subring on the main ring are whole, have no progeny, the secondary port of the host node of 2 sub-loop is all decontroled, and certainly will form data loopback (data message moves towards shown in double-head arrow, forms figure of eight data loopback) between the subring.
In order to eliminate this defective; Introduced above-mentioned SRPT status checkout mechanism; This SRPT status checkout mechanism is cooperated by Edge node and assistant edge node to be accomplished, and its objective is when fringe node detects among the SRPT and has no progeny, before the secondary port of two sub-loop host nodes is all decontroled; Block the edge port of two sub-loop fringe nodes, thereby avoid forming between subring data loopback.
Specifically, fringe node is the promoter and the policymaker of SRPT status checkout mechanism, the listener of assistant edge node channel status, and be responsible for when channel status changes, in time notifying fringe node.Two ports of fringe node through being connected into SRPT periodically send edge (EDGE-HELLO) message of shaking hands in main ring, this EDGE-HELLO message mails to the assistant edge node through each node on the main ring successively.If the assistant edge node can be received the EDGE-HELLO message at the appointed time; Show that to have 1 SRPT at least normal, the subring message can normally pass through, otherwise the assistant edge node can not receive EDGE-HELLO; Explain that 2 SRPT all interrupt, the subring message can't pass through.When detecting 2 SRPT and all interrupt, the assistant edge node sends main ring fault (MAJOR-FAULT) message from edge port through subring chain road direction fringe node immediately.After fringe node is received MAJOR-FAULT, block the edge port of oneself.
With dual-homing networking shown in Figure 2 is example; Behind two link failures shown in Figure 2, the secondary port of S5 and S6 is decontroled, and assistant edge node S3 does not receive the EDGE-HELLO message that fringe node S4 sends; Then S3 is that port 3 sends the MAJOR-FAULT message with port 4 through its edge port; Because S5 and S6 have decontroled secondary port, the MAJOR-FAULT message that therefore sends from the port 3 of S3 arrives the port 4 of fringe node through S6, and the MAJOR-FAULT message that sends from the port 4 of S3 is through the port 3 of S5 arrival fringe node; This moment, fringe node S3 blocked its port 3 and port 4, and the result is as shown in Figure 3 for obstruction.As can beappreciated from fig. 3, behind the port block, S5 and S6 still can communicate with S3, but can not communicate with S4, can not guarantee the largest connected property of looped network.
Equally, for the networking structure shown in Fig. 1, behind S3-S4 and two link failures of S1-S2; S6 decontrols secondary port, and assistant edge node S3 do not receive the EDGE-HELLO message that fringe node S4 sends, and then S3 is that port 3 sends the MAJOR-FAULT messages through its edge port; Because S6 has decontroled secondary port; Therefore S4 can receive the MAJOR-FAULT message, and this moment, fringe node S3 blocked its port 3, and the result is as shown in Figure 4 for obstruction.As can beappreciated from fig. 4, behind the port block, S6 can communicate with S5 and S4, but can not communicate with S3, can not guarantee the largest connected property of looped network.
It is thus clear that existing SRPT status checkout mechanism can not guarantee the largest connected property of looped network.
In addition; SRPT status checkout mechanism realizes through the Edge node and assistant edge node literary composition of transmitting messages mutually; Therefore must introduce the notion of Edge node and assistant edge node; And on the ring fringe node and an assistant edge node can only be arranged, otherwise the confusion that message sends will occur, thereby can't realize SRPT status checkout mechanism.Therefore existing ethernet ring network structure only supports subring to insert main ring from identical Edge node and assistant edge node, and in this case, networking as shown in Figure 5 is with unavailable.
Summary of the invention
In view of this, the invention provides a kind of Ethernet ring network system, can support a plurality of subrings to insert main ring,, and can not cause data loopback with raising networking flexibility property from diverse location.
This system comprises main ring and the one or more subring of being made up of node; Main ring includes a host node with each subring; Each host node is used for sending the HELLO message of shaking hands from the master port of self, receive the HELLO message that self sends at secondary port from self after, block secondary port; Each node in the main ring receives the HELLO message with institute and is regarded as data message when the HELLO message that receives from subring, confirms to transmit or block the HELLO message that is received according to self port status;
Different subrings insert main ring from identical or different node, and all nodes that subring and main ring intersect are called fringe node; Different subrings are configured to have different priority;
Said fringe node is used for when receiving the HELLO message that is derived from subring x, and reception HELLO message is forwarded to main ring and priority is not less than in the subring of subring x;
The host node of each subring in the system receives the HELLO message with institute and is regarded as data message when the HELLO message that receives from other subrings, confirms to transmit or block the HELLO message that is received according to self port status.
Preferably; Host node in each subring is further used for, and when not receiving the HELLO message that oneself sends at the fixed time in the t1, decontrols the obstruction of secondary port to non-book loop protocol message earlier; After waiting for one section setting-up time t2, decontrol the obstruction of secondary port again to the data message.
Preferably, the host node in each subring comprises control unit, control domain fault timer and data field fault timer;
The timing of said data field fault timer is greater than the timing of said control domain fault timer, and the timing of data field fault timer and control domain fault timer is all greater than transmission cycle of HELLO message;
Said control unit is used for after place subring host node sends the HELLO message, controls said control domain fault timer and said data field fault timer picks up counting; When said control domain fault timer expiry, decontrol the obstruction of the secondary port of place subring host node to non-book loop protocol message; When said data field fault timer expiry, decontrol the obstruction of the secondary port of place subring host node to the data message; When subring host node in place receives the HELLO message that oneself sends, cancel the timing of said control domain fault timer and said data field fault timer.
Preferably, the host node of main ring comprises control unit, data field fault timer and control domain fault timer;
The timing of said data field fault timer is greater than the timing of said control domain fault timer, and the timing of data field fault timer and control domain fault timer is all greater than transmission cycle of HELLO message;
Said control unit is used for when place main ring host node sends the HELLO message, controls said control domain fault timer and said data field fault timer picks up counting; When said control domain fault timer expiry, decontrol the obstruction of the secondary port antithetical phrase loop protocol message of place main ring host node; When said data field fault timer expiry, decontrol the obstruction of the secondary port of place main ring host node to the data message; When place main ring host node receives the HELLO message that oneself sends, cancel the timing of said control domain fault timer and said data field fault timer.
Preferably, each node of main ring is further used for, and when interim blocked state is kept scheduled time t1, decontrols earlier the obstruction of interim port blocked antithetical phrase loop protocol message, wait for one section setting-up time t2 after, decontrol the obstruction of interim port blocked again to the data message.
Preferably, the node in the main ring comprises control unit, control domain fault timer and data field fault timer;
The timing of said data field fault timer is greater than the timing of said control domain fault timer, and the timing of control domain fault timer and data field fault timer is all greater than transmission cycle of HELLO message;
Said control unit is used for when oneself state is moved to interim blocked state, controls said control domain fault timer and data field fault timer picks up counting; When said control domain fault timer expiry, decontrol the obstruction of the interim port blocked antithetical phrase loop protocol message of place node; When said data field fault timer expiry, decontrol the obstruction of the interim port blocked of place node to the data message.
Preferably, the timing of said data field fault timer is 4 times of the HELLO message transmission cycle; The timing of said control domain fault timer sends 2 times of cycle for the HELLO message.
The present invention also provides the fringe node equipment of one such application in Ethernet ring network system, can support a plurality of subrings to insert main ring from diverse location, with raising networking flexibility property, and can not cause data loopback.
This fringe node comprises message receiving element and message retransmission unit;
Said message receiving element is used for sending to the message retransmission unit with being received from the outside HELLO message of node;
Said message retransmission unit is used for when receiving the HELLO message that is derived from subring x, and the main ring and the priority that the HELLO message that is received are forwarded to the place system are not less than in the subring of subring x.
Preferably, said message retransmission unit comprises respective modules and forwarding module;
Said respective modules is used to preserve the source ring information of setting and the corresponding relation of outbound port; Source ring information is main ring, and then corresponding output port is for connecting the public port of main ring; Source ring is subring, and then corresponding output port comprises public port and the particular edge port that connects main ring, and said particular edge port is to be not less than the edge port that the subring that encircles in the source links to each other with priority;
Said forwarding module; Be used for after receiving the HELLO message; The source ring information that employing is confirmed according to reception HELLO message is mated in said respective modules, finds the outbound port of the HELLO message that receives, and transmits the HELLO message that is received from the outbound port that finds.
The present invention provides a kind of implementation method of Ethernet ring network system again, can support a plurality of subrings to insert main ring from diverse location, with raising networking flexibility property, and can not cause data loopback.
This method is applied to comprise in the Ethernet ring network system of main ring and one or more subrings; It is characterized in that different subrings insert main ring from identical or different node, all nodes that subring and main ring intersect are called fringe node;
This method also comprises: be different subring configuration different priorities;
Host node in main ring and each subring sends the HELLO message from master port, receive the handshake message that self sends at secondary port after, block secondary port;
Each node in the main ring receives the HELLO message with institute and is regarded as data message when the HELLO message that receives from subring, confirms to transmit or block the HELLO message that is received according to self port status;
Fringe node is forwarded to main ring and priority is not less than in the subring of subring x with the HELLO message that is received when receiving the HELLO message that is derived from subring x;
The host node of each subring confirms to transmit or block the HELLO message that is received according to host node self port status of this subring in the system when the HELLO message that receives from other subrings.
Preferably, this method further comprises: the corresponding relation that source ring information and outbound port are set in advance; Source ring information is main ring, and then corresponding output port is for connecting the public port of main ring; Source ring is subring, and then corresponding output port comprises public port and the particular edge port that connects main ring, and said particular edge port is to be not less than the edge port that the subring that encircles in the source links to each other with priority;
Saidly the HELLO message that is received is forwarded to main ring and priority is not less than in the subring of subring x and comprises: fringe node is after receiving the HELLO message; The source ring information that employing is confirmed according to reception HELLO message is mated in said corresponding relation; Find the outbound port of the HELLO message that receives, and transmit the HELLO message that is received from the outbound port that finds.
Preferably; This method further comprises: when the host node in each subring is not received the HELLO message that oneself sends in the t1 at the fixed time; Decontrol earlier the obstruction of secondary port to non-book loop protocol message, wait for one section setting-up time t2 after, decontrol the obstruction of secondary port again to the data message.
Preferably, the said obstruction of secondary port of decontroling earlier to non-book loop protocol message, wait for one section setting-up time t2 after, decontrol the obstruction of secondary port again to the data message, comprising:
Host node Sx when subring 1After sending the HELLO message, control a control domain fault timer and a data field fault timer picks up counting;
When said control domain fault timer expiry, decontrol host node Sx 1Secondary port to the obstruction of non-book loop protocol message; When said data field fault timer expiry, decontrol host node Sx 1Secondary port to the obstruction of data message;
As host node Sx 1When receiving the HELLO message that oneself sends, cancel the timing of said control domain fault timer and said data field fault timer;
The timing of said data field fault timer is greater than the timing of said control domain fault timer, and the timing of data field fault timer and control domain fault timer is all greater than transmission cycle of HELLO message.
Preferably, this method further comprises:
Host node Sx when main ring 2When sending the HELLO message, control a control domain fault timer and a data field fault timer picks up counting;
When said control domain fault timer expiry, decontrol host node Sx 2The obstruction of secondary port antithetical phrase loop protocol message; When said data field fault timer expiry, decontrol host node Sx 2Secondary port to the obstruction of data message;
As host node Sx 2When receiving the HELLO message that oneself sends, cancel the timing of said control domain fault timer and said data field fault timer;
The timing of said data field fault timer is greater than the timing of said control domain fault timer, and the timing of data field fault timer and control domain fault timer is all greater than transmission cycle of HELLO message.
Preferably; This method further comprises: when each node of main ring is kept scheduled time t1 at interim blocked state; Decontrol earlier the obstruction of interim port blocked antithetical phrase loop protocol message, wait for one section setting-up time t2 after, decontrol the obstruction of interim port blocked again to the data message.
Preferably, this method further comprises: the said obstruction of decontroling earlier interim port blocked antithetical phrase loop protocol message, wait for one section setting-up time t2 after, decontrol the obstruction of interim port blocked again to the data message, comprising:
The node Sx of main ring 3When oneself state is moved to interim blocked state, control a control domain fault timer and a data field fault timer picks up counting;
When said control domain fault timer expiry, decontrol node Sx 3The obstruction of interim port blocked antithetical phrase loop protocol message; When said data field fault timer expiry, decontrol node Sx 3Interim port blocked to the obstruction of data message;
The timing of said data field fault timer is greater than the timing of said control domain fault timer, and the timing of control domain fault timer and data field fault timer is all greater than transmission cycle of HELLO message.
Preferably, the timing of said data field fault timer is 4 times of the HELLO message transmission cycle; The timing of said control domain fault timer sends 2 times of cycle for the HELLO message.
Visible according to above technical scheme; Visible by above scheme; Therefore the present invention when the many subrings of needs insert, can select access node flexibly owing to cancelled the notion of assistant edge node; Rather than must fixingly insert from unique Edge node and assistant edge node, therefore can realize many subrings structure as shown in Figure 5.
And; The present invention is provided with subring priority; And allow the HELLO message in the low priority subring in the high priority subring, to transmit, thus the subring HELLO message forwarding path that need not pass through main ring increased, and subring HELLO message forwarding path must pass through main ring in the prior art.When link failure appears in main ring; When double-link failure (being a common link fails on the main ring, a non-common link fails) especially occurring, the subring HELLO message forwarding path of process main ring all is blocked; But the subring HELLO message forwarding path that need not pass through main ring that the present invention increases is not affected; The complete of HELLO message forwarding path makes corresponding subring host node can block secondary port, data loopback can not occur, guarantees the normal operation of looped network.
Secondly; The present invention's protocol massages and data message to non-subring on the subring host node distinguished; When not receiving the HELLO message that oneself sends in the given time, decontrol earlier the obstruction of secondary port to non-book loop protocol message, wait for one section setting-up time t2 after; Decontrol the obstruction of secondary port again to the data message, thus the moment data loopback when having avoided the main ring double-link failure.
In addition; The present invention's protocol massages and data message to subring on each node of main ring distinguished; When the interim blocked state of node is kept the scheduled time, decontrol earlier the obstruction of interim port blocked antithetical phrase loop protocol message, wait for one section setting-up time after; Decontrol the obstruction of interim port blocked again to the data message, thus the moment data loopback when having avoided the main ring double-link failure to recover.
Description of drawings
Fig. 1 is the networking sketch map of a kind of typical ethernet ring network in the prior art.
Fig. 2 is the networking sketch map of dual-homed ethernet ring network in the prior art.
Fig. 3 is in the prior art behind the dual-homed ethernet ring network double-link failure, carries out the port block sketch map after the SRPT status checkout mechanism.
Fig. 4 is in the prior art behind the list ring ethernet ring network double-link failure, carries out the port block sketch map after the SRPT status checkout mechanism.
Fig. 5 is a kind of a plurality of subring inserts main ring through different nodes an ethernet ring network sketch map.
The ethernet ring network sketch map that Fig. 6 provides for the embodiment of the invention.
Fig. 7 is the sketch map of fringe node in the embodiment of the invention.
Fig. 8 is the HELLO message forwarding path sketch map of main ring in the embodiment of the invention one.
Fig. 9 is the HELLO message forwarding path sketch map of subring 1 in the embodiment of the invention one.
Figure 10 is the HELLO message forwarding path sketch map of subring 2 in the embodiment of the invention one.
Figure 11 is the HELLO message forwarding path sketch map of the embodiment of the invention one double center chain road fault.
Figure 12 is the HELLO message forwarding path sketch map of subring 1 in the embodiment of the invention two.
Figure 13 is the HELLO message forwarding path sketch map of subring 2 in the embodiment of the invention two.
Figure 14 is the sketch map after the common link S1-S2 fault in the embodiment of the invention two.
Figure 15 is the sketch map after the common link S3-S4 fault in the embodiment of the invention two.
Figure 16 is the sketch map after common link S1-S2 and the S3-S4 fault in the embodiment of the invention two.
Figure 17 is the sketch map after the non-common link S1-S4 fault in the embodiment of the invention two.
Figure 18 is the sketch map after the embodiment of the invention two double center chain road faults.
Embodiment
Below in conjunction with the accompanying drawing embodiment that develops simultaneously, describe the present invention.
The present invention is a kind of implementation of ethernet ring network, and is as shown in Figure 6, and this scheme has been cancelled the notion of assistant edge node, owing to do not have the assistant edge node in the system, therefore also cancellation thereupon of SRPT testing mechanism of the prior art.Specifically, Ethernet ring network system of the present invention comprises main ring and one or more subring; A plurality of subrings insert main ring from identical or different node, and all nodes that subring and main ring intersect all are called fringe node; The port that fringe node inserts subring is an edge port, and two ports that insert main ring are public port.
The present invention has also introduced the notion of subring priority.The purposes of subring priority is; The HELLO message that the host node of low priority subring sends can be transmitted in the high priority subring; And when in the high priority subring, transmitting; The HELLO message that the host node of low priority subring sends is regarded as data message by the high priority subring, blocks or transmits according to data message.And the HELLO message that the host node of high priority subring sends can not be transmitted in the subring of low priority.In addition, identical with prior art is that the HELLO message that the host node of subring sends also will be transmitted in main ring.
Under the situation of introducing subring priority, the fringe node in the Ethernet ring network system of the present invention when receiving the HELLO message that is derived from subring x, is forwarded to main ring and priority is not less than in the subring of subring x with the HELLO message that is received.Wherein, the subring that is not less than subring x comprises subring x itself.
Host node in the subring receives the HELLO message with institute and is regarded as data message when the HELLO message that receives from other subrings, confirms to transmit or block the HELLO message that is received according to host node self port status of this subring.
Identical with prior art is; Host node in main ring and the subring also can be carried out the action of active detecting looped network state; Being each host node sends the HELLO message of shaking hands from self master port; Block secondary port receive the HELLO message that oneself sends at secondary port after, thereby cut off data loopback from self.And each node on the main ring will be regarded as data message from the HELLO message of subring, confirm to transmit or block the HELLO message that is received according to self port status.
In the present invention, can come the presentation protocol message source from which subring through precedence field is set in protocol massages.For example, corresponding RRPP agreement has reserved field in the field of RRPP protocol massages, can choose a reserved field wantonly and be defined as precedence field.Preferably, for fear of increasing field, also can utilize ring sign (Ring ID) field in the existing RRPP protocol massages to identify subring priority.For example, the Ring ID of subring 1 is 1, and then its priority is 1, and the Ring ID of subring 2 is 2, and then its priority is 2.The more little expression priority of numeral is big more, and perhaps the big more expression priority of numeral is big more, and numeral can be provided with the concrete corresponding relation of priority.
Visible by above scheme; The present invention is owing to cancelled the notion of assistant edge node; Therefore when the many subrings of needs insert; Can select access node flexibly, rather than must fixingly insert, therefore can realize many subrings structure as shown in Figure 5 from unique Edge node and assistant edge node.
And; The present invention is provided with subring priority; And allow the HELLO message in the low priority subring in the high priority subring, to transmit, thus the subring HELLO message forwarding path that need not pass through main ring increased, and subring HELLO message forwarding path must pass through main ring in the prior art.When link failure appears in main ring; When double-link failure (being a common link fails on the main ring, a non-common link fails) especially occurring, the subring HELLO message forwarding path of process main ring all is blocked; But the subring HELLO message forwarding path that need not pass through main ring that the present invention increases is not affected; The complete of HELLO message forwarding path makes corresponding subring host node can block secondary port, data loopback can not occur, guarantees the normal operation of looped network.And, can find out that from following each embodiment that enumerates RRPP ring net of the present invention can also guarantee the largest connected property of looped network.
The present invention also provides a kind of fringe node that is applied to above-mentioned Ethernet ring network system.Fig. 7 is the sketch map of fringe node in the embodiment of the invention.As shown in Figure 7, this fringe node comprises message receiving element and message retransmission unit, wherein,
The message receiving element is used for sending to the message retransmission unit with being received from the outside HELLO message of node;
The message retransmission unit is used for when receiving the HELLO message that is derived from subring x, and the main ring and the priority that the HELLO message that is received are forwarded to the place Ethernet ring network system are not less than in the subring of subring x.
In reality, which subring the message retransmission unit can judge the HELLO message source from according to the information in the HELLO message when receiving the HELLO message, and then according to the priority relationship between subring, confirms which subring to transmit the HELLO message that receives to.But such operation is too complicated, brings very big computation burden for fringe node equipment.
Therefore preferably, the present invention stores the corresponding relation that information and outbound port are encircled in the source that is provided with on the equipment according to the priority relationship between subring in advance on the edge of, after receiving the HELLO message, finds corresponding outbound port and forwarding through the coupling corresponding relation then.Specifically, the message retransmission unit comprises respective modules and forwarding module; Wherein,
Respective modules is used to preserve the corresponding relation that information and outbound port are encircled in the source that is provided with according to each subring priority relationship.This corresponding relation can adopt ACL to realize, the corresponding relation that message multicast mac address and outbound port perhaps are set is realized.In reality, the particular content of the corresponding relation of preserving in the respective modules need carry out pre-configured according to network configuration.
Wherein, corresponding relation can comprise: source ring information is main ring, and then corresponding output port is for connecting the public port of main ring; Source ring is subring, and then corresponding output port comprises public port and the particular edge port that connects main ring, and said particular edge port is to be not less than the edge port that the subring that encircles in the source links to each other with priority.
Forwarding module; Be used for after receiving the HELLO message; The source ring information that employing is confirmed according to reception HELLO message is mated in the corresponding relation that respective modules is preserved, and finds the outbound port of the HELLO message that receives, and transmits the HELLO message that is received from the outbound port that finds.
Wherein, according to reception HELLO message confirm source ring information mode can for: the MAC Address according to message carries is confirmed source ring information, or confirms source ring information according to the RingID field in the message.
Be that example E-Packets to the correspondence setting of fringe node of the present invention with according to corresponding relation and describes with the dual-homed Ethernet ring network system shown in Fig. 6 below.As shown in Figure 6, equipment S1~S4 forms main ring, and equipment S3, S4 and S5 form subring 1 and S5 is a host node, and equipment S3, S4 and S6 form subring 2 and S6 is a host node.The port 2 of each host node of unified Definition is secondary port.S4 and S3 are fringe node.The priority of supposing subring 1 is higher than subring 2.
So, the corresponding relation that on equipment S4, is provided with can for: when source ring information was subring 1, then corresponding output port was port one, port 2 and port 3; Source ring information is subring 2, and then corresponding output port is port one, port 2, port 3 and port 4.When source ring information was main ring, corresponding output port was port one and port 2.Need to prove that the port that receives the HELLO message is no longer transmitted the HELLO message that receives.
When S4 when port one/port 2 receives the HELLO message that is derived from main ring, transmit from port 2/ port one; When S4 when port 3 receives the HELLO message that is derived from subring 1, transmit from port one and port 2 according to corresponding relation; When S4 when port 4 receives the HELLO message that is derived from subring 2, transmit from port one, port 2 and port 3 according to corresponding relation.
Though fringe node has duplicated many parts with the HELLO message that is received; And forward through different ports; But under the normal situation of link; Have only a HELLO message can get back to source subring host node, other HELLO messages are blocked by certain node on forward-path, and any time has only a HELLO forward-path complete.Following embodiment will specify this point.
The present invention can be applied to start the ethernet ring network of RRPP agreement, also can be applied to start the ethernet ring network of other looped network agreements, as long as each node in the ethernet ring network is carried out corresponding operating according to such scheme, just in protection scope of the present invention.
Be example with the ethernet ring network that starts the RRPP agreement below, lift a plurality of embodiment implementation of the present invention and effect are described in detail.And in following examples, all the port one of defined node is a master port, and port 2 is secondary port.
Embodiment one
Present embodiment is an example with the dual-homing networking structure shown in Fig. 6.The HELLO message forwarding path of main ring and each subring at first is described.
Fig. 8 shows the HELLO message forwarding path of main ring in the present embodiment one.As shown in Figure 8, when main ring was complete, the HELLO message that the port one of main ring host node S1 sends was through each transmission node, received by the secondary port of host node, and the port 2 of S1 gets clogged, and this is identical with prior art.
Fig. 9 shows the HELLO message forwarding path of subring 1 in the present embodiment one.As shown in Figure 9, when main ring and subring were complete, the secondary port of main ring host node S1 was that port 2 gets clogged.The HELLO message that the port one of subring 1 host node S5 sends arrives fringe node S4, because the HELLO message is from subring 1, and the priority of subring 1 is greater than subring 2, and then S4 is forwarded to the HELLO message in the main ring, promptly through port one and port 2 forwarding HELLO messages.Because the port 2 of S1 gets clogged and subring HELLO message is taken as data message in main ring; Therefore the HELLO message of subring 1 stops forwarding after arriving S1, and after another road HELLO message arrives another edge device S3 through common link S3-S4, is forwarded through its port one, 2 and 4 by S3; HELLO message through port 4 is transmitted is got back to S5; Thereby a HELLO message forwarding path that forms be among Fig. 9 the path 2., this moment, S5 blocked its secondary port, to avoid forming data loopback.And S3 gets clogged at the S1 place through the message that port one and 2 sends.
It is thus clear that when main ring and subring were complete, the path among Fig. 9 was 1. obstructed, have only the path 2. logical.When common link S3-S4 fault, S1 decontrols secondary port, and then 2. the path is broken off, and the path is 1. logical.
Figure 10 shows the HELLO message forwarding path of subring 2 in the present embodiment one.Shown in figure 10, when main ring and subring were complete, the secondary port of main ring host node S1 was that port 2 gets clogged.The HELLO message that the port one of subring 2 host node S6 sends arrives fringe node S4; Because the HELLO message is from subring 2; And the priority of subring 2 is lower than subring 1, and then S4 is forwarded to the HELLO message in main ring and the subring 1, promptly transmits the HELLO message through port one, 2 and 3.The HELLO message is forwarded in main ring and subring 1, because the secondary port of S1 and S5 gets clogged, so the HELLO message is merely able to through another fringe node of common link S3-S4 arrival S3.S3 transmits these HELLO messages through its port one~4, has only the message of transmitting through port 3 to get back to S6, thus a HELLO message forwarding path that forms be among Figure 10 the path 2., this moment, S6 blocked its secondary port, to avoid forming data loopback.
It is thus clear that when main ring and subring were complete, the path among Figure 10 was 1. 3. obstructed with the path, have only the path 2. logical.When common link S3-S4 fault, S1 decontrols secondary port, and S5 still blocks, so subring 2 has only the path 1. logical; When on common link and the main ring during arbitrary other link failures, S5 decontrols its secondary port, so subring 2 has only the path 3. logical.
HELLO message forwarding path in conjunction with the main ring shown in Fig. 8~Figure 10 and each subring; Can learn; When main ring and all subrings were complete, subring 1 all had a HELLO message forwarding path complete with subring 2, so the secondary port of host node gets clogged in subring 1 and the subring 2; The data message loop can be do not formed, and communication can both be kept between each node.
When a common link fails was arranged on the main ring, the secondary port of S1 was decontroled and is blocked, and guarantees the connection of each node of main ring.Subring 1 still has a HELLO message forwarding path complete respectively with subring 2; Be among Fig. 9 the path 1. with Figure 10 in the path 1.; Therefore the secondary port of host node still blocks in subring 1 and the subring 2, can not form the data message loop, and can both keep communication between each node.
When a non-common link fails is arranged on the main ring, link S1-S4 fault for example, the secondary port of S1 is decontroled and is blocked, and guarantees the connection of each node of main ring.Subring 1 still has a HELLO message forwarding path complete respectively with ring 2; Be among Fig. 9 the path 2. with Figure 10 in the path 2.; Therefore the secondary port of host node still blocks in subring 1 and the subring 2, can not form the data message loop, and can both keep communication between each node.
When a common link and a non-common link fails are arranged on the main ring; Link S3-S4 shown in figure 11 and link S1-S4 fault; At this moment; The secondary port of main ring host node S1 is decontroled, and the HELLO message forwarding path of subring 1 simultaneously all interrupts, so the secondary port of subring 1 host node S5 is also decontroled.And still have the HELLO message forwarding path of path shown in 3. in the subring 2, so the secondary port of host node still blocks in the subring 2, the generation of anti-loops.And, as can beappreciated from fig. 11,, between six equipment of the S1~S6 in the system path can be arranged, thereby guarantee the largest connected property of looped network though two link failures are arranged in the main ring.
Embodiment two
It is example that present embodiment inserts main ring with subring 2 through different fringe nodes with subring 1.Referring to Figure 12, equipment S1~S4 forms main ring, and S7, S8, S1 and S2 have constituted subring 1, and inserts main ring through fringe node S1 and S2, and S5, S6, S3 and S4 constitute subring 2, and inserts main ring through fringe node S3 and S4.The priority of subring 1 is greater than subring 2.
The HELLO message forwarding path of main ring and each subring at first is described.
In the present embodiment, the HELLO message forwarding path of main ring also is in main ring.This is identical with prior art.When main ring was complete, the port 2 of S1 got clogged.
Figure 12 also shows the HELLO message forwarding path of subring 1 in the present embodiment two.Shown in figure 12, when main ring and subring were complete, the secondary port of main ring host node S1 was that port 2 gets clogged.The HELLO message that the port one of the host node S7 of subring 1 sends arrives fringe node S2; Because the HELLO message is from subring 1; And the priority of subring 1 is greater than subring 2, and then S2 is forwarded to the HELLO message in the main ring, promptly transmits the HELLO message through port one and port 2.Because the port 2 of S1 gets clogged, so during the port 2 of the HELLO message of subring 1 through S3 and S4 arrival S1, stop forwarding.And after the HELLO message on another road arrives S1 through common link S1-S2, be forwarded to S7 through its port 3 by S1, thereby form a HELLO message forwarding path be among Figure 12 the path 1., this moment, S7 blocked its secondary port, to avoid forming data loopback.
It is thus clear that when main ring and subring were complete, the path among Figure 12 was 2. obstructed, have only the path 1. logical.And when common link S1-S2 fault, S1 decontrols secondary port, and then 1. the path is broken off, and the path is 2. logical.
Figure 13 shows the HELLO message forwarding path of subring 2 in the present embodiment two.Shown in figure 13, when main ring and subring were complete, the secondary port of S1 got clogged.When the HELLO message that the port one of subring 2 host node S6 sends arrives fringe node S4; Because the HELLO message is from subring 2; And the priority of subring 2 is lower than subring 1, and then S4 is forwarded to this HELLO message in the main ring (because S4 does not connect subring 1), promptly transmits the HELLO message through port one and 2.Because the secondary port of S1 gets clogged, so the HELLO message is merely able to arrive another fringe node S3 through common link S3-S4.S3 transmits these HELLO messages through its port 2 and 3, has only the message of transmitting through port 3 to get back to S6, thus a HELLO message forwarding path that forms be among Figure 12 the path 1., this moment its secondary port of S6 obstruction.And the HELLO message of transmitting through the port 2 of S3 can stop forwarding because of the obstruction of S1 and S7.
It is thus clear that when main ring and subring were complete, the path among Figure 13 was 2. 3. obstructed with the path, have only the path 1. logical.When common link S3-S4 fault, S1 decontrols secondary port, and S7 still blocks, and the path among Figure 13 is 2. logical at this moment, and the path is 1. 3. obstructed with the path.When two common link fails, or during the non-common link fails of common link fails and main ring, the path among Figure 13 is 3. logical, and the path is 1. 2. obstructed with the path.
HELLO message forwarding path in conjunction with each subring shown in Figure 12~Figure 13; Can learn; When main ring and all subrings were complete, subring 1 all had a HELLO message forwarding path complete with subring 2, so the secondary port of host node gets clogged in subring 1 and the subring 2; Thereby prevent the data message loop, and can both keep communication between each node.
When the common link S1-S2 fault on the main ring, referring to Figure 14, the secondary port of S1 is decontroled and is blocked, and each node in the main ring can proper communication.For subring 1, the path of subring 1 is 1. obstructed, but because the relieving of the secondary port of S1, the path of subring 1 is 2. open-minded, so subring 1 possesses a complete HELLO message forwarding path, and the secondary port of S7 still blocks.For subring 2, link S1-S2 fault can not exert an influence to subring 2, and 1. the path of subring 2 keeps being communicated with.Therefore, subring 1 has a HELLO message forwarding path complete respectively with subring 2, and the secondary port of host node still blocks in subring 1 and the subring 2, can not form data loopback.And can both keep communication between eight equipment of the S1~S8 in the system, guaranteed the largest connected property of looped network.
When the common link S3-S4 fault on the main ring, referring to Figure 15, the secondary port of S1 is decontroled and is blocked, and each node in the main ring can proper communication.For subring 1, link S3-S4 fault can not exert an influence to subring 1, and 1. the path of subring 1 keeps being communicated with.For subring 2, the path of subring 2 is 1. obstructed, but because the relieving of the secondary port of S1, the path of subring 2 is 2. open-minded.Therefore, subring 1 has a HELLO message forwarding path complete respectively with subring 2, and the secondary port of host node still blocks in subring 1 and the subring 2, can not form data loopback.And can both keep communication between eight equipment of the S1~S8 in the system, guaranteed the largest connected property of looped network.
When common link S3-S4 on the main ring and the equal fault of S3-S4, referring to Figure 16, the host node S1 in the main ring decontrols its secondary port; Two HELLO message forwarding path is all obstructed in the subring 1, so host node S7 also decontrols its secondary port, and at this moment, 3. the path in the subring 2 is communicated with, and the secondary port of host node still blocks in the subring 2.As can be seen from Figure 16, between eight equipment of the S1~S8 in the system path can be arranged, thereby guarantee the largest connected property of looped network.
When one on the main ring non-common link S1-S4 fault, referring to Figure 17, the host node S1 in the main ring decontrols its secondary port; The path is 1. available in the subring 1, and the path of subring 2 is 1. available, so subring 1 has a HELLO message forwarding path complete respectively with subring 2, and the secondary port of host node still blocks in subring 1 and the subring 2, can not form the data message loop.And, can both keep communication between six equipment of the S1~S6 in the system, guaranteed the largest connected property of looped network.HELLO message forwarding path and Figure 17 after the non-common link S2-S3 fault are similar, do not detail here.
When a common link S3-S4 on the main ring and a logical non-common link S1-S4 fault, referring to Figure 18, host node S1 decontrols its secondary port in the main ring; The path is 1. available in the subring 1; Subring 2 host node S6 decontrol its secondary port.At this moment, shown in figure 18, there is not the data message loop in main ring, subring 1 and the subring 2, and, can both keep communication between eight equipment of the S1~S8 in the system, guaranteed the largest connected property of looped network.
In reality, when the main ring double-link failure, perhaps main ring all might produce the moment loop when double-link failure recovers.The present invention avoids the generation of moment loop through data field timer and control domain timer are set.
During at first to double-link failure moment loop avoid be described in detail.
In the prior art, host node does not receive the HELLO message at the appointed time, just decontrols secondary port.The timing of Fail timer was set to the HELLO message usually and sent the cycle multiple, is generally 3 times by the decision of Fail timer this stipulated time.The timing of the Fail timer of each node is identical in the system.
So; When main ring double-link failure shown in figure 11 takes place; The secondary port of the secondary port of subring 1 host node S5 and subring 2 host node S6 is not all decontroled, and the HELLO message of subring 2 is regarded as data message in subring 1, and therefore the path of subring 2 this moment is 3. still obstructed.When the Fail of S5 and S6 timer expiry; S5 and S6 decontrol secondary port simultaneously; 3. the path of subring 2 this moment is communicated with; Path connection 3. makes and the HELLO message of subring 2 and the secondary port that data message possibly arrive subring 2 host node S6 simultaneously because S6 does not also have enough time to block its secondary port, has therefore formed the moment data loopback.
During for fear of double-link failure moment loop generation, the present invention distinguishes the protocol massages of data message (data message that comprises main ring and subring) and non-subring on the subring host node, no longer make no exception.When not receiving the HELLO message that oneself sends at the appointed time, decontrol the obstruction of non-book loop protocol message earlier, after waiting for a period of time, decontrol the obstruction of data message again.
The scheme of the protocol massages of this differentiation data message and Fei Ben subring can realize through two Fail timers are set.Specifically; Two timers are set on each subring host node; One is data field Fail timer; One is control domain Fail timer, and the timing of data field Fail timer is greater than control domain Fail timer, and the timing of these two timers is all greater than transmission cycle of HELLO message.Preferably, the timing that control domain Fail timer can be set sends 2 times of cycle for the HELLO message, and the timing of data field Fail timer is 4 times that the HELLO message sends the cycle.
So, the control unit in the subring host node, when HELLO message that place subring host node sends, control control domain Fail timer and data field Fail timer pick up counting; When control domain Fail timer expiry, decontrol the obstruction of the secondary port of place subring host node to non-book loop protocol message; When data field Fail timer expiry, decontrol the obstruction of the secondary port of place subring host node to the data message; When subring host node in place receives the HELLO message that oneself sends, the timing of remove controls territory Fail timer and data field Fail timer.
Referring to Figure 11, adopt the scheme of two Fail timers, behind the main ring double-link failure; S5 and S6 can be when control domain Fail timer expiries; Decontrol the obstruction of HELLO message, 3. the HELLO message of subring 2 this moment arrives S6 along the path, so S6 blocks its secondary port to avoid data loopback; And the timing of two Fail timers of S6 cancellation, thereby cancelled the timer expiry state.When the data field Fail of S5 timer expiry, S5 decontrols blocked state, makes data message be able to circulation.
Because main ring has quick LINK-DOWN informing mechanism, the mechanism that therefore two Fail timers are set can not introduced host node.Certainly, the situation for fear of LINK-DOWN notice message dropping also can be provided with two Fail timers on the main ring host node.
Similar with the subring host node, when two Fail timers were set on the main ring host node, during the HELLO message that the main ring host node sends at the place of the control unit in the subring host node, control control domain Fail timer and data field Fail timer picked up counting; When control domain Fail timer expiry, decontrol the obstruction of the secondary port antithetical phrase loop protocol message of place main ring host node; When data field Fail timer expiry, decontrol the obstruction of the secondary port of place main ring host node to the data message; When place main ring host node receives the HELLO message that oneself sends, the timing of remove controls territory Fail timer and data field Fail timer.
Below, the moment loop of main ring when double-link failure recovers avoided being described in detail.
In the prior art; For fear of instant loop; When certain link during from fault recovery, the port of this both link ends is set to interim obstruction (Pre-forwarding) state earlier, then when the affirmation of receiving host node recovers message (COMPLETE-FLUSH-FDB); After perhaps the Pre-forwarding state is kept the stipulated time, just switch to connection (Link-Up) state from the Pre-forwarding state.The timing of Fail timer was set to the HELLO message usually and sent the cycle multiple, often is 3 times by the decision of Fail timer the said stipulated time.When port was in the Pre-forwarding state, the protocol massages of data message and Fei Ben ring all can not pass through.
So, when main ring double-link failure shown in figure 11 recovered, S1, S3 and S4 still were in the Pre-forwarding state on the main ring, and the protocol massages of subring and data message can't pass through main ring.After the Pre-forwarding of S1, S3 and S4 state is kept the stipulated time; Switch to connection (Link-Up) state simultaneously; Thereby decontrol secondary port separately; The protocol massages and the data message of subring 1 this moment get into subring 1 simultaneously, and the secondary port of subring this moment 1 host node S5 still is in the relieving state, has just formed the moment data loopback.
When double-link failure recovers moment loop generation, the present invention distinguishes data message and subring protocol massages on each node of main ring, no longer make no exception.When the Pre-forwarding of main link state is held time when reaching the stipulated time, decontrol the obstruction of subring protocol massages earlier, after waiting for a period of time, decontrol the obstruction of data message again.
The scheme of this differentiation data message and subring protocol massages can realize through two Fail timers are set.Specifically; Two timers are set on each node of main ring; One is data field Fail timer; One is control domain Fail timer, and the timing of data field Fail timer is greater than control domain Fail timer, and the timing of these two timers is all greater than transmission cycle of HELLO message.Preferably, the timing that control domain Fail timer can be set sends 2 times of cycle for the HELLO message, and the timing of data field Fail timer is 4 times that the HELLO message sends the cycle.
So, the control unit in each node of main ring, when oneself state was moved to the pre-forwarding state, control control domain Fail timer and data field Fail timer picked up counting; When control domain Fail timer expiry, decontrol the obstruction of interim port blocked antithetical phrase loop protocol message; When data field Fail timer expiry, decontrol the obstruction of interim port blocked to the data message.
Visible from the above, RRPP ring net of the present invention system can provide subnet access way flexibly, when the main ring link breaks down, can guarantee not occur data loopback, and can guarantee the largest connected property of looped network in maximum.And, because the present invention has cancelled SRPT status checkout mechanism, therefore if when ring network structure as shown in Figure 4 and fault, S3 can not block its port 3, thereby has guaranteed the largest connected property of looped network.
Based on above-described Ethernet ring network system of the present invention, the present invention also provides a kind of Ethernet ring network realization method, and this method is used and comprised in the Ethernet ring network system of main ring and one or more subrings; A plurality of subrings insert main ring from identical or different node, and the node that subring and main ring intersect is called fringe node.This method may further comprise the steps:
A, be different subrings configuration different priorities.
Host node in B, main ring and each subring sends the HELLO message from master port, receive the handshake message that self sends at secondary port after, block secondary port.
Each node in C, the main ring receives the HELLO message with institute and is regarded as data message when the HELLO message that receives from subring, confirms to transmit or block the HELLO message that is received according to self port status.
D, fringe node are forwarded to main ring and priority is not less than in the subring of subring x with the HELLO message that is received when receiving the HELLO message that is derived from subring x.Identical with prior art, fringe node only is forwarded in the main ring when receiving the HELLO message that is derived from main ring.
The host node of E, each subring receives the HELLO message with institute and is regarded as data message when the HELLO message that receives from other subrings, confirms to transmit or block the HELLO message that is received according to the master node port state.
More than each step sequence number and be not used in and limit the sequencing that step is carried out.
Wherein, the configuration step of said steps A can for: adopt RingID field of the prior art as subring priority tag of the present invention.
Said step B can encircle the corresponding relation realization of information and outbound port through the pre-configured source of coupling.This corresponding relation specifically describes at preamble, omits here.In this case; Said step B is: each fringe node is forwarded to main ring and priority with the HELLO message that is received and is not less than in the subring of subring x and comprises: fringe node is after receiving the HELLO message; The source ring information that employing is confirmed according to reception HELLO message is mated in said corresponding relation; Find the outbound port of the HELLO message that receives, and transmit the HELLO message that is received from the outbound port that finds.
The moment data loopback that possibly occur during for fear of the main ring double-link failure; When this method comprises the steps: that further the host node in each subring is not received the HELLO message that oneself sends in the t1 at the fixed time; Decontrol the obstruction of secondary port earlier to non-book loop protocol message; After waiting for one section setting-up time t2, decontrol the obstruction of secondary port again to the data message.
This step can realize through two Fail timers are set.Specifically, as the host node Sx of subring 1After sending the HELLO message, control a control domain Fail timer and a data field Fail timer picks up counting; When said control domain Fail timer expiry, decontrol host node Sx 1Secondary port to the obstruction of non-book loop protocol message; When said data field Fail timer expiry, decontrol the obstruction of the secondary port of host node Sx to the data message; As host node Sx 1When receiving the HELLO message that oneself sends, cancel the timing of said control domain Fail timer and said data field Fail timer.
Wherein, data field Fail timer is identical with preamble with the setting of control domain Fail timer, omits here.
The mechanism of this pair Fail timer also can be introduced the main ring host node, the moment data loopback that possibly cause when avoiding the LINK-DOWN message dropping.In this case, this method further comprises the steps:
Host node Sx when main ring 2When sending the HELLO message, control a control domain Fail timer and a data field Fail timer picks up counting;
When control domain Fail fault timer expiry, decontrol host node Sx 2The obstruction of secondary port antithetical phrase loop protocol message; When data field Fail timer expiry, decontrol host node Sx 2Secondary port to the obstruction of data message;
As host node Sx 2When receiving the HELLO message that oneself sends, the timing of remove controls territory Fail timer and data field Fail timer.
The moment data loopback that when the main ring double-link failure recovers, possibly occur; This method comprises the steps: that further each node of main ring is when interim blocked state is kept scheduled time t1; Decontrol the obstruction of interim port blocked antithetical phrase loop protocol message earlier; After waiting for one section setting-up time t2, decontrol the obstruction of interim port blocked again to the data message.
This step also can realize through two above-mentioned Fail timers are set.Specifically, as each node Sx of main ring 3When oneself state is moved to the pre-forwarding state, control a control domain Fail timer and a data field Fail timer picks up counting;
When control domain Fail timer expiry, decontrol node Sx 3The obstruction of interim port blocked antithetical phrase loop protocol message; When data field Fail timer expiry, decontrol node Sx 3Interim port blocked to the obstruction of data message.
In sum, more than being merely preferred embodiment of the present invention, is not to be used to limit protection scope of the present invention.All within spirit of the present invention and principle, any modification of being done, be equal to replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (15)

1. Ethernet ring network system, this system comprises main ring and the one or more subring of being made up of node; Main ring includes a host node with each subring; Each host node is used for sending the HELLO message of shaking hands from the master port of self, receive the HELLO message that self sends at secondary port from self after, block secondary port; Each node in the main ring receives the HELLO message with institute and is regarded as data message when the HELLO message that receives from subring, confirms to transmit or block the HELLO message that is received according to self port status;
It is characterized in that,
Different subrings insert main ring from identical or different node, and all nodes that subring and main ring intersect are called fringe node; Different subrings are configured to have different priority;
Said fringe node is used for when receiving the HELLO message that is derived from subring x, and reception HELLO message is forwarded to main ring and priority is not less than in the subring of subring x;
The host node of each subring in the system receives the HELLO message with institute and is regarded as data message when the HELLO message that receives from other subrings, confirms to transmit or block the HELLO message that is received according to self port status.
2. the system of claim 1; It is characterized in that; Host node in each subring is further used for, and when not receiving the HELLO message that oneself sends at the fixed time in the t1, decontrols the obstruction of secondary port to non-book loop protocol message earlier; After waiting for one section setting-up time t2, decontrol the obstruction of secondary port again to the data message.
3. system as claimed in claim 2 is characterized in that, the host node in each subring comprises control unit, control domain fault timer and data field fault timer;
The timing of said data field fault timer is greater than the timing of said control domain fault timer, and the timing of data field fault timer and control domain fault timer is all greater than transmission cycle of HELLO message;
Said control unit is used for after place subring host node sends the HELLO message, controls said control domain fault timer and said data field fault timer picks up counting; When said control domain fault timer expiry, decontrol the obstruction of the secondary port of place subring host node to non-book loop protocol message; When said data field fault timer expiry, decontrol the obstruction of the secondary port of place subring host node to the data message; When subring host node in place receives the HELLO message that oneself sends, cancel the timing of said control domain fault timer and said data field fault timer.
4. the system of claim 1 is characterized in that, the host node of main ring comprises control unit, data field fault timer and control domain fault timer;
The timing of said data field fault timer is greater than the timing of said control domain fault timer, and the timing of data field fault timer and control domain fault timer is all greater than transmission cycle of HELLO message;
Said control unit is used for when place main ring host node sends the HELLO message, controls said control domain fault timer and said data field fault timer picks up counting; When said control domain fault timer expiry, decontrol the obstruction of the secondary port antithetical phrase loop protocol message of place main ring host node; When said data field fault timer expiry, decontrol the obstruction of the secondary port of place main ring host node to the data message; When place main ring host node receives the HELLO message that oneself sends, cancel the timing of said control domain fault timer and said data field fault timer.
5. the system of claim 1; It is characterized in that; Each node of main ring is further used for, and when interim blocked state is kept scheduled time t1, decontrols the obstruction of interim port blocked antithetical phrase loop protocol message earlier; After waiting for one section setting-up time t2, decontrol the obstruction of interim port blocked again to the data message.
6. system as claimed in claim 5 is characterized in that, the node in the main ring comprises control unit, control domain fault timer and data field fault timer;
The timing of said data field fault timer is greater than the timing of said control domain fault timer, and the timing of control domain fault timer and data field fault timer is all greater than transmission cycle of HELLO message;
Said control unit is used for when oneself state is moved to interim blocked state, controls said control domain fault timer and data field fault timer picks up counting; When said control domain fault timer expiry, decontrol the obstruction of the interim port blocked antithetical phrase loop protocol message of place node; When said data field fault timer expiry, decontrol the obstruction of the interim port blocked of place node to the data message.
7. like claim 3 or 4 or 6 described systems, it is characterized in that the timing of said data field fault timer sends 4 times of cycle for the HELLO message; The timing of said control domain fault timer sends 2 times of cycle for the HELLO message.
8. Ethernet ring network realization method, this method is applied to comprise in the Ethernet ring network system of main ring and one or more subrings; It is characterized in that different subrings insert main ring from identical or different node, all nodes that subring and main ring intersect are called fringe node;
This method comprises:
Be different subring configuration different priorities;
Host node in main ring and each subring sends the HELLO message from master port, receive the handshake message that self sends at secondary port after, block secondary port;
Each node in the main ring receives the HELLO message with institute and is regarded as data message when the HELLO message that receives from subring, confirms to transmit or block the HELLO message that is received according to self port status;
Fringe node is forwarded to main ring and priority is not less than in the subring of subring x with the HELLO message that is received when receiving the HELLO message that is derived from subring x;
The host node of each subring confirms to transmit or block the HELLO message that is received according to host node self port status of this subring in the system when the HELLO message that receives from other subrings.
9. method as claimed in claim 8 is characterized in that, this method further comprises: the corresponding relation that source ring information and outbound port are set in advance; Source ring information is main ring, and then corresponding output port is for connecting the public port of main ring; Source ring is subring, and then corresponding output port comprises public port and the particular edge port that connects main ring, and said particular edge port is to be not less than the edge port that the subring that encircles in the source links to each other with priority;
Saidly the HELLO message that is received is forwarded to main ring and priority is not less than in the subring of subring x and comprises: fringe node is after receiving the HELLO message; The source ring information that employing is confirmed according to reception HELLO message is mated in said corresponding relation; Find the outbound port of the HELLO message that receives, and transmit the HELLO message that is received from the outbound port that finds.
10. method as claimed in claim 9; It is characterized in that; This method further comprises: when the host node in each subring is not received the HELLO message that oneself sends in the t1 at the fixed time; Decontrol earlier the obstruction of secondary port to non-book loop protocol message, wait for one section setting-up time t2 after, decontrol the obstruction of secondary port again to the data message.
11. method as claimed in claim 10 is characterized in that, the said obstruction of secondary port of decontroling earlier to non-book loop protocol message, wait for one section setting-up time t2 after, decontrol the obstruction of secondary port again to the data message, comprising:
Host node Sx when subring 1After sending the HELLO message, control a control domain fault timer and a data field fault timer picks up counting;
When said control domain fault timer expiry, decontrol host node Sx 1Secondary port to the obstruction of non-book loop protocol message; When said data field fault timer expiry, decontrol host node Sx 1Secondary port to the obstruction of data message;
As host node Sx 1When receiving the HELLO message that oneself sends, cancel the timing of said control domain fault timer and said data field fault timer;
The timing of said data field fault timer is greater than the timing of said control domain fault timer, and the timing of data field fault timer and control domain fault timer is all greater than transmission cycle of HELLO message.
12. method as claimed in claim 8 is characterized in that, this method further comprises:
Host node Sx when main ring 2When sending the HELLO message, control a control domain fault timer and a data field fault timer picks up counting;
When said control domain fault timer expiry, decontrol host node Sx 2The obstruction of secondary port antithetical phrase loop protocol message; When said data field fault timer expiry, decontrol host node Sx 2Secondary port to the obstruction of data message;
As host node Sx 2When receiving the HELLO message that oneself sends, cancel the timing of said control domain fault timer and said data field fault timer;
The timing of said data field fault timer is greater than the timing of said control domain fault timer, and the timing of data field fault timer and control domain fault timer is all greater than transmission cycle of HELLO message.
13. method as claimed in claim 8; It is characterized in that; This method further comprises: when each node of main ring is kept scheduled time t1 at interim blocked state; Decontrol earlier the obstruction of interim port blocked antithetical phrase loop protocol message, wait for one section setting-up time t2 after, decontrol the obstruction of interim port blocked again to the data message.
14. method as claimed in claim 13; It is characterized in that this method further comprises: the said obstruction of decontroling earlier interim port blocked antithetical phrase loop protocol message, wait for one section setting-up time t2 after; Decontrol the obstruction of interim port blocked again, comprising the data message:
The node Sx of main ring 3When oneself state is moved to interim blocked state, control a control domain fault timer and a data field fault timer picks up counting;
When said control domain fault timer expiry, decontrol node Sx 3The obstruction of interim port blocked antithetical phrase loop protocol message; When said data field fault timer expiry, decontrol node Sx 3Interim port blocked to the obstruction of data message;
The timing of said data field fault timer is greater than the timing of said control domain fault timer, and the timing of control domain fault timer and data field fault timer is all greater than transmission cycle of HELLO message.
15., it is characterized in that the timing of said data field fault timer sends 4 times of cycle for the HELLO message like claim 11 or 12 or 14 described methods; The timing of said control domain fault timer sends 2 times of cycle for the HELLO message.
CN2009100928230A 2009-09-09 2009-09-09 Ethernet ring network system and border node thereof and Ethernet ring network realization method Active CN101656664B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2009100928230A CN101656664B (en) 2009-09-09 2009-09-09 Ethernet ring network system and border node thereof and Ethernet ring network realization method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2009100928230A CN101656664B (en) 2009-09-09 2009-09-09 Ethernet ring network system and border node thereof and Ethernet ring network realization method

Publications (2)

Publication Number Publication Date
CN101656664A CN101656664A (en) 2010-02-24
CN101656664B true CN101656664B (en) 2012-05-23

Family

ID=41710763

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2009100928230A Active CN101656664B (en) 2009-09-09 2009-09-09 Ethernet ring network system and border node thereof and Ethernet ring network realization method

Country Status (1)

Country Link
CN (1) CN101656664B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104283758B (en) * 2014-08-04 2018-07-17 新华三技术有限公司 A kind of method and apparatus that boss's ring is associated with by data vlan
CN108234200B (en) * 2017-12-21 2019-08-27 中国联合网络通信集团有限公司 Communication processing method and device based on access ring
CN110336728B (en) * 2019-07-26 2021-08-06 新华三大数据技术有限公司 Message processing method, node equipment, electronic equipment and readable storage medium

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101499949A (en) * 2008-02-01 2009-08-05 杭州华三通信技术有限公司 Failure recovery processing method in intersected ring network, node and intersected ring network
CN101534232A (en) * 2008-03-10 2009-09-16 中兴通讯股份有限公司 Method for preventing internet storm occurring in Ether multi-ring network
CN101677281A (en) * 2008-09-19 2010-03-24 中兴通讯股份有限公司 Ether multiple ring control channel allocation and message transmission method thereof
CN101677267A (en) * 2008-09-17 2010-03-24 中兴通讯股份有限公司 Method for retransmitting protocol frames in transannular way and sharing nodes in multiple rings of Ethernet

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101499949A (en) * 2008-02-01 2009-08-05 杭州华三通信技术有限公司 Failure recovery processing method in intersected ring network, node and intersected ring network
CN101534232A (en) * 2008-03-10 2009-09-16 中兴通讯股份有限公司 Method for preventing internet storm occurring in Ether multi-ring network
CN101677267A (en) * 2008-09-17 2010-03-24 中兴通讯股份有限公司 Method for retransmitting protocol frames in transannular way and sharing nodes in multiple rings of Ethernet
CN101677281A (en) * 2008-09-19 2010-03-24 中兴通讯股份有限公司 Ether multiple ring control channel allocation and message transmission method thereof

Also Published As

Publication number Publication date
CN101656664A (en) 2010-02-24

Similar Documents

Publication Publication Date Title
US8244838B2 (en) Industrial controller employing the network ring topology
CN102377601B (en) LSP (Local Switched Path) fault notification method and LSP fault notification device
JP4628945B2 (en) Layer 2 network
CN101146014B (en) Fault-tolerant Ethernet network
CN101677267B (en) Method for retransmitting protocol frames in transannular way and sharing nodes in multiple rings of Ethernet
JP5168230B2 (en) Communication system, edge router, and signal transfer method
CN101075935A (en) Method, system and apparatus for realizing loop protection
CN100454880C (en) Method and system of realizing rig network protection
US8264954B2 (en) Method and device for operating a network and communication system comprising such device
CN101227371B (en) Method and apparatus of backup switch among same level switch equipments
CN103124224A (en) Multiple-fault-tolerant Ethernet network for industrial control
CN101841450B (en) Method and system for realizing stable communication by forming multiple ring topologies into intersecting ring
CN103428060A (en) Seamless redundancy realization method of loop network
CN102055658B (en) Method and equipment for realizing failure protection in rapid ring protection protocol single-ring networking
CN103107940A (en) Redundant gateway system for device level ring networks
CN102130912B (en) Rapid ring protection protocol (RRPP)-based multiple virtual local area network (VLAN) register protocol (MVRP) realization method and equipment
CN101989930B (en) Method and switching equipment for realizing Ethernet double loop
CN101656664B (en) Ethernet ring network system and border node thereof and Ethernet ring network realization method
US8959386B2 (en) Network and expansion unit and method for operating a network
US8493844B2 (en) Method for the reconfiguration of a communications network
EP2525527B1 (en) Network relay device and network relay method
JP2008022271A (en) Network apparatus
JP5441566B2 (en) Communication terminal, communication system, and node switching method
JP2006174422A (en) Communication apparatus and fault notification method
JP5520741B2 (en) Communication device

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
CP03 Change of name, title or address

Address after: 310052 Binjiang District Changhe Road, Zhejiang, China, No. 466, No.

Patentee after: Xinhua three Technology Co., Ltd.

Address before: 310053 Hangzhou hi tech Industrial Development Zone, Zhejiang province science and Technology Industrial Park, No. 310 and No. six road, HUAWEI, Hangzhou production base

Patentee before: Huasan Communication Technology Co., Ltd.

CP03 Change of name, title or address