CN102150053A

CN102150053A - Protection for provider backbone bridge traffic engineering

Info

Publication number: CN102150053A
Application number: CN2009801356260A
Authority: CN
Inventors: N.布拉格
Original assignee: Nortel Networks Ltd
Current assignee: Nortel Networks Ltd
Priority date: 2008-09-11
Filing date: 2009-06-22
Publication date: 2011-08-10
Also published as: CA2735000A1; EP2329281A4; JP5485998B2; EP2329281A1; JP2012502583A; KR20110073445A; WO2010030428A1

Abstract

Network protection between endpoints includes both end-to-end and local section protection. A primary path between the endpoints includes a plurality of links, and is protected on an end-to-end basis by a protection path associated with different links. At least one sections of the primary path is also protected on a local basis, where a ''section'' is a link, ring, trunk or other portion of a network. One or more sections of the protection path may also be protected on a local basis. In response to detection of a fault condition, section protection is utilized to overcome the fault if possible, and otherwise end-to-end protection is utilized. Invoking section protection does not imply a switchover from the primary end-to-end path to the protection end-to-end path. Rather, the primary end-to-end path is rerouted in the affected section.

Description

Be used for the protection of provider's backbone bridge traffic engineering

Technical field

The present invention relates generally to network service, and the protection that relates more particularly to be used for provider's backbone bridge traffic engineering.

The cross reference of related application

Require the right of priority of the U.S. Provisional Patent Application 61/096,011 that is entitled as " SECTION PROTECTION FOR PROVIDER BACKBONE BRIDGE TRAFFIC ENGINEERING " of submission on September 11st, 2008.

Background technology

Provider's backbone bridge traffic engineering (" PBB-TE ") is used end-to-end 1:1 protection example, is referred to herein as " end-to-end protection (e2e protection) ".This provides the protection mechanism that can be used on the robust in the connection of arbitrary mess mesh network.But, the bottom physical connection of many networks is usually used fiber optic loop, and it makes the protection on the ring topology become material circumstance.In cascade loop network as shown in Figure 1, the node of coupling is to being used to the communication link of ring adjacent on the interconnect logic.Main path between the end points crosses the link between a plurality of (being the main path node) of node centering of coupling.The different links between the node (promptly protecting path node) of node centering correspondence of coupling are crossed in the protection path.Every paths among the PBB-TE is associated with VLAN ID (VID).Under the situation about breaking down in the link of main path or node, change the VID that uses on the Frame by the head end end points, flow is switched to the protection path.

Although the work of end-to-end 1:1 protection example is good, it has some defectives.For example, different rings may be associated with the different geographic region that different tissues in the bearer is managed and operated, and each tissue may wish do not need with other situation about organizing and coordinating under scheduling maintenance shut down.In addition, in the communication deployment, it is enough obvious that the failure-frequency on the end-to-end path of expansion may become long, and the probability of malfunction second time that occurred in before the first time, fault was repaired on the protection path becomes unacceptable.The end-to-end protection example of expansion can be used for alleviating this situation, but needed protection number of paths is as the quantity of ring and the function of protection domain (for example striding the protection of the multiple failure of a plurality of rings) and phenomenal growth.For example, need four end-to-end paths to be used for a pair of ring of uncoupling cascade, because there are four possible paths.Under the situation of three rings cascade, need six paths to protect any ring, and need eight paths to protect simultaneous single fault on each ring fully last two faults independently.Safeguard that a large amount of paths may be problematic, because forwarding state directly relates to the quantity in the path of being set up, and every paths has increased continuity failure management (" CFM ") session to terminal node.

Summary of the invention

According to embodiments of the invention, a kind of being used for provides the method for protection to comprise at provider's bridge backbone network: in response to fault, utilize segment protection under possible situation, wherein define section between entrance and exit point, described section is not crossed over whole provider bridge backbone network; And under the situation that this fault can't overcome with segment protection, utilize end-to-end protection.

According to another embodiment of the present invention; be used between end points, utilizing provider's bridge backbone network to come communicating devices to comprise: entrance node and exit point node; they are operated together so that the protection at the section of the network between entrance node and the exit point node to be provided; wherein said section is not crossed over whole provider's bridge backbone network; described network is operated in response to fault; under possible situation, utilize segment protection, and under the situation that fault can't overcome with segment protection, utilize the end-to-end protection between the end points.

The advantage related with the present invention comprises to be used Internet resources more efficiently and strengthens protection.All need in node, set up forwarding state with the shared path of target flow although be used for each paths of the required mulitpath of general protection; but according to the employed 1:1 of PBB-TE " head end switching " model, data traffic is only propagated on an end-to-end path at any time.This makes the use poor efficiency (its deployment that had before also made the protection of segment protection and end-to-end path combine is uneconomical) of Internet resources of previous 1+1 segment protection embodiment obtain overcoming.In addition; by utilizing segment protection at main end-to-end path and protection end-to-end path; fault of each ring be can allow, the node failure (its make two ring faults) of a coupling, the situation that the status information amount in the storage of the node place of coupling is doubled comprised.But, should understand and the invention is not restricted to ring structure.

According to detailed description and drawings, these and other advantage of the present invention will be more clear.

Description of drawings

Fig. 1 shows end-to-end 1:1 protection example.

Fig. 2 shows the example with end-to-end protection and segment protection that only is applied to main path.

Fig. 3 shows the segment protection at main path and protection path.

Fig. 4 shows and acts on the basis of response is made in description to the various failure condition shown in Fig. 5 to 10 embodiments of the invention with using.

Fig. 5 show main road by in upstream failure.

Fig. 6 show main road by in downstream fault.

Fig. 7 shows at each bridge to be needed under the situation of single forwarding database, and how Fig. 5 conflicts potentially mutually with the response shown in 6.

Fig. 8 shows the solution based on the VID exchange to problem shown in Figure 7.

Fig. 9 and 10 shows (sympathetic switch) solution is switched in the resonance of problem shown in Figure 7.

Embodiment

Data communication network can comprise various computing machines, server, node, router, switch, bridge, hub, act on behalf of and be coupled to and be configured to other network equipment of Data transmission each other.These equipment will be called as " node " at this.By utilizing internodal one or more communication link to transmit protocol Data Unit between node, as the Internet protocol packets, ethernet frame, data cell, section or other data logic association, data transmit by data communication network.Concrete protocol Data Unit can be handled by a plurality of nodes, and when it on the network it the source and its destination between can pass a plurality of communication links when advancing.

Details described herein is in order to describe the embodiment that goes out shown in the present, to be considered to useful and understandable description to principle of the present invention and design aspect so that provide.Do not make a try and configuration aspects of the present invention is shown with the more details more required for the present invention than basic comprehension.

Fig. 2 shows the protection example with end-to-end protection and segment protection under the situation of two end nodes 200,202 that connect via three cascade rings, and to (204,206), (208,210) interconnect described three cascade rings by the node that mates.The main path 212 related with link 214,216,218 is by protecting based on end-to-end with the related protection path 220 of link 222,224,226.One or more sections of main path are also protected based on the part, and wherein " section " is link, ring, main line or the other parts of network.For example, link 214 can be protected via link 222 by path 228, and link 216 can be protected via link 224 by path 230, and link 218 can be protected via link 226 by path 232.Passing the malfunction of every paths of section independently monitors by the CFM session between the section end points.In response to the detection of fault state, under possible situation, utilize segment protection to overcome this fault, otherwise utilize end-to-end protection.For example, before end-to-end CFM session indication fault, timeout period can be set to longer than the section timeout period of correspondence, is not triggered thereby make unnecessary end-to-end protection switch.Notice that the call area segment protect is not the switching that means from main end-to-end path to the protection end-to-end path.Or rather, only in affected section, main end-to-end path is re-routed.Therefore, at least some problems that are associated with end-to-end 1:1 protection example can be alleviated.Although Fig. 2 shows ring especially, segment protection mechanism as described herein can be used on any topology, and unique restriction is to form two paths of protection section can not intersect between the protection open and close.

It must be understood that; all need in node, set up forwarding state although be used for the every paths of the required mulitpath of general protection; with the path that takies of indication flow, model means that flow was once only advanced on an end-to-end path at any time but PBB-TE is employed 1:1 " head end switching ".This makes the use poor efficiency (its deployment that protection of segment protection and end-to-end path was combined is uneconomical) of Internet resources of previous 1+1 segment protection embodiment obtain overcoming.

Can be closed on the corresponding node at the entrance and exit place of the node of single coupling or this section at the protection of specifying section.For example, primary link 214 and closed on node 204, this node 204 and node 206 couplings via the protection path 228 of the correspondence of link 222 carrying.Therefore, the fault in the link 214 can be used for the link 228 of segment protection and do not need to switch to leave or need node 206 to get involved this handoff procedure on one's own initiative from node 204 to overcome by utilization.Under the situation of less possible node 204 faults; because main end-to-end CFM session failure on the path 212; end-to-end protection will be called, and flow will be switched to link and the node related with protecting path 220, comprise and use node 206 rather than node 204.This advantageously alleviated with matched node between the synchronous relevant problem of protection.

With reference to figure 3, in one embodiment, main path and protection path both have been set up segment protection.More specifically, except the path and link described about Fig. 2, link 222 is protected via link 214 by path 300, and link 224 is protected via link 216 by path 302, and link 226 is protected via link 218 by path 304.At the protection of each section on main path and protection path by open and close independently.Keep the connective check of CFM session at main path between the end points 240,242 and protection path, so that determine the state of two end-to-end paths continuously.Two one of CFM session 306(are clockwise; another is counterclockwise) VID that utilizes main end-to-end path to use determines the state of two paths between

node

200 and 204, and two additional C FM session 307 utilizes the employed VID of protection end-to-end path to determine the state of two paths between node 200 and 206.The VID that two CFM sessions 308 utilize main end-to-end path to use determines the state of two paths between

node

204 and 208, and two additional C FM session 309 utilizes the state of protecting the employed VID of end-to-end paths to determine two paths between node 206 and 210.Each segment protection CFM session need not created at each end-to-end path.For example, two sessions 308 can be used for entering and leaving via

node

204 and 208 all routes of adapter ring.Be used for the segment protection in main path and protection path by utilization, can allow fault of each ring, comprise the node failure (its make two ring faults) of a coupling, the situation that the status information amount of the node place storage of coupling is doubled.

Fig. 4 shows and will be used as the embodiments of the invention of description to the basis of the response of the various different faults situations shown in Fig. 5 to 10.Do not losing under the general situation, only showing at flow and flow to right all situations effect down from a left side, and flowing to use " upstream " and " downstream " with respect to this.As shown in Figure 4, backbone edge bridges (BEB) 400,402 communicates by

ring.Ring docks

404a, 404b, 406a, 406b place in the backbone core bridges (BCB) of coupling.Each BCB comprises four provider's network ports (PNP), for example 408a, 408b, 408c, 408d.Each BEB comprises client's backbone port (CBP) 410 and two PNP 412a, 412b.Main path 420 between the BEB is protected by the segment protection path 422 between end-to-end protection path 424 and BCB 404a and the 406b.The segment protection that is used for the end-to-end protection path is not shown, thereby clearer figure is provided.For detection failure and trigger protection switch, following between keep the CFM session: (BEB 400, BCB 404b); (BEB 400, BCB 404a); (BCB 404b, BCB 406b); (BCB 404a, BCB 406a); (BCB 406b, BEB 402); And (BCB 406a, BEB 402).

Fig. 5 shows the fault 500 in the main path 420, and it is the upstream of BCB 404a.(BEB 400 in response to passing through; BCB 404a) fault that arrives of the fault detect of CFM session, flow by BEB 400 along at 420 segment protection path along first section, pass PNP 412a and also re-route to BCB 404a subsequently to BCB 404b.BCB 404a turns back to main path 420 with the flow through re-routing at PNP 408d place.

Fig. 6 shows the main road of BCB 404a by the fault in the downstream.In response to this failure condition, flow re-routes via the segment protection that is used for second section.More specifically, BCB 404a re-routes the flow that receives its PNP 408c place (flow is through proceeding to BCB 404b and BCB406b from here).BCB 406b is transmitted to BCB 406a with this flow, and BCB 406a turns back to main path with the flow through re-routing at its PNP 408d place.Opening a BCB 404a place through the frame that re-routes in protection changes never in any form; They can and use (related with main end-to-end path) identical VID value really.

Fig. 7 shows node 404b that how response shown in Fig. 5 and 6 to require BCB coupling has the segment protection of using that is used for main end-to-end path under varying environment two forward-paths.Although described two paths can not be simultaneously effectively, but BCB 404b does not have direct means to determine which paths will be used and when use, and keeping this shortage with regard to state and complicacy, effectively to relate to (active involvement) be favourable.Set up these two forward-paths simultaneously and show conflict potentially.Particularly, different routes are associated with (the terminal B EB's 402) identical address at BCB 404b place.This is a problem, because BCB typically has only a forwarding database (FDB).A simple proposal to this problem is at each PNP place different forwarding databases to be arranged.Yet this is opposite with IEEE 802.1ah with ethernet bridge standard IEEE 802.1Q, and FDB can be applicable to all bridge ports in these standards, and this meets under the situation of these standards in expectation be problematic.

Fig. 8 shows another solution to problem shown in Figure 7.Shown solution is based on the VID swap operation.According to this solution, on the segment protection path, from the upstream node to the partner node, perhaps from the partner node to the downstream node, cross over and carry out the VID exchange on the BCB switch.In the example that illustrates especially, the VID exchange is carried out at the PNP of BCB 404b 408a place, and carry out at PNP 408d place again, the flow that makes PNP 408d place exist has identical VID with the flow that enters PNP 408a, but in BCB 404b, there is not address conflict, because each path is associated with different VID.Therefore, FDB of each VLAN can be kept according to ethernet standard.

Fig. 9 and 10 shows respectively another protection handoff solution at downstream and upstream failure, wherein sets up the additional state of CFM form of session on the node of coupling, but clear and definite synchronous to still not needing between as 404a and 404b at the node of coupling.In this solution, additional CCM session is kept between the matched node of the protection on the matched node of opening the segment protection on the main end-to-end path and closed this protection end-to-end path (clockwise and counterclockwise).Whether the downstream part that these between the matched node " diagonal line " session allows the upstream matched node on the segment protection path directly to be determined to this segment protection path of segment protection closing point can operate.For example, between BCB 406a and BCB 404b, keep diagonal line CCM session 900.This diagonal line session allows this network to determine directly whether BCB 404b can arrive the section closing point at the PNP 408d place of BCB 406a.If fault 902 occurs in the main path; then BCB 404b is forwarded to BCB 406a with flow clockwise around the protection path; because partner's matched node 404a also directly detects fault from the fault of the main CCM session of itself and BCB 406a, and flow is switched on the segment protection path.Upstream failure 1000 will can not detected by CCM session 900, so the default ground of flow (that is, when CFM session 900 correctly moves) transmitted by BCB 404b, just look like that main path fault on the ring of upstream is the same, and promptly fault 1000 is assert.Notice under the situation of the fault 1001 on the link between the node of coupling (it all disconnects two paths); the behavior of BCB 404b is unimportant; because two rings all break down now, pass consequent malfunction on the main end-to-end path of two sections and can only switch by end-to-end protection and recover.

Although describe the present invention, those skilled in the art will appreciate that modification and the distortion that under the situation that does not break away from inventive concept disclosed herein, can make illustrated embodiment by above exemplary embodiment.And, although described preferred embodiment is described together with various illustrative structures, one of skill in the art will recognize that and can use various ad hoc structures to embody this system.Correspondingly, the present invention should not be considered by limiting beyond the scope and spirit of claims.

Claims

1. method that is used for providing at provider's bridge backbone network protection comprises:

In response to fault, under possible situation, utilize segment protection, wherein between entrance and exit point, define section, described section is not crossed over whole provider bridge backbone network; And

Under the situation that this fault can't overcome with segment protection, utilize end-to-end protection.

2. the method for claim 1; wherein end-to-end protection comprises main path and protection path; and segment protection comprises the protection at least one section of described main path, and described method comprises at the step of open and close segment protection independently of each section on the main path.

3. the method for claim 1, comprise another step: utilize the continuity failure managing conversation to determine to pass the malfunction in all paths of section, described all paths can be used for entering and leaving via first and second nodes of the described section of provider's bridge backbone network all flows of this section.

4. the method for claim 1; wherein segment protection is closed on the node of a pair of single coupling at every end place of section, and described method comprises another step: utilize segment protection and need not the node of described single coupling is synchronous with their corresponding matched node.

5. the method for claim 1 comprises another step: utilize in matched node on the end-to-end main path with at the continuity failure managing conversation between another matched node on the end-to-end protection path and determine in the segment protection path and have between the downstream protection closing point of corresponding section main path whether have connectedness.

6. the method for claim 1 comprises: by keeping a plurality of forwarding databases that are used at least one node, allow this at least one node to participate at least two protection sections and need not be synchronous with other node.

7. the method for claim 1; comprise: determine by the first continuity failure managing conversation whether this fault can overcome with end-to-end protection; and determine by the second continuity failure managing conversation whether this fault can overcome with segment protection, wherein be longer than timeout period at the second continuity failure managing conversation at the timeout period of the first continuity failure managing conversation.

8. the method for claim 1; comprise: open a little in each segment protection; malfunction and end-to-end vlan identifier based on each paths; to be placed on the main path that passes described section at the frame of given destination or pass on the protection path of described section, and need not change vlan identifier.

9. method as claimed in claim 5; comprise: at the node place of the protection that participates in the upstream and downstream section; based on the state of end-to-end vlan identifier and downstream section, flow is placed on the path of protection closing point of upstream zone or is placed on the path of protection closing point of downstream section.

10. one kind is utilized provider's bridge backbone network communicating devices between end points, comprising:

Entrance node and exit point node; they are operated together so that the protection at the section of the network between entrance node and the exit point node to be provided; wherein said section is not crossed over whole provider bridge backbone network; described network is operated in response to fault; so that under possible situation, utilize segment protection, and under the situation that fault can't overcome with segment protection, utilize end-to-end protection between the end points.

11. device as claimed in claim 10, wherein end-to-end protection comprises main path and protection path, and segment protection comprises the protection at least one section of described main path, and at the segment protection of each section by open and close independently.

12. device as claimed in claim 10, wherein the continuity failure managing conversation is used to determine the malfunction in all paths of passing section, and described all paths can be used for entering and leaving via first and second nodes of the described section of provider's bridge backbone network all flows of this section.

13. device as claimed in claim 10, wherein segment protection is closed on the node of a pair of single coupling at every end place of section, and wherein: utilize segment protection and need not the node of described single coupling is synchronous with their corresponding matched node.

14. device as claimed in claim 10 wherein utilizes in matched node on the end-to-end main path with at the continuity failure managing conversation between another matched node on the end-to-end protection path and determines in the segment protection path and have between the downstream protection closing point of corresponding section main path whether have connectedness.

15. device as claimed in claim 10 wherein by keeping a plurality of forwarding databases that are used at least one node, allows this at least one node to participate at least two protection sections and need not be synchronous with other node.

16. device as claimed in claim 10; wherein use the first continuity failure managing conversation to determine whether this fault can overcome with end-to-end protection; and use the second continuity failure managing conversation to determine whether this fault can overcome with segment protection, and wherein be longer than timeout period at the second continuity failure managing conversation at the timeout period of the first continuity failure managing conversation.

17. device as claimed in claim 10; wherein open a place in each segment protection; malfunction and end-to-end vlan identifier based on each paths; to be placed on the main path that passes described section at the frame of given destination or pass on the protection path of described section, and not change vlan identifier.

18. device as claimed in claim 14; node place in the protection that participates in the upstream and downstream section; based on the state of end-to-end vlan identifier and downstream section, flow is placed on the path of protection closing point of upstream zone or is placed on the path of protection closing point of downstream section.