Tunnel establishing method and device
Technical field
The present invention relates to traffic engineering (TE, Traffic Engineering) technical field, be specifically related to tunnel establishing method and device.
Background technology
Network congestion is the subject matter that affects the backbone network performance.Congested reason may be that Internet resources are not enough, also may the unbalanced local congestion that causes of network resource loads.Traffic engineering (TE, Traffic Engineering) solves is because that load imbalance causes is congested.
TE is by the flow of Real Time Monitoring network and the load of network element, dynamically adjust traffic management parameters, routing parameter and resource constraint parameter etc., make network operation state transition to perfect condition, the use of Optimizing Network Resources, that avoids that load imbalance causes is congested.
Multiprotocol label switching (MPLS, Multi-Protocol Label Switching) itself has the characteristic that some are different from Interior Gateway Protocol (IGP, Interior Gateway Protocol), wherein just has the TE of realization needed, for example:
MPLS supports explicit tag switching path (LSP, Label Switched Path) route;
LSP is more traditional, and single IP forwarding of packets is more convenient for administering and maintaining;
The tag distribution protocol of Constraint-based routing (CR-LDP, Constraint-based Routed Label Distribution Protocol) can be realized the various strategies of TE;
Resource consumption based on the TE of MPLS is lower than other implementation.
MPLS TE combines MPLS technology and TE technology, carries out resource reservation by setting up the lsp tunnel that arrives specified path, makes network traffics get around congested node, reaches the purpose of balancing network flow.In the situation that the resource anxiety, MPLS TE can seize low priority lsp tunnel bandwidth resources, satisfies the demand of large bandwidth LSP or responsible consumer.Simultaneously, when a certain node of lsp tunnel fault or network occurs when congested, MPLS TE can by backup path and fast heavy-route (FRR, Fast ReRoute) protection is provided.
Use MPLS TE network manager only need to set up some LSP and bypass congested node, just can eliminate network congestion.Along with the growth of LSP quantity, can also use special off-line tools to carry out traffic analysis.
For simplicity, below provide the basic conception of MPLS TE:
1, lsp tunnel
For a LSP, in case stamp label at Ingress node to message, the forwarding of flow has just been determined by label fully.Flow is transparent to the intermediate node of LSP, and in this sense, a LSP can be regarded as a lsp tunnel.
2, MPLS traffic engineering tunnel
Disposing heavy-route (Reroute) when maybe needing flow transmitted by mulitpath, may need to use many lsp tunnels.In TE, one group of such lsp tunnel is called traffic engineering tunnel.
MPLS TE mainly realizes two class functions:
1, the processing of static CR-LSP: create and delete static CR-LSP.The bandwidth of static CR-LSP needs manual the appointment.
2, dynamically CR-LSP processes: comprise the processing to three kinds of dissimilar CR-LSP: basic CR-LSP, backup CR-LSP and quick heavy-route CR-LSP.
The processing of static CR-LSP is fairly simple.For dynamic CR-LSP, MPLS TE mainly comprises four parts in realization:
1, issue contains the information of TE attribute
MPLS TE need to understand the dynamic TE association attributes of every link, this can be by expanding to realize to existing IGP agreement with link-state algorithm, such as ospf (OSPF, Open Shortest Path First) expansion of agreement and Intermediate System-to-Intermediate System (IS-IS, Intermediate System to Intermediate System) agreement.
OSPF after the expansion and Intermediate System to Intermediate System have increased link bandwidth, the TE association attributes such as painted in link connection state.Wherein, the link of the maximum reservable bandwidth of link and each priority not to be reserved bandwidth particularly important.
Every equipment is collected the TE relevant information of one's respective area or every link of this rank all devices, the generation traffic engineering database (TEDB, TE DataBase).
2, calculating path
Use the Routing Protocol of link-state algorithm to calculate the shortest path that arrives each node of network by SPF (SPF, Shortest Path First) algorithm.
MPLS TE uses SPF (CSPF, the Constraint-based Shortest Path First) algorithm of Constraint-based to calculate the shortest path that satisfies TE attribute specification that arrives certain node.
The CSPF algorithm is derivative next from the SPF algorithm, and CSPF has two initial conditions:
1) need the bandwidth of the LSP of foundation, painted, foundation/keep the constraintss such as priority, explicit path, these are all in the configuration of the porch of LSP.
2)TEDB。
The computational process of CSPF is exactly to require first the link among the TEDB to be sheared for LSP, and the link that does not satisfy TE attribute specification is cut; Adopt again the SPF algorithm to seek the shortest path that satisfies TE attribute specification to the LSP Ingress node.
3, set up the path
The signaling that lsp tunnel is set up in support comprises CR-LDP and the RSVP (RSVP-TE, Resource Reserved Vation Protocol-Traffic Engineering) of expanding based on traffic engineering.They can both carry the bandwidth, part explicit route of LSP, the constrained parameters such as painted, and the function that both finish is the same.
Realize internally, CR-LDP sets up LSP by transmission control protocol (TCP, Transmission Control Protocol), and RSVP-TE then sets up LSP by original (Raw) IP and connects.
The RSVP technology has experienced development for many years, its architecture, protocols and relatively ripe to the support scheme of miscellaneous service; CR-LDP then is new technology, and is with the obvious advantage aspect extensibility.
4, E-Packet
Use the tunnel of setting up to E-Packet.
The technical scheme of at present, carrying out bandwidth resources optimization is MPLS TE.Fig. 1 has provided the networking diagram of MPLS TE, common IGP Route Selection is on the basis of not realizing MPLS TE, and take Fig. 1 as example, meeting is with SPF, need to arrive DUT4 if namely there is customer flow to enter from DUT1, will select shortest path DUT1-DUT5-DUT4 so.This link only has the bandwidth of 50M, because the essence of route learning, the triggering route is heavy not preferred as long as this link breaks down, so no matter there are how many customer flows all can select this link, so just be easy to occur the situation of network congestion, the flow of DUT1-DUT5-DUT4 is very busy, and DUT1-DUT2-DUT3-DUT4 has the bandwidth of 100M very idle.Certainly route can allow flow switch to the link of above-mentioned 100M by revising cost (COST) value mode, but so just again all flows has all been switched, and still can not carry out load balancing.And MPLS TE just can carry out dynamic adjustments according to the amount of bandwidth of current every link, rationally carries out load balancing, some link can not occur busy especially, the situation that some link is very idle.
The process that the MPLS traffic engineering tunnel is set up is as follows:
Step 01: to existing Routing Protocol as: OSPF, IS-IS expand, make it when the issue routing iinformation, carry the information such as maximum link bandwidth, maximum reservable bandwidth, so that more full and accurate ground of the router awareness network situation in the network, and according to network condition planning traffic forwarding path.
Step 02: after having obtained the information such as routing iinformation, maximum link bandwidth, maximum reservable bandwidth, MPLS TE calculates the path of satisfying the flow bandwidth requirement by CSPF.
Step 03: set up LSP, and notice LSP the node of process reserve certain bandwidth resources.The LSP that sets up among the MPLSTE is called CR-LSP.
MPLS TE can by RSVP-TE and two kinds of tag distribution protocols of CR-LDP, set up LSP.As shown in Figure 1, set up DUT1-DUT5-DUT4, bandwidth is the traffic engineering tunnel of 40M, and take the RSVP-TE tag distribution protocol as example, CR-LSP sets up process and can simply be described as:
1) DUT1 produces path (Path) message of carrying the label solicited message, sends to DUT4 along the path hop-by-hop that calculates by CSPF.
2) after DUT4 receives Path message, generation is carried obligate information such as above-mentioned bandwidth 40M and the reservation of label (Resv) message, the reverse path hop-by-hop that sends along Path message returns DUT1, simultaneously, Resv message is carried out resource reservation on LSR (LSR, Label Switched Router) on the way.
3) when DUT1 received Resv message, CR-LSP was successfully established.
Step 04: traffic forwarding, namely search Label Forwarding Information Base, switch label, E-Packet along the CR-LSP that sets up.
The shortcoming of prior art is as follows:
In existing the realization, need to guarantee when MPLS TE sets up traffic engineering tunnel that the bandwidth of every the link on the link of setting up all satisfies the user bandwidth demand, if there is a link not satisfy user bandwidth, traffic engineering tunnel just can not set up.As shown in Figure 1:
The traffic engineering tunnel that to set up first a bandwidth on the DUT1 be 70M is to destination DUT4, according to the algorithm of setting up at present traffic engineering tunnel, selects the path of satisfying bandwidth, i.e. DUT1-DUT2-DUT3-DUT4; The traffic engineering tunnel that to set up a bandwidth on the DUT1 be 40M again is to destination DUT4, and is same according to the algorithm of setting up at present traffic engineering tunnel, selects the path of satisfying bandwidth, i.e. DUT1-DUT5-DUT4.In realizing at present, if have another one the user, need to be in the DUT1 traffic engineering tunnel that to set up a bandwidth be 40M, then this requirement can't be satisfied, because there has not been the bandwidth of a link to have satisfied such bandwidth demand, this just becomes the bottleneck that present technology realizes.
Summary of the invention
The invention provides tunnel establishing method and device, more reasonably to utilize the bandwidth resources in the network.
Technical scheme of the present invention is achieved in that
A kind of tunnel establishing method, the method comprises:
Detect and configured required bandwidth and destination sign on the local tunnel interface;
Attempt setting up the TE tunnel of traffic engineering that satisfies required bandwidth from this locality to the destination;
Judge whether to be successfully established, if this flow process finishes; Otherwise, in all links from this locality to the destination, the link group of following condition is satisfied in selection: the remaining bandwidth sum of all links in this link group is not less than required bandwidth, utilizes the remaining bandwidth of every link of this link group to set up respectively a little tunnel of TE;
The little tunnel binding of all TE that to set up in this link group is a large tunnel of TE.
Described will be further to comprise after the large tunnel of TE at the little tunnel binding of all TE that this link group is set up:
To entering the flow of described local tunnel interface, carry out load balancing according to the bandwidth value in every little tunnel of TE.
Described in all links from this locality to the destination, the link group of following condition is satisfied in selection: the remaining bandwidth sum of all links in this link group is not less than required bandwidth, utilizes the remaining bandwidth of every link of this link group to set up respectively a little tunnel of TE and comprises:
A: the little building tunnel factor of initialization TE count makes count=0;
B: calculate b=a/2
Count+1, wherein a is required bandwidth, b is the bandwidth in the little tunnel of pre-established TE;
C: attempt setting up the little tunnel of the TE that satisfies bandwidth b from this locality to the destination, judge whether to be successfully established, if, execution in step D; Otherwise, Count is added 1, return step B;
D: the bandwidth sum that judges whether all little tunnels of built vertical TE is not less than a, if, determine that traffic engineering tunnel is set up to finish, all little tunnels of the TE that has set up are bundled into a large tunnel of TE; Otherwise, return step C.
Among the step C, when judge setting up success, Count is added 1 further comprises before returning step B:
Whether judge the building tunnel number of times less than default building tunnel maximum times, if carry out and described Count is added 1 action of returning step B; Otherwise, determine this building tunnel failure.
Among the step D, when the bandwidth sum of judging all little tunnels of built vertical TE during less than a, further comprise before returning step C:
Judge that whether the building tunnel number of times is less than default building tunnel maximum times, if carry out the described action of returning step C; Otherwise, determine this building tunnel failure.
Described traffic engineering tunnel is multiprotocol label switching MPLS traffic engineering tunnel.
A kind of building tunnel device, this device comprises:
Detection module: detect and configured required bandwidth and destination sign on the local tunnel interface, required bandwidth and destination sign are sent to bandwidth allocation module;
Bandwidth allocation module: receive required bandwidth and destination sign, attempt setting up the traffic engineering tunnel that satisfies required bandwidth from this locality to the destination; Judge whether to be successfully established, if this flow process finishes; Otherwise, in all links from this locality to the destination, the link group of following condition is satisfied in selection: the remaining bandwidth sum of all links in this link group is not less than required bandwidth, utilizes the remaining bandwidth of every link of this link group to set up respectively a little tunnel of TE; The little tunnel binding of all TE that to set up in this link group is a large tunnel of TE.
Described device further comprises: forwarding module,
Described bandwidth allocation module is further used for, and sign and the bandwidth value in every little tunnel of TE sent to forwarding module,
Described forwarding module receives sign and the bandwidth value in described every little tunnel of TE, to entering the flow of described local tunnel interface, carries out load balancing according to the bandwidth value in every little tunnel of TE.
Described bandwidth allocation module comprises:
Computing module: receive required bandwidth and destination sign, required bandwidth and destination sign is transmitted to sets up module, and the little building tunnel factor of initialization TE count, make count=0, calculate b=a/2
Count+1, b outputed to sets up module; When receiving when setting up the renewal count indication that module sends, count is added 1, recomputate b=a/2
Count+1, b outputed to sets up module; Wherein, a is required bandwidth, and b is the bandwidth in the little tunnel of pre-established TE;
Set up module: receive required bandwidth and destination sign, when receiving the b that computing module sends, attempt setting up the little tunnel of the TE that satisfies bandwidth b from this locality to the destination, judge whether to be successfully established, if success judges whether that the bandwidth sum in all little tunnels of built vertical TE is not less than a, if be not less than, determine that traffic engineering tunnel is set up and finish, all little tunnels of the TE that has set up are bundled into a large tunnel of TE; If less than, return and carry out the action that the little tunnel of TE that satisfies bandwidth b from this locality to the destination is set up in described trial; If failure sends renewal Count indication to computing module.
The described module of setting up is further used for, and when judgement is set up unsuccessfully, whether judges the building tunnel number of times less than default building tunnel maximum times, if carry out the described action of upgrading the Count indication that sends to computing module; Otherwise, determine this building tunnel failure; When the bandwidth sum of judging all little tunnels of built vertical TE during less than a, judge that whether the building tunnel number of times is less than default building tunnel maximum times, if carry out described returning and carry out the action that the little tunnel of TE that satisfies bandwidth b from this locality to the destination is set up in described trial; Otherwise, determine this building tunnel failure.
Compared with prior art, the present invention can realize finer bandwidth resource allocation, has more reasonably utilized the bandwidth resources in the network.
Description of drawings
Fig. 1 is the networking exemplary plot of existing MPLS TE;
Fig. 2 is the method flow diagram of setting up traffic engineering tunnel provided by the invention;
The method flow diagram of setting up traffic engineering tunnel that Fig. 3 provides for the embodiment of the invention one;
Fig. 4 is the composition diagram of traffic engineering tunnel apparatus for establishing provided by the invention.
Embodiment
The present invention is further described in more detail below in conjunction with drawings and the specific embodiments.
Fig. 2 is the method flow diagram of setting up traffic engineering tunnel provided by the invention, and as shown in Figure 2, its concrete steps are as follows:
Step 201: detect the user required bandwidth a of configure user and destination sign on local tunnel interface.
Step 202: attempt setting up the traffic engineering tunnel that satisfies bandwidth a from this locality to the destination.
Here, setting up traffic engineering tunnel can adopt any prior art to realize such as: CSPF algorithm.
Step 203: judge whether to be successfully established, if this flow process finishes; Otherwise, execution in step 204.
Step 204: judge that whether have the link group that satisfies following condition: the remaining bandwidth sum of all links in this link group is not less than bandwidth a in all links from this locality to the destination, if, execution in step 205; Otherwise, determine that traffic engineering tunnel sets up unsuccessfully, this flow process finishes.
Step 205: in the link group that satisfies condition, select a link group, carry out RSVP in the remaining bandwidth of every link of this link group, namely set up the little tunnel of TE, wherein, the bandwidth reserved sum in all links of this link group is a.
Here, when in the link group that satisfies condition, selecting a link group, can select arbitrarily one, also can preset a selective rule, according to one of this rules selection.
Step 206: will bind in the little tunnel of all TE that this link group is set up, and bind a large tunnel of TE.
Step 207: the flow to from the subscriber equipment of step 201, carry out load balancing according to the bandwidth in each little tunnel of bar TE.
For example: if set up m (m 〉=2) the little tunnel of bar TE, the bandwidth in each little tunnel of bar TE is respectively b
1, b
2, b
3..., b
M-1, b
m, then for the flow that comes from subscriber equipment, according to b
1: b
2: b
3: ...: b
M-1: b
mRatio share respectively the little tunnel 1 of TE, 2,3 ..., on m-1, the m.
The tunnel establishing method flow chart that Fig. 3 provides for the embodiment of the invention one, as shown in Figure 3, its concrete steps are as follows:
Step 301: control plane detects the user required bandwidth a of configure user and destination sign on local tunnel interface.
Control plane is the plane that can be used for finishing building tunnel.
Step 302: control plane uses the CSPF algorithm, attempts setting up the MPLS traffic engineering tunnel that satisfies bandwidth a from this locality to the destination.
Step 303: control plane judges whether to be successfully established, if this flow process finishes; Otherwise, execution in step 304.
Step 304: the little building tunnel factor of control plane initialization MPLS TE count makes count=0.
Step 305: control plane calculates b=a/2
Count+1, b is the bandwidth in the little tunnel of pre-established TE.
Step 306: control plane uses the CSPF algorithm, attempts setting up the little tunnel of MPLS TE that satisfies bandwidth b from this locality to the destination.
Step 307: control plane judges whether to be successfully established, if, execution in step 308; Otherwise, execution in step 311.
In the present embodiment, can preset a building tunnel maximum times n
MaxHere, before execution in step 311, judge first building tunnel number of times<n
MaxWhether set up, if, execution in step 311; Otherwise, determine that traffic engineering tunnel sets up unsuccessfully, this flow process finishes.
Step 308: control plane judges whether that the bandwidth sum in all little tunnels of built vertical MPLS TE is not less than a, if, execution in step 309; Otherwise, return step 306.
In the present embodiment, can preset a building tunnel maximum times n
MaxHere, before returning step 306, judge first building tunnel number of times<n
MaxWhether set up, if return step 306; Otherwise, determine that traffic engineering tunnel sets up unsuccessfully, this flow process finishes.
Step 309: control plane is determined that the MPLS traffic engineering tunnel is set up and is finished, and all little tunnels of MPLS TE of having set up are bundled into a large tunnel of MPLS TE.
Step 310: Forwarding plane is to the flow from the subscriber equipment of step 301, carries out load balancing according to the bandwidth value in each little tunnel of bar MPLS TE, and this flow process finishes.
Step 311: control plane adds 1 with Count, returns step 305.
Below provide application example embodiment illustrated in fig. 3:
As shown in Figure 1, the user need to walk the flow of 40M between DUT1~DUT4, and in the current network, and two links of DUT1-DUT2-DUT3-DUT4 and DUT1-DUT5-DUT4 are the bandwidth of remaining 30M and 10M respectively, and then to set up process as follows for traffic engineering tunnel:
01: calculate b=40M/2=20M, attempt setting up the little tunnel of article one TE that bandwidth is 20M, that is, according to the CSPF algorithm, calculate and satisfy bandwidth requirement to DUT1-DUT2-DUT3-DUT4, then carry out RSVP, obtain tunnel tunnel1, at this moment, bandwidth and be 20M.
02: attempt setting up the little tunnel of second TE that bandwidth is 20M, finding does not have link to satisfy bandwidth requirement, and then execution in step 03.
Step 03: calculate b=20M/2=10M, attempt setting up the second traffic engineering tunnel that bandwidth is 10M, calculate and satisfy bandwidth requirement to DUT1-DUT2-DUT3-DUT4, then carry out RSVP, obtain tunnel tunnel2, this Time Bandwidth and be 20+10=30M.
04: attempt setting up the 3rd the little tunnel of TE that bandwidth is 10M, find that DUT1-DUT5-DUT4 satisfies bandwidth requirement, then carry out RSVP, obtain tunnel tunnel3, this Time Bandwidth and be 40M.
Through said process, set up respectively three little tunnels of TE of 20M, 10M, 10M, the total bandwidth 40M in three little tunnels of TE satisfies the user bandwidth requirement, and these three little tunnels of TE are tied to a large tunnel of TE.Then the flow of on Forwarding plane this user being sent is made 2: 1: 1 load balancing, namely generate 4 prefix forwarding-table items, wherein, the outgoing interface of 2 prefix entries points to tunnel1, the outgoing interface of 1 prefix entries points to tunnel2, and the outgoing interface of 1 prefix entries points to tunnel2; The different characteristic of the stream of sending according to the user more all is assigned to all levellings on these 4 prefix entries, practical manifestation out be exactly with 2: 1: 1 ratio flow to be sent into the little tunnel of these three TE to transmit.
In actual applications, also can be by the user according to the remaining bandwidth of current each link and user's bandwidth demand, the bandwidth in the number in the little tunnel of manual configuration TE and each little tunnel of bar TE, and sign and the bandwidth value in each little tunnel of bar TE be input to control plane, control plane becomes a large tunnel of TE with the little tunnel binding of each bar TE.
For example: the user need to walk the flow of 40M between DUT1~DUT4, and in the current network, and two links of DUT1-DUT2-DUT3-DUT4 and DUT1-DUT5-DUT4 are the bandwidth of remaining 30M and 10M respectively, and then to set up process as follows for traffic engineering tunnel:
The user configures following information at the tunnel interface of DUT1:
Tunnel 1 (tunnel1): bandwidth is that 30M, destination are: DUT4;
Tunnel 2 (tunnel2): bandwidth is that 10M, destination are: DUT4.
Control plane can calculate the bandwidth demand that link DUT1-DUT2-DUT3-DUT4 satisfies tunnel 1 according to above-mentioned configuration information, and link DUT1-DUT5-DUT4 satisfies the bandwidth demand in tunnel 2, and a large tunnel is bound in tunnel 1,2.Then the flow of on Forwarding plane this user being sent is made 3: 1 load balancing, namely generates 4 prefix forwarding-table items, and wherein, the outgoing interface of 3 prefix entries points to tunnel1, and the outgoing interface of 1 prefix entries points to tunnel2; The different characteristic of the stream of sending according to the user more all is assigned to all levellings on these 4 prefix entries, practical manifestation out be exactly with 3: 1 ratio flow to be sent into these two little tunnels to transmit.
In addition, identical with the processing of existing traffic engineering tunnel when the little tunnel of TE breaks down or is preempted among the present invention, namely trigger and rebuild the little tunnel of this TE, if reconstruction failure then triggers again and rebuilds the large tunnel of TE.
The composition diagram of the building tunnel device that Fig. 4 provides for the embodiment of the invention, as shown in Figure 4, it mainly comprises: detection module 41 and bandwidth allocation module 42, wherein:
Detection module 41: detect the user and configured the required bandwidth of user and destination sign at local tunnel interface, the required bandwidth of user and destination sign are sent to bandwidth allocation module 42.
Bandwidth allocation module 42: receive the required bandwidth of user and destination sign that detection module 41 is sent, attempt setting up the traffic engineering tunnel that satisfies the required bandwidth of user from this locality to the destination; Judge whether to be successfully established, if this flow process finishes; Otherwise, in all links from this locality to the destination, the link group of following condition is satisfied in selection: the remaining bandwidth sum of all links in this link group is not less than the required bandwidth of user, utilizes the remaining bandwidth of every link of this link group to set up respectively a little tunnel of TE; The little tunnel binding of all TE that to set up in this link group is a large tunnel of TE.
Device shown in Figure 4 further comprises forwarding module, and bandwidth allocation module 42 is further used for, and sign and the bandwidth value in every little tunnel of TE sent to forwarding module; Forwarding module receives sign and the bandwidth value in this every the little tunnel of TE, to the flow from subscriber equipment, carries out load balancing according to the bandwidth value in every little tunnel of TE.
In actual applications, bandwidth allocation module 42 can comprise: computing module and set up module, wherein:
Computing module: receive the bandwidth request that detection module 41 is sent, this bandwidth request is transmitted to sets up module, and the little building tunnel factor of initialization TE count, make count=0, calculate b=a/2
Count+1, b outputed to sets up module; When receiving when setting up the renewal count indication that module sends, count is added 1, recomputate b=a/2
Count+1, b outputed to sets up module; Wherein, a is the required bandwidth of user.
Set up module: receive the bandwidth request that computing module is sent, when receiving the b that computing module sends, attempt setting up the little tunnel of the TE that satisfies bandwidth b from this locality to the destination, judge whether to be successfully established, if success judges whether that the bandwidth sum in all little tunnels of built vertical TE is not less than a, if be not less than, determine that traffic engineering tunnel is set up and finish, all little tunnels of the TE that has set up are bundled into a large tunnel of TE; If less than, return and carry out the action that the little tunnel of TE that satisfies bandwidth b from this locality to the destination is set up in described trial; If failure sends renewal Count indication to computing module.
Set up module and be further used for, when judgement is set up unsuccessfully, whether judge the building tunnel number of times less than default building tunnel maximum times, if carry out the described action of upgrading the Count indication that sends to computing module; Otherwise, determine this building tunnel failure; When the bandwidth sum of judging all little tunnels of built vertical TE during less than a, judge that whether the building tunnel number of times is less than default building tunnel maximum times, if carry out described returning and carry out the action that the little tunnel of TE that satisfies bandwidth b from this locality to the destination is set up in described trial; Otherwise, determine this building tunnel failure.
The above only is preferred embodiment of the present invention, and is in order to limit the present invention, within the spirit and principles in the present invention not all, any modification of making, is equal to replacement, improvement etc., all should be included within the scope of protection of the invention.