CN1960342A - Method for putting apart a bandwidth in advance, and equipment for implementing bandwidth put apart in advance - Google Patents
Method for putting apart a bandwidth in advance, and equipment for implementing bandwidth put apart in advance Download PDFInfo
- Publication number
- CN1960342A CN1960342A CNA2006101606175A CN200610160617A CN1960342A CN 1960342 A CN1960342 A CN 1960342A CN A2006101606175 A CNA2006101606175 A CN A2006101606175A CN 200610160617 A CN200610160617 A CN 200610160617A CN 1960342 A CN1960342 A CN 1960342A
- Authority
- CN
- China
- Prior art keywords
- bandwidth
- queues
- bandwidth resources
- scheduling
- formation
- 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.)
- Pending
Links
Images
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The method comprises: in the first, making traffic shaping for the normal queue group; the overall bandwidth of said traffic shaping equals or is less than the difference between the overall bandwidth of the output port and the reserved overall bandwidth of the bandwidth ensured resource queue group; making traffic scheduling for the normal queue group and the queues at each output port composes of bandwidth ensured resource queue group.
Description
Technical field
The present invention relates to the Resource Allocation in Networks technical field, because the device that relates to a kind of RSVP method and realize RSVP.
Background technology
Network congestion is the subject matter that influences the backbone network performance.Congested reason may be the Internet resources deficiency, also may be that network resource loads is unbalanced, causes local congestion.In order to guarantee the service quality of some premium-class traffics, can keep certain bandwidth in advance on the network equipment, for those data flow, thereby even guarantee that these data flow also can take the bandwidth of reservation under congestion situation.
The RSVP technology has obtained application in a lot of data communication field, at present, MPLS TE (Multiprotocol Label Switch Traffic Engineering, multi-protocol label switching traffic engineering) the strongest to the needs of RSVP technology, therefore, be example only below, introduce existing RSVP technical scheme with the RSVP technology among the MPLS TE.MPLS TE combines MPLS technology and TE technology, by for the lsp tunnel that arrives specified path carries out RSVP, guarantees the bandwidth of the lsp tunnel that priority is high, makes network traffics get around congested node, reaches the purpose of balancing network flow.
Before the existing RSVP technical scheme of concrete introduction, at first introduce several basic conceptions relevant: SHAPING (traffic shaping), WRR scheduling and SP scheduling with the RSVP technology.Wherein, " SHAPING " is meant on the network equipment flow done shaping, exceed the part delayed delivery of designated bandwidth, guarantee that flow meets the designated bandwidth requirement; " WRR scheduling " is meant a kind of dispatching method on the network equipment, be referred to as the weights polling dispatching, by specifying certain weight to different queue, guarantee to dispatch these formations according to the ratio of weights, if some formations do not have message, other formations can be according to the weights of other formations than sharing the shared bandwidth of these formation weights so." SP scheduling " is meant a kind of dispatching method on the network equipment, is referred to as strict priority scheduling, by giving the different flow assigned priority, comes the flow of strict guarantee high priority to obtain scheduling before the flow of low level earlier.
Below in conjunction with the schematic diagram directly perceived of bandwidth reservation approach in the existing MPLS TE technology shown in Figure 1, carry out more specifically existing RSVP technical scheme introduction.
Shown in the figure 0 to n formation be pairing each scheduling queue of outbound port, it is divided into two groups generally: common queue group (0 formation) and assurance bandwidth resources set of queues (1 to n formation), each formation in the assurance bandwidth resources set of queues is corresponding to a lsp tunnel that needs RSVP.With regard to scheduling mode, adopt the SP scheduling, wherein, the priority of common queue group is minimum, guarantees that the individual queue priority in the bandwidth resources set of queues all is higher than the priority of common queue group.And, all do the SHAPING shaping to guaranteeing each formation in the bandwidth resources set of queues, the amount of bandwidth of shaping is exactly the required RSVP value of corresponding lsp tunnel, thereby guarantees that the bandwidth that each lsp tunnel data flow takies can be above the restriction of RSVP value.
Be higher than the common queue group owing to guarantee the SP dispatching priority of bandwidth resources set of queues, even therefore under congestion situation, pairing each lsp tunnel of assurance bandwidth resources set of queues that priority is high also can obtain priority scheduling.Further, all done the SHAPING shaping owing to guarantee in the bandwidth resources set of queues each formation again,, can guarantee that also each lsp tunnel data flow can both take the bandwidth of reserving separately even therefore under congestion situation based on RSVP value separately.
This shows that above-mentioned existing bandwidth reservation approach can satisfy the demand of bandwidth reserved, even under congestion situation, those pairing lsp tunnel data flow of formation after by the SHAPING shaping still can take the bandwidth of reserving separately (being the size of shaping).But, also just because of these pairing formations of lsp tunnel that need guarantee bandwidth resources respectively by the SHAPING shaping, make the bandwidth of any one lsp tunnel data flow wherein can surpass the restriction of RSVP value at no time.Therefore, a part of lsp tunnel bandwidth is in the time of the free time in above-mentioned each lsp tunnel, and all the other lsp tunnels can not take the bandwidth of described idle lsp tunnel, thereby causes the bandwidth resources waste.
Summary of the invention
The object of the present invention is to provide a kind of method and apparatus of RSVP, can either guarantee when congested that the RSVP of premium-class traffic needs, can make full use of the idle bandwidth resource again, improve the bandwidth resources utilance.
For solving the problems of the technologies described above, the objective of the invention is to be achieved through the following technical solutions: a kind of RSVP method, described method comprises: the common queue group is done traffic shaping, and the total bandwidth of described traffic shaping is equal to or less than the poor of outbound port total bandwidth and the reservation total bandwidth that guarantees the bandwidth resources set of queues; Flow scheduling is carried out in each outbound port formation to common queue group and assurance bandwidth resources set of queues formation.
Preferably, described assurance bandwidth resources set of queues comprises a plurality of formations, guarantees that employing is based on the flow scheduling of weights between the bandwidth resources formation at each, and the ratio of individual queue weights equals the ratio of individual queue reserved bandwidth value in the described assurance bandwidth resources set of queues.
Preferably, adopt strict priority SP flow scheduling between described common queue group and assurance bandwidth resources set of queues, described assurance bandwidth resources set of queues is done as a whole when participating in the SP scheduling.
Preferably, in described common queue group with guarantee to adopt between the bandwidth resources set of queues flow scheduling based on weights.
Preferably, described common queue group comprises a plurality of formations, and the step of the common queue group being done traffic shaping is specially:
Traffic shaping is done in each formation in the described common queue group, and the bandwidth sum of individual queue traffic shaping is equal to or less than the poor of outbound port total bandwidth and the reservation total bandwidth that guarantees the bandwidth resources set of queues; Or
Described common queue group integral body is done traffic shaping, and the bandwidth of described traffic shaping is equal to or less than the poor of outbound port total bandwidth and the reservation total bandwidth that guarantees the bandwidth resources set of queues.
A kind of device of realizing RSVP comprises the bandwidth calculation unit, in order to calculate the bandwidth difference of port total bandwidth and the reservation total bandwidth that guarantees the bandwidth resources set of queues; The traffic shaping unit is done traffic shaping in order to the bandwidth difference that provides according to described bandwidth calculation unit to the common queue group, and the total bandwidth of described traffic shaping is equal to or less than above-mentioned bandwidth difference; Scheduling unit is in order to carry out flow scheduling to each outbound port formation of described common queue group and assurance bandwidth resources set of queues formation.
Preferably, described assurance bandwidth resources set of queues comprises a plurality of formations, described scheduling unit guarantees that the flow scheduling that adopts between the bandwidth resources formation is specially the flow scheduling based on weights at each, and the ratio of individual queue weights equals the ratio of individual queue reserved bandwidth value in the described assurance bandwidth resources set of queues.
Preferably, described scheduling unit is in the common queue group and guarantee that the flow scheduling that adopts between the bandwidth resources set of queues is specially strict priority SP flow scheduling, and described assurance bandwidth resources set of queues is done as a whole when participating in the SP scheduling.
Preferably, described scheduling unit is in the common queue group with guarantee that the flow scheduling that adopts between the bandwidth resources set of queues is specially the flow scheduling based on weights.
Preferably, described common queue group comprises a plurality of formations, and the total bandwidth of described common queue group traffic shaping is the bandwidth sum of each common queue traffic shaping; Or described common queue group is made the bandwidth of the as a whole traffic shaping that carries out.
Above technical scheme as can be seen, among the present invention by the common queue group is done traffic shaping, the total bandwidth of traffic shaping is equal to or less than the poor of outbound port total bandwidth and the reservation total bandwidth that guarantees the bandwidth resources set of queues, make under congestion situation, guarantee that the bandwidth resources set of queues need not to do traffic shaping, the message that enters common queue can not take the bandwidth reserved that guarantees the bandwidth resources set of queues yet.Just because of guaranteeing that the bandwidth resources set of queues need not to do traffic shaping, make when port has the remaining bandwidth resource, enter the message that guarantees the bandwidth resources set of queues and can take the remaining bandwidth resource, improved the port bandwidth resource utilization.
Description of drawings
Fig. 1 is the schematic diagram directly perceived of bandwidth reservation approach in the existing MPLS TE technology;
Fig. 2 is the schematic diagram directly perceived of bandwidth reservation approach first preferred implementation of the present invention;
Fig. 3 is the pairing RSVP method flow diagram of first preferred implementation shown in Figure 2;
Fig. 4 is the schematic diagram directly perceived of bandwidth reservation approach second preferred implementation of the present invention;
Fig. 5 is the pairing RSVP method flow diagram of second preferred implementation shown in Figure 4;
Fig. 6 is the schematic diagram directly perceived of bandwidth reservation approach the 3rd preferred implementation of the present invention;
Fig. 7 is the pairing RSVP method flow diagram of the 3rd preferred implementation shown in Figure 4;
Fig. 8 is the schematic diagram directly perceived of bandwidth reservation approach the 4th preferred implementation of the present invention;
Fig. 9 is the pairing RSVP method flow diagram of the 4th preferred implementation shown in Figure 4;
Figure 10 realizes the device example structure schematic diagram of RSVP for the present invention.
Embodiment
Technical solution of the present invention is further disclosed below in conjunction with a plurality of preferred implementations of the present invention and corresponding accompanying drawing.Need explanation, though following each embodiment still guarantees that as needs the data flow of bandwidth resources (being bandwidth reserved) is that example describes with lsp tunnel in the MPLS TE technology, but those skilled in the art are to be appreciated that, technical solution of the present invention is not limited to this, and any have the field of RSVP demand all can adopt technical scheme provided by the invention.
See also Fig. 2, it is for the schematic diagram directly perceived of bandwidth reservation approach first preferred implementation of the present invention.As can be seen from the figure, outbound port is to there being 0 to n outbound port formation, and wherein, 0 formation is common queue (being called the common queue group), uses for the data flow that does not need bandwidth reserved; 1 to n formation is used for the data flow that respectively needs bandwidth reserved for guaranteeing bandwidth resources formation (be called and guarantee the bandwidth resources set of queues).This preferred implementation is actual to have comprised two specific embodiments, and please in the lump in conjunction with consulting Fig. 3, it is the pairing RSVP method flow diagram of first preferred implementation shown in Figure 2 for this reason.
Step 310: every bandwidth resources capacity addition that lsp tunnel need guarantee on the outbound port, calculate the summation of the bandwidth resources that need assurance, promptly guarantee the reservation total bandwidth that the bandwidth resources set of queues is required.Then, deduct the reservation total bandwidth, calculate the remaining bandwidth value with the total bandwidth on this outbound port.
Step 320: the remaining bandwidth value bandwidth value of this value (or less than) as the bandwidth that Shaping (with regard to shape) shaping is carried out in 0 formation, has been prevented that thus common message from too much taking the bandwidth that guarantees the bandwidth resources formation.Guarantee that the individual queue in the bandwidth resources set of queues does not need to carry out traffic shaping.
Step 330: in 0 formation and do to carry out between the as a whole assurance bandwidth resources set of queues SP scheduling, further, guaranteeing that bandwidth resources set of queues inside carries out the flow scheduling based on weights.Wherein, guarantee that the ratio of individual queue weights in the bandwidth resources set of queues equals the ratio of individual queue reserved bandwidth value.In other words, can think that all outbound port formations carry out two-layer scheduling, first level is to 0 formation and does to carry out between the as a whole assurance bandwidth resources set of queues SP scheduling; Then, when assurance bandwidth resources set of queues is dispatched, the concrete flow scheduling that adopts based on weights.
As for the SP dispatching priority of 0 formation with guarantee to concern that not restriction can be that the former is lower than the latter in the present embodiment, also can be that the former is higher than the latter between the SP dispatching priority of bandwidth resources set of queues.
Have multiplely based on the flow scheduling mode of weights, therefore guarantee that respectively the weights set-up mode of bandwidth resources formation is also varied.For example, in a kind of flow scheduling WRR scheduling based on weights, each guarantees that the weights of bandwidth resources formation are exactly the ratio of their reserved bandwidth values.Particularly, suppose that outbound port has 3 lsp tunnel data flow and needs bandwidth reserved, the bandwidth that they need to reserve separately is respectively 200Mbps, 100Mbps, 200Mbps, so the weights of the assurance bandwidth resources formation of these 3 lsp tunnel correspondences are respectively 2,1,2.Find out easily, also these three weights that guarantee the bandwidth resources formation can be set to respectively: 40%, 20%, 40%; Perhaps 2/5,1/5,2/5.This shows, though have multiple based on the concrete mode of the flow scheduling of weights, the setting that causes guaranteeing bandwidth resources individual queue weights also has multiple concrete mode, still, equals the ratio of individual queue reserved bandwidth value as long as guarantee the ratio of bandwidth resources individual queue weights.
Implementation by above three steps as can be seen, after having come, the common message that need not to guarantee the RSVP resource can enter into 0 formation, because by traffic shaping is carried out in 0 formation, limited the shared bandwidth of these messages, even the SP dispatching priority of 0 formation is than the SP dispatching priority height that guarantees the bandwidth resources set of queues, these common messages also can't take other bandwidth that guarantees the bandwidth resources formation when outbound port is congested.After needs guarantee that the lsp tunnel message of RSVP resource enters into 1 to n part or all of formation, if it is congested that port takes place, because the weights method of salary distribution of WRR scheduling mode and individual queue, can guarantee provides its required bandwidth reserved for each guarantees the pairing lsp tunnel data flow of bandwidth resources formation.If it is congested that port does not take place, also has residual bandwidth resources, guarantee do not have lsp tunnel data flow to be sent in the bandwidth resources formation as some, and/or 0 do not have message to be sent in the formation, then all the other guarantee that lsp tunnel message in bandwidth resources formations just can be dispatched based on WRR and take above-mentioned residual bandwidth resources, maximizedly utilize whole port bandwidth resource.This shows that the technical scheme that present embodiment provides both can guarantee when congested that the RSVP of premium-class traffic needed, and can make full use of the idle bandwidth resource again, improved the bandwidth resources utilance.
In addition, also has a kind of replacement scheme (i.e. second specific embodiment in first preferred implementation) for above-mentioned specific embodiment, adopt the flow scheduling based on weights to replace aforementioned SP scheduling between 0 formation and assurance bandwidth resources set of queues (looking as a whole), all the other technical schemes remain unchanged.Owing to the inflow-rate of water turbine shaping is carried out in 0 formation, therefore, no matter 0 formation and guarantee how to carry out between the bandwidth resources set of queues weights and distribute, even the weights of 0 queue assignment are very big, the shared bandwidth of common message that enters 0 formation also has no idea to surpass the maximum of shaping, therefore, under congestion situation, the message in 0 formation can not take the reservation total bandwidth that guarantees the bandwidth resources set of queues.Therefore, equal the ratio of individual queue reserved bandwidth value as long as guarantee the ratio of each weights in the bandwidth resources set of queues, so just can guarantee provides its required bandwidth resource for each guarantees the pairing lsp tunnel data flow of bandwidth resources formation.Under non-congestion situation, need to guarantee that the lsp tunnel message of bandwidth resources still can take the remaining bandwidth resource based on the WRR scheduling, maximizedly utilize whole port bandwidth resource.
See also Fig. 4, it is for the schematic diagram directly perceived of bandwidth reservation approach second preferred implementation of the present invention.
Comparison diagram 2 and Fig. 4 as can be seen, both difference parts are: the common queue group comprises a plurality of common queue (only being among Fig. 2).Owing to be provided with a plurality of formations in the common queue group, therefore can carry out more flexile scheduling to common message.For step identical in two execution modes only in brief, will not give unnecessary details, describe the step of difference emphatically.See also Fig. 5, it is the RSVP method flow diagram of the present invention's second preferred implementation correspondence shown in Figure 4.It is noted that actually in the method flow diagram shown in Figure 4 to have comprised a plurality of specific embodiments, as multiple viable option is arranged in step 520 and the step 530.
Step 510: every bandwidth resources capacity addition that lsp tunnel need guarantee on the outbound port, calculate the summation of the bandwidth resources that need assurance, promptly guarantee the reservation total bandwidth that the bandwidth resources set of queues is required.Then, deduct the reservation total bandwidth, calculate the remaining bandwidth value with the total bandwidth on this outbound port.
Step 520: the total bandwidth of the remaining bandwidth value bandwidth value of this value (or less than) being carried out traffic shaping as the common queue group.Guarantee that individual queue does not need to carry out traffic shaping in the bandwidth resources set of queues, particularly, each guarantees that bandwidth resources formation bandwidth reserved separately need not to do shaping.Because the common queue group comprises a plurality of common queue, therefore when being done traffic shaping, the common queue group at least dual mode can be arranged.Even this dual mode can guarantee message in the common queue group under congestion situation, can not take the bandwidth resource that guarantees the bandwidth resources set of queues yet.
One all carries out traffic shaping to each common queue, and the bandwidth that each common queue is done traffic shaping can be provided with according to actual needs, there is no concrete restriction.But the bandwidth sum that will guarantee all common queue traffic shapings is equal to or less than the poor of outbound port total bandwidth and the reservation total bandwidth that guarantees the bandwidth resources set of queues.
They are two years old, look the common queue group as a whole, promptly each common queue is not done traffic shaping, but the common queue group is done whole traffic shaping, the bandwidth of described traffic shaping is equal to or less than the poor of outbound port total bandwidth and the reservation total bandwidth that guarantees the bandwidth resources set of queues.In other words, to the not concrete restriction of the shared bandwidth of each common queue data flow, can distribute according to concrete scheduling mode, but no matter adopt which kind of scheduling mode to dispatch individual queue in the common queue group, under congested or non-congestion situation, the total bandwidth that the data flow in the whole common queue group takies can not surpass the bandwidth of traffic shaping.
Step 530: making as a whole common queue group and doing to carry out between the as a whole assurance bandwidth resources set of queues SP scheduling or the WRR scheduling; Further, guaranteeing that bandwidth resources set of queues inside carries out the flow scheduling based on weights, wherein, guarantee that the ratio of individual queue weights in the bandwidth resources set of queues equals the ratio of individual queue reserved bandwidth value.Can adopt in common queue group inside SP scheduling or WRR scheduling or other scheduling modes all can, and there is no concrete restriction for the SP dispatching priority of each common queue or the weights of distribution.
Certainly, also can the common queue group not looked as a whole participation with the SP scheduling that guarantees between the bandwidth resources set of queues, but participate in directly with the SP scheduling that guarantees between the bandwidth resources set of queues, for example, the SP dispatching priority that 0 formation is set is second, the SP dispatching priority of 1 formation is the 4th, the SP dispatching priority of 2 formations is the 3rd, the SP dispatching priority that guarantees the bandwidth resources set of queues is first, so when carrying out flow scheduling, will priority scheduling enter the message (its inner WRR of employing scheduling) that guarantees the bandwidth resources set of queues, and then dispatch 0 formation successively, message in 2 formations and 1 formation.Generally speaking, as long as done traffic shaping at common queue, even the priority of messages of common queue is higher than the priority of RSVP formation also it doesn't matter so, common message still can't take and guarantee the required bandwidth reserved of bandwidth resources formation.
Because the total bandwidth that the common queue group is done traffic shaping is equal to or less than the poor of outbound port total bandwidth and the reservation total bandwidth that guarantees the bandwidth resources set of queues, also can not take the bandwidth that other guarantee the bandwidth resources formations even therefore can guarantee the common message that enters common queue when outbound port is congested.Simultaneously, because each guarantees that the ratio of the weights of bandwidth resources formation equals the ratio of individual queue reserved bandwidth value, even therefore under congestion situation, each guarantees that LSP tunnel message can take required separately bandwidth reserved in bandwidth resources formation.Further, under non-congestion situation, because therefore the flow scheduling that is based on weights that adopts between each assurance bandwidth resources formation can utilize abundant remaining bandwidth resource.This shows, how to dispatch for common queue group inside, can't influence the effect of bandwidth reserved method of the present invention: both can when congested, guarantee the RSVP needs of premium-class traffic, can make full use of the idle bandwidth resource again, improve the bandwidth resources utilance.
Earlier figures 2 and two execution modes of the present invention shown in Figure 4, what all adopt is two-layer scheduling mode: the scheduling between common queue group and the assurance bandwidth resources set of queues is a level; , guarantee the scheduling of bandwidth resources set of queues inside be a level, also may comprise the internal schedule of the common queue group of same level thereafter.Provide the embodiment of only utilizing a level scheduling below, see also Fig. 6, it is the schematic diagram directly perceived of bandwidth reservation approach the 3rd preferred implementation of the present invention.As can be seen from the figure, to common queue (0 formation) with guarantee that the bandwidth resources set of queues carries out the flow scheduling based on weights together.Please combination is referring to Fig. 7 in the lump, and it is the RSVP method flow diagram of the 3rd preferred implementation correspondence shown in Figure 6.
Step 710: every bandwidth resources capacity addition that lsp tunnel need guarantee on the outbound port, calculate the summation of the bandwidth resources that need assurance, promptly guarantee the reservation total bandwidth that the bandwidth resources set of queues is required.Then, deduct the reservation total bandwidth, calculate the remaining bandwidth value with the total bandwidth on this outbound port.
Step 720: the bandwidth of the remaining bandwidth value bandwidth value of this value (or less than) being carried out Shaping (with regard to shape) shaping as 0 formation.Guarantee that the individual queue in the bandwidth resources set of queues does not need to carry out traffic shaping.
Step 730: 0 formation and each are guaranteed that bandwidth resources formation carries out the flow scheduling based on weights, and each guarantees that the ratio of the weights of bandwidth resources formation equals the ratio of reserved bandwidth value separately.As for the not concrete restriction of the weights of distributing to 0 formation, because traffic shaping has been carried out in 0 formation, so no matter 0 formation weights what are, with the weights that guarantee the bandwidth resources formation such as what, enter the bandwidth resources that the message of 0 formation takies, can surpass the restriction of traffic shaping bandwidth at no time, in other words, even under congestion situation, can not take yet and guarantee the required reservation total bandwidth of bandwidth resources formation.Therefore, equal the ratio of reserved bandwidth value separately, just can guarantee that each lsp tunnel message can take required separately bandwidth resource at least as long as can guarantee respectively to guarantee the ratio of the weights of bandwidth resources formation.In addition, this execution mode also can extend to the situation that the common queue group comprises a plurality of common queue, with the situation of a common queue (0 formation) in like manner, the weights of each common queue distribute and there is no concrete restriction.
See also Fig. 8, it is for the schematic diagram directly perceived of bandwidth reservation approach the 3rd preferred implementation of the present invention.As can be seen from the figure, the common queue group comprises 0 to 2 three formation, guarantees that the bandwidth resources set of queues only comprises one 1 formation, directly adopts SP scheduling or WRR scheduling when carrying out flow scheduling.Please combination is referring to Fig. 9 in the lump, and it is the RSVP method flow diagram of the 4th preferred implementation correspondence shown in Figure 8.
Step 910: the total bandwidth on the outbound port is deducted bandwidth reserved (i.e. the required bandwidth reserved of 1 formation), calculate the remaining bandwidth value.
Step 920: the bandwidth of the remaining bandwidth value bandwidth value of this value (or less than) being carried out traffic shaping as the common queue group.1 formation does not need to carry out traffic shaping.When being carried out traffic shaping, the common queue group can adopt the described dual mode of step 520 all can.
Step 930: flow scheduling based on weights is carried out in each common queue and 1 formation that guarantees bandwidth resources.Perhaps carry out the SP scheduling, but need to guarantee that the SP scheduling priority of 1 formation is higher than the priority of each common queue.
When the flow scheduling that adopts based on weights, under congestion situation, owing to the common queue group has been carried out traffic shaping, the shared bandwidth resources of message that therefore enter the common queue group can not surpass the bandwidth of traffic shaping, and then have guaranteed the required bandwidth reserved of 1 formation.Under non-congestion situation, outbound port has the remaining bandwidth resource, as not having message to be sent in the part or all of formation in the common queue group, so because the existence of WRR scheduling mode makes that the lsp tunnel message in 1 formation just can take the remaining bandwidth resource.
When adopting the SP scheduling, under congestion situation, because the priority of 1 formation is the highest, therefore the lsp tunnel message in 1 formation of meeting priority scheduling can guarantee the bandwidth resource that it is required; Under non-congestion situation, outbound port has the remaining bandwidth resource, as not having message to be sent in the part or all of formation in the common queue group, the so same the highest quilt of priority owing to 1 formation is preferably dispatched, and not by traffic shaping, thereby make the lsp tunnel message that enters 1 formation can take the remaining bandwidth resource.
The preferred implementation schematic diagram directly perceived of Fig. 2, Fig. 4, Fig. 6 and bandwidth reservation approach of the present invention shown in Figure 8 can be achieved by a kind of RSVP device.For this reason, the invention also discloses a kind of device of realizing the existing RSVP of bandwidth, because this device and earlier figures 3, Fig. 5 and RSVP method shown in Figure 9 are based on same inventive concept, therefore both have identical or the corresponding techniques feature, and those skilled in the art are easy to realize RSVP device of the present invention on referring to the basis of preamble content.
Please referring to Figure 10, it is the structural representation of RSVP device preferred embodiment of the present invention.Described device comprises bandwidth calculation unit 11, traffic shaping unit 12 and scheduling unit 13.Below in conjunction with the operation principle of this device, further introduce its internal structure.This device can be realized previously described each preferred bandwidth reservation techniques scheme, when introducing each concrete part, can mention a plurality of preferred implementations, and particular content can be referring to corresponding techniques feature in each method execution mode of preamble.
At first, calculate the bandwidth difference of the reservation total bandwidth of port total bandwidth and assurance bandwidth resources set of queues by bandwidth calculation unit 11.Described assurance bandwidth resources set of queues can comprise the formation of one or more assurance bandwidth resources.If comprise the formation of a plurality of assurance bandwidth resources, then each guarantees that the bandwidth resources formation all has its own required bandwidth reserved, and each guarantees that the required bandwidth reserved sum of bandwidth resources formation is the reservation total bandwidth that guarantees the bandwidth resources set of queues.
Secondly, according to the bandwidth difference that bandwidth calculation unit 11 provides the common queue group is done traffic shaping by traffic shaping unit 12, the total bandwidth of described traffic shaping is equal to or less than above-mentioned bandwidth difference.The common queue group can comprise one or more common queue.When comprising a plurality of common queue, can do traffic shaping to each common queue respectively, no matter also can do whole traffic shaping to the common queue group, be which kind of, is equal to or less than above-mentioned bandwidth difference as long as guarantee the total bandwidth of common queue group traffic shaping.
At last, realize flow scheduling is carried out in each outbound port formation of common queue group and assurance bandwidth resources set of queues formation by scheduling unit.Concrete scheduling mode has multiple, and the scheme that can adopt two levels to dispatch also can only adopt a level to dispatch.
If adopt two levels to carry out flow scheduling, in the common queue group with guarantee to adopt between the bandwidth resources set of queues SP or based on the flow scheduling of weights, guaranteeing the inner flow scheduling that adopts based on weights of bandwidth resources set of queues, as long as guarantee that respectively the ratio of bandwidth resources formation weights equals the ratio of individual queue reserved bandwidth value.Whether dispatch, adopt which kind of concrete scheduling mode and each common queue SP dispatching priority setting or weights to distribute separately as for common queue group inside, all without limits.Because the total bandwidth that the common queue group is done traffic shaping is equal to or less than the poor of outbound port total bandwidth and the reservation total bandwidth that guarantees the bandwidth resources set of queues, when outbound port is congested, can not take the bandwidth that other guarantee the bandwidth resources formations even therefore can guarantee the common message that enters common queue yet.Simultaneously, because each guarantees that the ratio of the weights of bandwidth resources formation equals the ratio of individual queue reserved bandwidth value, even therefore under congestion situation, each guarantees that LSP tunnel message can take required separately bandwidth reserved in bandwidth resources formation.Further, under non-congestion situation, because therefore the flow scheduling that is based on weights that adopts between each assurance bandwidth resources formation can utilize abundant remaining bandwidth resource.This shows, how specifically to dispatch for common queue group inside, can't influence bandwidth reserved device of the present invention and wish the effect that reaches: both can when congested, guarantee the RSVP needs of premium-class traffic, and can make full use of the idle bandwidth resource again, and improve the bandwidth resources utilance.
If adopt a level to carry out flow scheduling, when guaranteeing that the bandwidth resources set of queues comprises a plurality of formation, can adopt flow scheduling to each outbound port formation (comprise common queue and guarantee the bandwidth resources formation), but need guarantee respectively that equally the ratio of bandwidth resources formation weights equals the ratio of individual queue reserved bandwidth value based on weights.When guaranteeing that the bandwidth resources set of queues includes only a formation, can adopt flow scheduling or SP to dispatch to each outbound port formation (comprise common queue and guarantee the bandwidth resources formation) based on weights, if adopt the SP scheduling, guarantee that then the priority of bandwidth resources formation will be higher than each common queue priority.Scheduling scheme shown in this section still can satisfy the beneficial effect of RSVP device of the present invention: both can guarantee the RSVP needs of premium-class traffic when congested, can have made full use of the idle bandwidth resource again, and improve the bandwidth resources utilance.
More than introduced the device of the realization RSVP under bandwidth reserved method of the present invention and the same inventive concept by a plurality of preferred implementations.Those skilled in the art are to be appreciated that, though be that example describes only in above-mentioned each preferred implementation with SP scheduling with based on the flow scheduling of weights, but be not limited thereto, guaranteeing that the bandwidth resources set of queues comprises under a plurality of formation situations, everyly not only can adopt the SP scheduling but also can replace with the flow scheduling part of employing based on weights, as the scheduling of common queue group inside or the scheduling between common queue group (looking as a whole) and the assurance bandwidth resources set of queues (looking as a whole), all can be suitable for other scheduling modes; Equal the ratio of bandwidth reserved separately as long as guarantee the ratio of the weights of individual queue in the bandwidth resources set of queues.
More than to RSVP method provided by the present invention and realize that the device of RSVP is described in detail, used specific case herein principle of the present invention and execution mode are set forth, the explanation of above embodiment just is used for helping to understand method of the present invention and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, the part that all can change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.
Claims (10)
1, a kind of RSVP method is characterized in that:
The common queue group is done traffic shaping, and the total bandwidth of described traffic shaping is equal to or less than the poor of outbound port total bandwidth and the reservation total bandwidth that guarantees the bandwidth resources set of queues;
Flow scheduling is carried out in each outbound port formation to common queue group and assurance bandwidth resources set of queues formation.
2, according to right 1 described method, it is characterized in that: described assurance bandwidth resources set of queues comprises a plurality of formations, guarantee that employing is based on the flow scheduling of weights between the bandwidth resources formation, the ratio of individual queue weights equals the ratio of individual queue reserved bandwidth value in the described assurance bandwidth resources set of queues at each.
3, according to right 2 described methods, it is characterized in that: adopt strict priority SP flow scheduling between described common queue group and assurance bandwidth resources set of queues, described assurance bandwidth resources set of queues is done as a whole when participating in the SP scheduling.
4, according to right 2 described methods, it is characterized in that: between described common queue group and assurance bandwidth resources set of queues, adopt flow scheduling based on weights.
5, according to any described method in the claim 1 to 4, it is characterized in that: described common queue group comprises a plurality of formations, and the step of the common queue group being done traffic shaping is specially:
Traffic shaping is done in each formation in the described common queue group, and the bandwidth sum of individual queue traffic shaping is equal to or less than the poor of outbound port total bandwidth and the reservation total bandwidth that guarantees the bandwidth resources set of queues; Or
Described common queue group integral body is done traffic shaping, and the bandwidth of described traffic shaping is equal to or less than the poor of outbound port total bandwidth and the reservation total bandwidth that guarantees the bandwidth resources set of queues.
6, a kind of device of realizing RSVP is characterized in that comprising:
The bandwidth calculation unit is in order to calculate the bandwidth difference of port total bandwidth and the reservation total bandwidth that guarantees the bandwidth resources set of queues;
The traffic shaping unit is done traffic shaping in order to the bandwidth difference that provides according to described bandwidth calculation unit to the common queue group, and the total bandwidth of described traffic shaping is equal to or less than above-mentioned bandwidth difference;
Scheduling unit is in order to carry out flow scheduling to each outbound port formation of described common queue group and assurance bandwidth resources set of queues formation.
7, according to right 6 described devices, it is characterized in that: described assurance bandwidth resources set of queues comprises a plurality of formations, described scheduling unit guarantees that the flow scheduling that adopts between the bandwidth resources formation is specially the flow scheduling based on weights at each, and the ratio of individual queue weights equals the ratio of individual queue reserved bandwidth value in the described assurance bandwidth resources set of queues.
8, according to right 7 described devices, it is characterized in that: described scheduling unit is in the common queue group and guarantee that the flow scheduling that adopts between the bandwidth resources set of queues is specially strict priority SP flow scheduling, and described assurance bandwidth resources set of queues is done as a whole when participating in the SP scheduling.
9, according to right 7 described devices, it is characterized in that: described scheduling unit is in the common queue group and guarantee that the flow scheduling that adopts between the bandwidth resources set of queues is specially the flow scheduling based on weights.
10, according to any described device in the claim 6 to 9, it is characterized in that: described common queue group comprises a plurality of formations,
The total bandwidth of described common queue group traffic shaping is the bandwidth sum of each common queue traffic shaping; Or described common queue group is made the bandwidth of the as a whole traffic shaping that carries out.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2006101606175A CN1960342A (en) | 2006-11-29 | 2006-11-29 | Method for putting apart a bandwidth in advance, and equipment for implementing bandwidth put apart in advance |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2006101606175A CN1960342A (en) | 2006-11-29 | 2006-11-29 | Method for putting apart a bandwidth in advance, and equipment for implementing bandwidth put apart in advance |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1960342A true CN1960342A (en) | 2007-05-09 |
Family
ID=38071837
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2006101606175A Pending CN1960342A (en) | 2006-11-29 | 2006-11-29 | Method for putting apart a bandwidth in advance, and equipment for implementing bandwidth put apart in advance |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1960342A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102014061A (en) * | 2010-11-25 | 2011-04-13 | 福建星网锐捷网络有限公司 | Upstream flow control method and device for intranet core switch |
| CN102594663A (en) * | 2012-02-01 | 2012-07-18 | 中兴通讯股份有限公司 | Queue scheduling method and device |
| CN102843296A (en) * | 2012-08-28 | 2012-12-26 | 深圳市共进电子股份有限公司 | Method for acceleration of designated terminal in local area network (LAN) and router |
| CN101616096B (en) * | 2009-07-31 | 2013-01-16 | 中兴通讯股份有限公司 | Method and device for scheduling queue |
| WO2016202097A1 (en) * | 2015-06-15 | 2016-12-22 | 中兴通讯股份有限公司 | Method for obtaining assured bandwidth of user-side port and network monitoring system |
| WO2017118422A1 (en) * | 2016-01-08 | 2017-07-13 | 中兴通讯股份有限公司 | Method and system for distributing preserved bandwidth of tunnel egress |
| CN109391555A (en) * | 2017-08-08 | 2019-02-26 | 迈普通信技术股份有限公司 | Method for dispatching message, device and communication equipment |
| CN111740922A (en) * | 2020-08-21 | 2020-10-02 | 浙江巨化信息技术有限公司 | Data transmission method, device, electronic equipment and medium |
-
2006
- 2006-11-29 CN CNA2006101606175A patent/CN1960342A/en active Pending
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101616096B (en) * | 2009-07-31 | 2013-01-16 | 中兴通讯股份有限公司 | Method and device for scheduling queue |
| CN102014061A (en) * | 2010-11-25 | 2011-04-13 | 福建星网锐捷网络有限公司 | Upstream flow control method and device for intranet core switch |
| CN102014061B (en) * | 2010-11-25 | 2012-11-21 | 福建星网锐捷网络有限公司 | Upstream flow control method and device for intranet core switch |
| CN102594663A (en) * | 2012-02-01 | 2012-07-18 | 中兴通讯股份有限公司 | Queue scheduling method and device |
| CN102843296A (en) * | 2012-08-28 | 2012-12-26 | 深圳市共进电子股份有限公司 | Method for acceleration of designated terminal in local area network (LAN) and router |
| CN102843296B (en) * | 2012-08-28 | 2015-08-26 | 深圳市共进电子股份有限公司 | A kind of method that in local area network (LAN), designated terminal accelerates and router |
| WO2016202097A1 (en) * | 2015-06-15 | 2016-12-22 | 中兴通讯股份有限公司 | Method for obtaining assured bandwidth of user-side port and network monitoring system |
| CN106330476A (en) * | 2015-06-15 | 2017-01-11 | 中兴通讯股份有限公司 | Method for acquiring user side port guaranteed bandwidth and network monitoring system |
| CN106330476B (en) * | 2015-06-15 | 2021-02-19 | 中兴通讯股份有限公司 | Method for obtaining guaranteed bandwidth of user side port and network monitoring system |
| WO2017118422A1 (en) * | 2016-01-08 | 2017-07-13 | 中兴通讯股份有限公司 | Method and system for distributing preserved bandwidth of tunnel egress |
| CN106961399A (en) * | 2016-01-08 | 2017-07-18 | 中兴通讯股份有限公司 | Tunnel exit port reserves the distribution method and system of bandwidth |
| CN106961399B (en) * | 2016-01-08 | 2021-01-08 | 中兴通讯股份有限公司 | Method and system for distributing reserved bandwidth of tunnel exit port |
| CN109391555A (en) * | 2017-08-08 | 2019-02-26 | 迈普通信技术股份有限公司 | Method for dispatching message, device and communication equipment |
| CN109391555B (en) * | 2017-08-08 | 2022-08-12 | 迈普通信技术股份有限公司 | Message scheduling method and device and communication equipment |
| CN111740922A (en) * | 2020-08-21 | 2020-10-02 | 浙江巨化信息技术有限公司 | Data transmission method, device, electronic equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1960342A (en) | Method for putting apart a bandwidth in advance, and equipment for implementing bandwidth put apart in advance | |
| CN100596359C (en) | Traffic control method and device for switch service flow | |
| CN101060473A (en) | A method for realizing the multicast/broadcast service resource reserve | |
| CN1848804A (en) | Data message transmitting method and node equipment based on MPLS TE tunnel | |
| CN1961544A (en) | Priority based multiplexing of data packet transport | |
| CN1809022A (en) | Ethernet device stacking method and its system | |
| CN1848769A (en) | Distribution-tuning mechanism for link aggregation group management | |
| CN1617521A (en) | Allocating bandwidth using resilient packet ring (RPR) fairness mechanism | |
| CN101052000A (en) | Method for realizing separate service in light burst exchange network | |
| CN1852246A (en) | Method for multi-cast broadcasting flow distribution node to receive multi-cast broadcasting flow | |
| CN101039468A (en) | Apparatus and method for scheduling service | |
| CN1968186A (en) | Message sending scheduling apparatus and method | |
| CN105959241B (en) | A kind of spaceborne photoelectricity hybrid switching United Dispatching method | |
| CN1859279A (en) | Method for dispatching variable length data packet queue in crossbar switching matrix | |
| CN1949701A (en) | Method, system and apparatus for service switching | |
| CN101034967A (en) | Multi-network interface reverse multiplexing method | |
| CN102594654A (en) | Method and device for queue scheduling | |
| CN1761228A (en) | Method for implementing transmission and protection for service in MPLS ring net | |
| CN1725729A (en) | Method for tranfer message on packet ring | |
| CN1874293A (en) | Information transmission method of optical network of automatic exchange | |
| CN1901519A (en) | Optimizing system and method for medium transmission on different transmission channels | |
| CN1735257A (en) | Method for allocating resources in communication system | |
| CN102769566A (en) | Method and device for configuring multilevel scheduling system, and method and device for changing configuration of multilevel scheduling system | |
| CN100345426C (en) | A push and pull based pure distributed data exchange method | |
| CN1512715A (en) | Method for establishing distributively controlled dynamic link in optical Internet |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |