CN101102275B - Method for multi-level dispatching on Ethernet switching chip - Google Patents
Method for multi-level dispatching on Ethernet switching chip Download PDFInfo
- Publication number
- CN101102275B CN101102275B CN2007101230034A CN200710123003A CN101102275B CN 101102275 B CN101102275 B CN 101102275B CN 2007101230034 A CN2007101230034 A CN 2007101230034A CN 200710123003 A CN200710123003 A CN 200710123003A CN 101102275 B CN101102275 B CN 101102275B
- Authority
- CN
- China
- Prior art keywords
- formation
- bandwidth
- level
- speed
- secondary 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 24
- 230000015572 biosynthetic process Effects 0.000 claims description 96
- 238000005755 formation reaction Methods 0.000 description 75
- 230000000694 effects Effects 0.000 description 9
- 238000005516 engineering process Methods 0.000 description 6
- 238000007726 management method Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000000205 computational method Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- HRULVFRXEOZUMJ-UHFFFAOYSA-K potassium;disodium;2-(4-chloro-2-methylphenoxy)propanoate;methyl-dioxido-oxo-$l^{5}-arsane Chemical compound [Na+].[Na+].[K+].C[As]([O-])([O-])=O.[O-]C(=O)C(C)OC1=CC=C(Cl)C=C1C HRULVFRXEOZUMJ-UHFFFAOYSA-K 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The method comprises: 1) setting total bandwidth for each level 1 queue; setting the ensured bandwidth and un-ensured bandwidth for each level 2 queue in each level 1 queue; setting a scheduling algorithm between all level 2 queues in same level 1 queue; 2) reading out the actual rate of level 1 queue level 2 queue at one moment; 3) sequentially calculating the final rate of each level 2 queue in each level 1 queue; 4) according to calculated the final rate of each level 2 queue, limiting the speed for the level 2 queue whose actual rate is more than the final rate; 5) waiting a preset time, restarting the scheduling from the step 2.
Description
Technical field
The present invention relates to the quality services technology of Ethernet switch, more specifically, relate to a kind of method that on Ethernet switching chip, realizes multi-stage scheduling.
Background technology
Along with the network application develop rapidly, novel business constantly occurs, for example, non-traditional data communication services such as ip voice and video conference, the unification of three nets that can merge voice, video, data is the trend of network development.For the voice with heterogeneous networks service request, video and data application communication, require the IP network core to have the ability of distinguishing different communication, and then different services is provided for it, reach the purpose of Differentiated Services.The appearance of QoS (Qualityof Service) service quality technology is to be devoted to the purpose that reaches Differentiated Services, guarantee service quality.QoS is intended to the different demands at various application, for it provides different service quality.
Present popular Diffserv model, mainly be to utilize message to be distinguished based on two layers COS (IEEE802.1p) field with based on the dscp field of IP layer, distribute different priority to it, when the network generation is congested, dispatch according to certain drop policy such as SP (Strict Priority), WRR (Weighted Round Robin) scheduling etc. selectively abandon the message of different priorities, thereby the service quality technology at different business is provided.
Along with further developing of broadband access technology, single dependence COS field or dscp field realize that service quality seems and more and more can not satisfy the demand of operator.Many times, operator is a kind of role who sells bandwidth, utilizes technology such as VLAN, sells bandwidth to different service providers; And in same service provider's packet, owing to also exist multiple business, its priority has nothing in common with each other, at this moment need carry out the quality services of Diffserv equally, this is actual to be that the requirement switching equipment provides a kind of multilevel queue scheduling function, and most Ethernet switching chip can only accomplish to be difficult to satisfy the demand of operator by the single-stage queue scheduling based on COS value or DSCP value.
On the other hand, though the most functions that speed limit is provided of Ethernet switching chip reach the QoS effect by the speed of controlling particular flow rate, this function often can only be carried out speed limit to the data flow of certain specific characteristic, as shown in Figure 1, be difficult to reach the effect of multistage speed limit.
Thereby, in order to address the above problem, be badly in need of a kind of method that can on Ethernet switching chip, realize multi-stage scheduling.
Summary of the invention
In order to address the above problem, the invention provides a kind of method that on Ethernet switching chip, realizes multi-stage scheduling, be used on Ethernet switching chip, realizing complicated traffic management.
The invention provides a kind of method that on Ethernet switching chip, realizes multi-stage scheduling, it may further comprise the steps: step 1, the total bandwidth of each one-level formation is set, and the dispatching algorithm between each secondary formation in the assurance bandwidth of each the secondary formation in the one-level formation, non-assured bandwidth and the same one-level formation; Step 2 reads one-level formation and the secondary formation actual speed rate a moment; Step 3 is calculated the final speed of each secondary formation in each one-level formation successively, and wherein, in step 3, calculating may further comprise the steps: deduct the part that surpasses the assurance bandwidth in all secondary formation actual flows from total bandwidth; According to predefined queue scheduling algorithm, remaining bandwidth is distributed to each secondary formation to satisfy the non-assured bandwidth of each secondary formation, obtain the assigned actual non-assured bandwidth of each secondary formation; And add actual non-assured bandwidth with the bandwidth that guarantees of each secondary formation, to obtain the final speed of secondary formation; Step 4 according to the final speed of each secondary formation that calculates, is carried out speed limit to actual speed rate greater than the secondary formation of final speed; And step 5, wait for the scheduled time, dispatch again from step 2 beginning.
In step 2, reading of actual speed rate is to be realized by counter that exchange chip provided.
In step 4, speed limit is that the speed limit function by exchange chip realizes.
Wherein, guarantee bandwidth is no matter whether network is congested, exchange chip guarantees that formation is accessible at most and goes out group speed.Non-assured bandwidth is that formation exceeds the maximum that the speed that guarantees bandwidth can reach.The secondary formation is other formation of at least one even lower level with hierarchical relationship.
Thereby, the method that is proposed above utilizing, can on common exchange chip, realize complicated traffic management, when the Business Stream of a plurality of service providers passes through same exchange chip, can be in the single service provider of restriction flow, guarantee that its inner different business carries out independently Differentiated Services according to the priority and the dispatching algorithm of in advance setting, thereby reached good effect that this is originally only to rely on the own dispatching office of chip to accomplish.In fact Shang Mian method can be promoted, and when the assurance bandwidth of all formations all is set to 0, just can reach the effect of coming bandwidth-hogging according to priority fully between the formation.On the other hand, top method is a kind of second-level dispatching, if demand is arranged, constraint can be set again, is certain such as two operators' Business Stream summation, and top final rate calculations method is changed a little, then can realize three grades of more complicated scheduling.
Other features and advantages of the present invention will be set forth in the following description, and, partly from specification, become apparent, perhaps understand by implementing the present invention.Purpose of the present invention and other advantages can realize and obtain by specifically noted structure in the specification of being write, claims and accompanying drawing.
Description of drawings
Accompanying drawing is used to provide further understanding of the present invention, and constitutes the part of specification, is used from explanation the present invention with embodiments of the invention one, is not construed as limiting the invention.In the accompanying drawings:
Fig. 1 is the schematic diagram of the single-stage scheduling method that provides of Ethernet switching chip of the prior art;
Fig. 2 is a schematic diagram of realizing the multi-stage scheduling pattern on the Ethernet switching chip among the present invention; And
Fig. 3 is a flow chart of realizing the method for multi-stage scheduling according to of the present invention on Ethernet switching chip.
Embodiment
Below in conjunction with accompanying drawing the preferred embodiments of the present invention are described, should be appreciated that preferred embodiment described herein only is used for description and interpretation the present invention, and be not used in qualification the present invention.
Technical problem to be solved by this invention is: realize having proposed a kind of method that software and hardware resources is realized multi-stage scheduling that fully utilizes by complicated traffic management on Ethernet switching chip.
At first, suppose on the switch, have two service providers to provide service by it simultaneously now, the bandwidth that Virtual network operator is sold to service provider (SP, service provider) A is 30Mbps, and the bandwidth of B is 20Mbps.With regard to traditional speed limit management method, usually have different feature (more commonly utilizing VLANID to distinguish) from different service provider traffics, at this time on exchange chip, utilize the function of speed limit, to other limiting speed 30Mbps of the flow point of two kinds of different characteristics and 20Mbps, also just can satisfy the demands.
In the time of but more, user's request is not limited thereto, because each service provider self may provide multiple different business, and voice for example, video and data or the like, the priority of these different business is also different usually.For service provider B, suppose voice, video, three kinds of service priority of data need carry out SP (strict priority) scheduling from high to low, and present three kinds of service traffics are 10Mbps, should have only preceding two kinds of Business Streams to pass through in theory, the 10Mbps data service should all be dropped, but is impossible such according to the effect of traditional speed limit, chip only can be from this 30Mbps flow at random selection 20Mbps give by.
Based on this, the present invention is by the following technical solutions: the physical transfer ability of a port is divided into several thick one-level formations, these one-level formations can make a distinction with different one-level stream features (for example VLANID or other data segment) between them corresponding to different service providers.In same one-level inner queue, can mark off some thin pipelines again, normally utilize COS value (or further feature, here be called the secondary streams feature) distinguish, miscellaneous service is included into different secondary formations, the quantity of the secondary formation that can support of system is here decided by the hardware resource of exchange chip, can select different dispatching algorithms such as SP, WRR etc. between the formation.
For each top secondary formation, use for reference the QoS thought of ATM exchange, introduce two parameters, guarantee bandwidth and non-assured bandwidth.The meaning that guarantees bandwidth be say no matter whether congested, exchange chip guarantees group speed that goes out that the multipotency of this formation reaches, the meaning of non-assured bandwidth be this formation exceed that a part of speed of guaranteeing bandwidth the maximum that can reach, but this part bandwidth is not ensured, if take place congested, the flow that surpass to guarantee bandwidth according to Congestion Level SPCC can be dropped some or even all.Most Ethernet switching chip all have tally function, utilize this function periodically to count the stream of special characteristic, calculate all secondary formations, the speed S0 of the Business Stream of just specific one-level inner queue ownership different queue.
According to each the secondary formation actual speed rate that records above and each formation bandwidth parameter setting, calculate, draw under the situation of multi-stage scheduling individual queue in this one-level formation and finally can allow the speed passed through.Computational methods are, the summation that at first cuts all secondary formations assurance bandwidth in this formation from this one-level formation total bandwidth the inside (has in fact also retrained interior all the secondary formations of same one-level formation here and has guaranteed that the summation of bandwidth can not be greater than one-level formation total bandwidth, the practical significance of this constraint and assurance bandwidth is contradiction not), if one-level formation total bandwidth also has residue, then remaining bandwidth calculates according to predefined dispatching algorithm between each secondary formation in this formation.For a secondary formation, the theoretical value S1 that calculates adds the assurance bandwidth, just obtains the final speed S of this formation.
The speed limit function of utilizing exchange chip to provide, final speed S in conjunction with each secondary formation that calculates above, speed limit is carried out in these formations (Business Stream that has just mated the specific stream feature of specific firsts and seconds), also just can reach the effect of multi-stage scheduling.
Fig. 2 is a schematic diagram of realizing the multi-stage scheduling pattern on the Ethernet switching chip among the present invention.As shown in Figure 2, suppose to have A and two service providers' of B business process on a switch, two service provider's Business Stream total bandwidths are respectively 30Mbps and 20Mbps, and three kinds of Business Streams 1,2,3 are respectively arranged, and priority from high to low.According to actual needs, it is as follows the assurance bandwidth and the non-assured bandwidth of various traffic streams to be set:
Herein, be that strict priority scheduling (SP) is an example with dispatching algorithm between the formation.
Utilize the counter of exchange chip, each secondary formation is periodically counted with computation rate, at certain time point, the speed that exchange chip records each secondary formation is as follows:
The final speed of each secondary formation of calculation services merchant A at first, according to the computational methods that propose above, at first cut all assurance bandwidth from total bandwidth, also surplus 24Mbps, then according to priority, at first the non-assured bandwidth of Business Stream 1 can satisfy 24Mbps, but flow 1 and itself remove that flow has only 9Mbps outside the assurance, therefore this moment, remaining bandwidth also had 15Mbps, distributed to the non-assured bandwidth 10Mbps of stream 2, and remaining 5Mbps then distributes to stream 3.The rest may be inferred, and the final speed that can calculate all secondary formations is as follows:
According to top result of calculation, can on exchange chip, carry out speed limit to each secondary formation, also just reached this effect of multi-stage scheduling constantly, periodically repeat aforesaid operations, just can reach good multi-stage scheduling effect.
Fig. 3 is a flow chart of realizing the method for multi-stage scheduling according to of the present invention on Ethernet switching chip.As shown in Figure 3, this method may further comprise the steps:
Step S302 is provided with the total bandwidth of each one-level formation and the dispatching algorithm between each secondary formation in the assurance bandwidth of each the secondary formation in the one-level formation, non-assured bandwidth and the same one-level formation;
Step S304 reads one-level formation and the secondary formation actual speed rate a moment;
Step S306 calculates the final speed of each secondary formation in each one-level formation successively;
Step S308 according to the final speed of each secondary formation that calculates, carries out speed limit to actual speed rate greater than the secondary formation of final speed; And
Step S310 waits for the scheduled time, begins to dispatch again from step S304.
In step S304, reading of actual speed rate is to be realized by counter that exchange chip provided.
In step S306, calculating may further comprise the steps: deduct the part that surpasses the assurance bandwidth in all secondary formation actual flows from total bandwidth; According to predefined queue scheduling algorithm, remaining bandwidth is distributed to each secondary formation to satisfy the non-assured bandwidth of each secondary formation, obtain the assigned actual non-assured bandwidth of each secondary formation; And add actual non-assured bandwidth with the bandwidth that guarantees of each secondary formation, to obtain the final speed of secondary formation.
In step S308, speed limit is that the speed limit function by exchange chip realizes.
Wherein, guarantee bandwidth is no matter whether network is congested, exchange chip guarantees that formation is accessible at most and goes out group speed.Non-assured bandwidth is that formation exceeds the maximum that the speed that guarantees bandwidth can reach.The secondary formation is other formation of at least one even lower level with hierarchical relationship.
Thereby, use method of the present invention, on common exchange chip, realize complicated traffic management, when the Business Stream of a plurality of service providers passes through same exchange chip, can be in the single service provider of restriction flow, guarantee that its inner different business carries out independently Differentiated Services according to the priority and the dispatching algorithm of in advance setting, thereby reached good effect.
Be the preferred embodiments of the present invention only below, be not limited to the present invention, for a person skilled in the art, the present invention can have various changes and variation.Within the spirit and principles in the present invention all, any modification of being done, be equal to replacement, improvement etc., all should be included within protection scope of the present invention.
Claims (6)
1. a method that realizes multi-stage scheduling on Ethernet switching chip is characterized in that,
May further comprise the steps:
Step 1 is provided with the total bandwidth of each one-level formation and the dispatching algorithm between each described secondary formation in the assurance bandwidth of each the secondary formation in the described one-level formation, non-assured bandwidth and the same described one-level formation;
Step 2 reads described one-level formation and the described secondary formation actual speed rate a moment;
Step 3 is calculated the final speed of each described secondary formation in each described one-level formation successively;
Step 4 according to the final speed of each the described secondary formation that calculates, is carried out speed limit to described actual speed rate greater than the described secondary formation of described final speed; And
Step 5 is waited for the scheduled time, dispatches again from described step 2 beginning, and wherein, in step 3, described calculating may further comprise the steps:
From described total bandwidth, deduct and do not surpass the part that guarantees bandwidth in all described secondary formation actual flows;
According to predefined described queue scheduling algorithm, remaining bandwidth is distributed to each described secondary formation to satisfy the non-assured bandwidth of each described secondary formation, obtain the assigned actual non-assured bandwidth of each described secondary formation; And
The bandwidth that guarantees with each described secondary formation adds the above actual non-assured bandwidth, to obtain the final speed of described secondary formation.
2. method according to claim 1 is characterized in that, in step 2, reading of described actual speed rate is to be realized by counter that exchange chip provided.
3. method according to claim 1 is characterized in that, in step 4, described speed limit is that the speed limit function by exchange chip realizes.
4. according to each described method in the claim 1 to 3, it is characterized in that the described bandwidth that guarantees is no matter whether network is congested, exchange chip guarantees that formation is accessible at most and goes out group speed.
5. according to each described method in the claim 1 to 3, it is characterized in that described non-assured bandwidth is that formation exceeds the maximum that the speed of described assurance bandwidth can reach.
6. method according to claim 1 is characterized in that, described secondary formation is other formation of at least one even lower level with hierarchical relationship.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2007101230034A CN101102275B (en) | 2007-06-25 | 2007-06-25 | Method for multi-level dispatching on Ethernet switching chip |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2007101230034A CN101102275B (en) | 2007-06-25 | 2007-06-25 | Method for multi-level dispatching on Ethernet switching chip |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101102275A CN101102275A (en) | 2008-01-09 |
| CN101102275B true CN101102275B (en) | 2010-08-04 |
Family
ID=39036369
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2007101230034A Active CN101102275B (en) | 2007-06-25 | 2007-06-25 | Method for multi-level dispatching on Ethernet switching chip |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101102275B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8218442B2 (en) * | 2008-09-11 | 2012-07-10 | Juniper Networks, Inc. | Methods and apparatus for flow-controllable multi-staged queues |
| CN101594307B (en) * | 2009-06-30 | 2012-12-05 | 中兴通讯股份有限公司 | Multi-queue based scheduling method and system |
| CN101692648B (en) * | 2009-08-14 | 2012-05-23 | 中兴通讯股份有限公司 | Queue scheduling method and system |
| CN102158420B (en) * | 2011-05-25 | 2013-10-23 | 杭州华三通信技术有限公司 | Service traffic scheduling method based on priority queue and device thereof |
| CN102594662A (en) * | 2012-01-20 | 2012-07-18 | 中兴通讯股份有限公司 | Realization method and device of quality of service |
| CN103166875B (en) * | 2013-03-12 | 2015-10-28 | 浙江大学 | A kind of double-layer isolation method of the Internet resources based on programmable switch |
| CN103873387A (en) * | 2014-03-11 | 2014-06-18 | 重庆邮电大学 | Industrial Ethernet-oriented deterministic communication scheduling method |
| CN106789739A (en) * | 2017-02-24 | 2017-05-31 | 上海斐讯数据通信技术有限公司 | A kind of multistage multiple row can configure schedule management method and system |
| CN111371690A (en) * | 2020-02-24 | 2020-07-03 | 迈普通信技术股份有限公司 | Flow regulation and control method and device, network equipment and computer readable storage medium |
| CN114245469B (en) * | 2022-02-23 | 2022-06-03 | 南京风启科技有限公司 | Multi-stage scheduling method supporting multiple time periods |
| CN117294599A (en) * | 2022-06-16 | 2023-12-26 | 华为技术有限公司 | Bandwidth adjustment method and related device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1422475A (en) * | 2000-02-08 | 2003-06-04 | 天然网络技术有限公司 | Multi-level scheduling method for multiplexing packets in a communications network |
| CN1625871A (en) * | 2002-04-12 | 2005-06-08 | 西门子公司 | Method for Commonly Checking the Bandwidth of Each Traffic Group |
-
2007
- 2007-06-25 CN CN2007101230034A patent/CN101102275B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1422475A (en) * | 2000-02-08 | 2003-06-04 | 天然网络技术有限公司 | Multi-level scheduling method for multiplexing packets in a communications network |
| CN1625871A (en) * | 2002-04-12 | 2005-06-08 | 西门子公司 | Method for Commonly Checking the Bandwidth of Each Traffic Group |
Non-Patent Citations (1)
| Title |
|---|
| ITU-T Study Group 15.ITU-T G.983.4,A broadband optical access system with increased service capability using dynamic bandwidth assignment.2001,第5-7页第1.3-1.4节,第37-42页. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101102275A (en) | 2008-01-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101102275B (en) | Method for multi-level dispatching on Ethernet switching chip | |
| EP1974509B1 (en) | Multi-priority multi-color markers for traffic metering | |
| CN106789660B (en) | A QoS-aware approach to traffic management in software-defined networking | |
| US6678248B1 (en) | Policy based quality of service | |
| US7697436B2 (en) | Bandwidth allocation | |
| Roberts | Internet traffic, QoS, and pricing | |
| US6859438B2 (en) | Policy based quality of service | |
| US7020143B2 (en) | System for and method of differentiated queuing in a routing system | |
| US6493317B1 (en) | Traffic engineering technique for routing inter-class traffic in a computer network | |
| CN101938403B (en) | Assurance method of multi-user and multi-service quality of service and service access control point | |
| EP2575329B1 (en) | Proportional bandwidth sharing of the excess part in a MEF Traffic Profile | |
| CN101406023B (en) | Method and system for implementing multi-protocol label switching network differential service flux engineering | |
| CN101692648B (en) | Queue scheduling method and system | |
| US20120195200A1 (en) | Method and apparatus for hierarchical policing | |
| CN101478475B (en) | Method for implementing HQoS technique in T-MPLS network | |
| CN101436987A (en) | Method and apparatus for scheduling and configuring QoS service | |
| CN102308535B (en) | hierarchical rate color marker | |
| CN110912736A (en) | Resource allocation method and device | |
| JP2014187421A (en) | Communication device and packet scheduling method | |
| US11336582B1 (en) | Packet scheduling | |
| CN101599905B (en) | Method, device and system for realizing addition of traffic shaping token | |
| CN101127723B (en) | QoS guarantee method for multi-protocol tag exchange L3 virtual private network | |
| US8416689B2 (en) | Meter-based hierarchical bandwidth sharing | |
| CN100476790C (en) | A method and system for resource bundling in communication network | |
| CN102546395A (en) | Service scheduling method and service scheduling device based on Layer 2 Virtual Private Networks (L2VPN) |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |