[go: up one dir, main page]

CN101848167A - Weighted fair queue dispatching method and device based on category - Google Patents

Weighted fair queue dispatching method and device based on category Download PDF

Info

Publication number
CN101848167A
CN101848167A CN201010195729A CN201010195729A CN101848167A CN 101848167 A CN101848167 A CN 101848167A CN 201010195729 A CN201010195729 A CN 201010195729A CN 201010195729 A CN201010195729 A CN 201010195729A CN 101848167 A CN101848167 A CN 101848167A
Authority
CN
China
Prior art keywords
message
grade
interface
classification
token
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.)
Granted
Application number
CN201010195729A
Other languages
Chinese (zh)
Other versions
CN101848167B (en
Inventor
熊途
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Star Net Ruijie Networks Co Ltd
Ruijie Networks Co Ltd
Original Assignee
Beijing Star Net Ruijie Networks Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Star Net Ruijie Networks Co Ltd filed Critical Beijing Star Net Ruijie Networks Co Ltd
Priority to CN2010101957290A priority Critical patent/CN101848167B/en
Publication of CN101848167A publication Critical patent/CN101848167A/en
Application granted granted Critical
Publication of CN101848167B publication Critical patent/CN101848167B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The invention discloses a weighted fair queue dispatching method and a device based on a category. The method comprises the following steps of: distributing a grade token representing the message sending grade for a message according to the message buffering capacity of the message category to which the received message belongs and the capacity representing a category token bucket that the message category allows to use bandwidth; determining the use condition of interface bandwidth in a current dispatching period according to the sending flow rate of the message passing through the interface and an interface token bucket representing the total bandwidth of the interface; determining the allowed sending grade corresponding to the use condition of the interface bandwidth in the current dispatching period according to the corresponding relation of the preset use condition of the interface bandwidth and the allowed sending grade; and buffering the message carrying the grade token of the allowed sending grade to a message sending queue. The method controls the sending of the message through the category token bucket and the interface token bucket, thereby the number of the message buffering queues is reduced, the queue polling of the message dispatching is decreased, the system resource is saved and the system performance is improved.

Description

Class-based Weighted Fair Queuing method and device
Technical field
The present invention relates to technical field of data transmission, refer to a kind of class-based Weighted Fair Queuing method and device especially.
Background technology
Class-based Weighted Fair Queuing (Class-based Weighted Fair Queuing, CBWFQ) be a kind of in service quality (QoS) queue mechanism, CBWFQ has increased the customization formation, be to Weighted Fair Queuing (Weighted Fair Queuing, WFQ) expansion of standard feature, CBWFQ also comprised Low Latency Queueing (Low Latency Queue, LLQ), can guarantee the network bandwidth according to user definition between the disparate networks data flow according to setting regular fair allocat.
In the prior art, class-based Weighted Fair Queuing is the classifying rules according to message, and the message of difference classification is put into different formations, carries out the associative operation of formation bandwidth constraints when message leaves equipment interface, and goes out group message.Be the realization principle schematic of class-based Weighted Fair Queuing in the prior art as shown in Figure 1, the transmission scheduling of packet specifically comprises following process:
Need classify according to the predefined classifying rules of user by the packet that certain equipment interface sends.
Wherein, the packet that conforms with user-defined same matched rule will be taken as same data flow and put into same CBWFQ formation, do not conform with will handling of user-defined matched rule according to the WFQ classifying rules, for example, identical source IP address, purpose IP address, source medium access control (Media AccessControl, MAC) address, target MAC (Media Access Control) address, source port number, the destination slogan, protocol type, COS (Type of Service, TOS) message belongs to same data flow, and each data flow is assigned in the WFQ formation of a reservation.The for example formation shown in Fig. 11, formation 2 ..., formation N etc.
When sending, carry out out group scheduling process, determine to leave the packet of interface and send according to the allocated bandwidth regular allocation bandwidth that CBWFQ sets.
As seen, existing Weighted Fair Queuing, when user definition during the classification of n data, will use n formation to carry out the buffering of data message; Simultaneously, can use the WFQ classifying rules to handle again, be equivalent to and increased a plurality of buffer queues for non-user-defined data qualification.Along with the increase of message queue quantity, when message leaves interface, owing to need each message queue of polling dispatching, frequent queue scheduling handover operation will appear, the decreased performance that causes system, in case and the systematic function deficiency, service quality also can significantly descend or can not get fully to guarantee.
Summary of the invention
The embodiment of the invention provides a kind of class-based Weighted Fair Queuing method and device, and is frequent in order to solve the polling operation that exists in the class-based Weighted Fair Queuing process in the prior art, the problem that causes systematic function to descend.
The present invention includes following content:
A kind of class-based Weighted Fair Queuing method comprises:
According to the message buffering amount of message classification under the message that receives and the capacity of the classification token bucket that characterizes this message classification permission utilized bandwidth, distribute the grade token that characterizes message transmission grade for the message that receives; And
According to the transmitted traffic of the message that passes through interface and the capacity of the interface token bucket that characterizes the interface total bandwidth, determine the interface bandwidth operating position in the current dispatching cycle;
According to the corresponding relation of the transmission grade of predefined interface bandwidth operating position and permission, determine the transmission grade of the permission corresponding with the interface bandwidth operating position in the described current dispatching cycle;
Message with the grade token of the transmission grade correspondence of carrying described permission that receives is buffered in the message transmit queue.
A kind of class-based Weighted Fair Queuing device comprises:
The token distribution module is used for distributing the grade token that characterizes message transmission grade for the message that receives according to the message buffering amount of message classification under the message that receives and the capacity of the classification token bucket that characterizes this message classification permission utilized bandwidth;
The bandwidth monitor module is used for determining the interface bandwidth operating position in the current dispatching cycle according to the transmitted traffic of the message that passes through interface and the capacity of the interface token bucket that characterizes the interface total bandwidth;
The grade determination module is used for the corresponding relation according to the transmission grade of predefined interface bandwidth operating position and permission, determines the transmission grade of the permission corresponding with the interface bandwidth operating position in the described current dispatching cycle;
Send scheduler module, be used for the message of the grade token of the transmission grade correspondence of carrying described permission that will receive, be buffered in the message transmit queue.
A kind of network equipment comprises: above-mentioned class-based Weighted Fair Queuing device.
Beneficial effect of the present invention is as follows:
Class-based Weighted Fair Queuing method and device that the embodiment of the invention provides, according to the message buffering amount of message classification under the message that receives and the capacity of the classification token bucket that characterizes this message classification permission utilized bandwidth, distribute the grade token that characterizes message transmission grade for the message that receives; And, determine the interface bandwidth operating position in the current dispatching cycle according to the transmitted traffic of the message that passes through interface and the capacity of the interface token bucket that characterizes the interface total bandwidth; According to the corresponding relation of the transmission grade of predefined interface bandwidth operating position and permission, determine the transmission grade of the permission corresponding with the interface bandwidth operating position in the described current dispatching cycle; With carrying the message of grade token of the transmission grade of described permission, be buffered in the message transmit queue.This method sends by classification token bucket and interface token bucket control message, realized sending according to the Congestion Level SPCC scheduling and the control message of interface, and make the classification message needn't enter buffer queue separately more separately, the message buffering that only need allow to carry the respective level token enters to send at least one unified message transmit queue and gets final product, this has effectively reduced the message buffering number of queues, thereby reduced the formation poll of dispatching message, avoided frequent scheduling handover operation, save system resource, improved systematic function.
Description of drawings
Fig. 1 is a class-based Weighted Fair Queuing principle schematic in the prior art;
Fig. 2 is the flow chart of class-based Weighted Fair Queuing method in the embodiment of the invention one;
Fig. 3 is the flow chart of class-based Weighted Fair Queuing method in the embodiment of the invention two;
Fig. 4 is the taxonomic structure schematic diagram of classification token in the embodiment of the invention two;
The flow chart that Fig. 5 joins the team for message in the embodiment of the invention two;
Fig. 6 for message in the embodiment of the invention two go out that team sends flow chart;
Fig. 7 is the structural representation of class-based Weighted Fair Queuing device in the embodiment of the invention.
Embodiment
The class-based Weighted Fair Queuing method that the embodiment of the invention provides, at first distribute the different grade tokens that send grade for the message that receives, then, whether the grade token decision of carrying according to the Congestion Level SPCC and the message of interface sends the message that receives, and realizes the dispatching message of class-based Weighted Fair Queuing.The specific implementation process describes by the following examples.
Embodiment one:
The class-based Weighted Fair Queuing method that the embodiment of the invention one provides, its flow process as shown in Figure 2, execution in step is as follows:
Step S101: the message that receives is classified.
After receiving the message of other network equipments transmissions, at first message is classified, determine the classification number of message, and carry out class indication according to user-defined rule.
Wherein, the message that satisfies user-defined classifying rules is included into corresponding classification according to user-defined classifying rules, the message that does not satisfy user-defined classifying rules is included into corresponding classification according to the classifying rules of Weighted Fair Queuing acquiescence, and adds class indication for the message that receives.
Step S102: characterize the grade token that message sends grade for sorted message distributes.
During for sorted message allocation level token, generally the capacity that allows the classification token bucket of utilized bandwidth according to message buffering amount and this message classification of sign according to message classification under the message that receives characterizes the grade token that message sends grade for the message that receives distributes.
Wherein, the pairing classification token bucket of all kinds of messages has a pre-assigned capability value, and the capacity summation of the pairing classification token bucket of each message classification is smaller or equal to the capacity of the interface token bucket that characterizes interface bandwidth.Set at least one according to the capacity of each classification token bucket respectively for each message classification and send grade threshold, with the token distribution condition of the classification token bucket that characterizes such message.
For the message allocation level token that receives the time, specifically determine the message buffering situation of this message classification earlier according to the classification of the message that receives, generally can monitor the message buffering amount of the message classification under the message that receives.Then, the magnitude relationship of the transmission grade threshold of the message buffering amount of this message classification and this message classification is relatively determined the transmission grade of each message of receiving respectively, and is distributed the grade token of corresponding transmission grade for each message.
Preferable, after being message allocation level token, upgrade token quantity in this message classification corresponding token bucket according to the message buffering amount of this message classification.
Therefore also can be described as residue token, be the highest grade of transmission the grade token in the classification token bucket of the message classification under the message distribution that receives according to the message classification under the message that receives.
Step S103:, determine the interface bandwidth operating position in the current dispatching cycle according to the transmitted traffic of the message that passes through interface and the capacity of the interface token bucket that characterizes the interface total bandwidth.Specifically comprise:
Pass through the transmitted traffic of the message of interface in the monitoring current dispatching cycle.
Judge the magnitude relationship of transmitted traffic and predefined at least one flow grade thresholding, determine to characterize the flow grade of interface bandwidth operating position; Wherein, the flow grade thresholding is set according to the interface token bucket capacity.For example:, think that the interface bandwidth operating position is in this thresholding corresponding flow grade when transmitted traffic during smaller or equal to the flow grade thresholding.
Step S104: the transmission grade of determining the permission corresponding with the interface bandwidth operating position in the current dispatching cycle.
Preestablish the corresponding relation of the transmission grade of interface bandwidth operating position and permission, specifically can set the transmission grade of the pairing permission of each flow grade.After step S103 determines the interface bandwidth operating position, just can determine the transmission grade of the permission corresponding according to the corresponding relation of the transmission grade of predefined interface bandwidth operating position and permission with the interface bandwidth operating position in the current dispatching cycle.
Specifically, determine the transmission grade of the pairing permission of present flow rate grade of the interface bandwidth operating position in the sign current dispatching cycle according to each flow grade of predefined interface and the corresponding relation of the transmission grade that is allowed.
Step S105: whether the grade token that the message that judgement receives carries is the grade token of the interior transmission grade that is allowed of current dispatching cycle.
That is to say,, judge whether each message that receives can be sent out according to the grade token that the transmission grade and the message of the permission of determining carries.Specifically be whether the grade token that carries according to the message that receives is that the grade token of the transmission grade that allowed in the current dispatching cycle is determined.
If think that then the message that receives allows to be sent out execution in step S106; Otherwise, think that the message that receives does not allow to be sent out execution in step S107.
Step S106: with message buffering to the message transmit queue.
Be about to carry the message of the grade token of the transmission grade that interface allows, be buffered in the message transmit queue.Wherein, the message transmit queue can have only one, also can be two or more according to priority level initializing.The message transmit queue of this moment and classification are irrelevant, that is to say in the different classes of transmit queue according to classification buffering message again.
Step S107: abandon the message that receives.
Not meeting the message that sends grade can not be sent out.
Embodiment two
The class-based Weighted Fair Queuing method that the embodiment of the invention two provides is that example describes to comprise two message transmit queues, its realization flow as shown in Figure 3, execution in step is as follows:
Step S201: the message that receives is classified.
Concrete dividing mode is with step S101.
For example: the message that the network equipment receives may comprise N classification.
Step S202: characterize the grade token that message sends grade for the sorted message that receives distributes.
Concrete assigning process is with step S102.
The distribution structure of classification token can be as shown in Figure 4.Assignable total token number is determined according to interface bandwidth, that is to say according to the capacity of interface token bucket and determines.Then token is distributed to each message classification, obtained the classification token bucket of each classification.Comprising message classification of determining according to user-defined classifying rules and the message classification determined according to the classifying rules of Weighted Fair Queuing acquiescence.At first, be that user-defined classification (being categories of customization) distributes token, for example, classification token 1 ..., classification token N etc., remaining token redistribution is given the message classification of acquiescence.For example: the residue token carries out the token granting according to message TOS priority, can be divided into 0-7 totally eight grades of classifications, and the classification number according to current existence distributes then.
For example: interface bandwidth is 500M, and the capacity (i.e. this message classification allow utilized bandwidth) of distributing to the token bucket 1 of classification A is 50M, distribute to classification B token bucket 2 capacity be 40M ..., the capacity of distributing to the token bucket N of classification X is 10M, or the like.The capacity summation of the token bucket of all categories is not more than interface bandwidth, i.e. the capacity of interface token bucket.The corresponding available tokens quantity of preserving own current token bucket of each classification.
Token bucket 1 with classification A is an example, the transmission grade threshold that is provided with is respectively 50M and 60M, by these two grade thresholds message is sent grade and be divided into three grades, wherein, the message buffering amount is during smaller or equal to 50M, the bandwidth of expression this message classification of distributing to does not use as yet, and corresponding grade token is a green tokens; The message buffering amount is during greater than 50M and less than 60M, the bandwidth of expression this message classification of distributing to has used, but when interface bandwidth allows, can allow to send this part message, corresponding grade token is a yellow tokens, the message buffering amount is during greater than 60M, the bandwidth of expression this message classification of distributing to has used, and its buffering capacity has surpassed the grade threshold scope of the permission utilized bandwidth setting of distributing to this message classification, then have only when interface bandwidth is quite idle, just can allow to send this part message, corresponding grade token is red token.Above-mentioned transmission grade threshold can be set according to the buffer capacity of system.
Then when the message that receives is classification A, distribute green, yellow or red grade token for it according to the message buffering amount of this message classification.We can say that also according to such other message buffering amount and transmission grade thresholding be that message is painted, is green, yellow, redness etc. with packet labeling.When message buffering amount or the token of saying token bucket distribute when reaching corresponding transmission grade thresholding, the color of corresponding transmission grade for message.
As shown in Figure 3, token bucket 2,3 ... the operating position of N and token bucket 1 are similar, give unnecessary details no longer one by one.The message that receives so all has been added the grade token of different colours.
Step S203:, determine the interface bandwidth operating position in the current dispatching cycle according to the transmitted traffic of the message that passes through interface and the capacity of the interface token bucket that characterizes the interface total bandwidth.Specifically comprise:
For example: still be that 500M is an example with the interface bandwidth, setting the flow grade thresholding is 400M and 500M, when message transmitted traffic during smaller or equal to 400M, confirm that flow grade is green grade, be that the interface operating position is for idle, during less than 500M, confirm that flow grade is yellow grade greater than 400M when the message transmitted traffic, promptly the interface operating position is normal; When flow during more than or equal to 500M, confirm that flow grade is red grade, promptly the interface operating position is for blocking up.The flow grade thresholding is used to characterize interface and whether is in congestion state, and it can be adjusted according to the physical circuit transmitting capacity of interface, can carry out initialization according to interface initialization ability.
After having set the flow grade threshold value, then by passing through the transmitted traffic of the message of interface in the monitoring current dispatching cycle, confirm the operating position of interface bandwidth, with the flow grade threshold ratio of setting, confirm that interface bandwidth is in green, the yellow still flow grade of redness, thereby can know the Congestion Level SPCC of interface.
Step S204: the transmission grade of determining the permission corresponding with the interface bandwidth operating position in the current dispatching cycle.
Preestablish the corresponding relation of the transmission grade of interface bandwidth operating position and permission, specifically can set the transmission grade of the pairing permission of each flow grade.For example: when the flow grade that characterizes the interface bandwidth operating position is green, can sends the grade token that carries and be green, yellow and all red messages; When flow grade is yellow, can sends the grade token that carries and be green, yellow message; When flow grade is red, can only sends the grade token that carries and be green message.
Then after step S203 has determined the pairing flow grade of interface bandwidth operating position, just can determine to allow to send the transmission grade of message, for example: allow to send the message that carries green grade token, or allow to send the message that carries green and yellow grade token, or the message that carries any grade token can both send.
Step S205: whether the grade token that the message that judgement receives carries is the grade token of the interior transmission grade that is allowed of current dispatching cycle.
That is to say that the token of the transmission grade whether the grade token that carries according to the message that receives is in the current dispatching cycle to be allowed determines whether message can be sent out.
If think that then the message that receives allows to be sent out execution in step S207; Otherwise, think that the message that receives does not allow to be sent out execution in step S206.
For example: determine the present flow rate grade for yellow according to bandwidth usage, then if message carries is green and yellow grade token, then this message allows to be sent out, if message carries is red grade token, then this message can not be sent out.
Step S206: abandon the message that receives.
The token that carries is not to allow the message of the grade token of transmission grade to be dropped.
Step S207: according to the priority of the message classification under the message of the permission transmission of determining, the message buffering that will allow to send is to the corresponding priorities formation.
The priority query that sends with message comprises that low delay message queue and quality services QoS common queue are example.Wherein, the low buffer queue that postpones cushions satisfying the low message within transmitting capacity that postpones rule, and common message buffering formation cushions the message that is within the transmitting capacity beyond the non-low delay, then the message flow process of joining the team comprises as shown in Figure 5:
Step S207a: judge whether message belongs to the message classification of Low Latency Queueing.
If, execution in step S207b; Otherwise, execution in step S207c.
Step S207b: the message buffering that will allow to send is to Low Latency Queueing.
Step S207c: the message buffering that will allow to send is to quality services QoS common queue.
For example:, then be low delayed data message flow if current data stream coupling is voice or video flowing; If be web-page requests or ftp flow then for common message flow.The user branch time-like of message is set, needs to formulate interior transmitting capacity of unit interval for it simultaneously, if do not carry out the distribution of transmitting capacity, then the idle bandwidth of docking port is evenly seized.
Preferable, the transmitting capacity of each classification embodies by message amount or the bit stream size that allows to send, and disposes definitely by the user, carries out strict guarantee for the bandwidth of custom queuing, then evenly seizes for unappropriated bandwidth.For example: when the user can define and satisfy current stream and be video flowing, its bandwidth that takies was 20% (being scaled bit number according to the interface total bandwidth) of interface bandwidth or is 100,000 packets/second (packets/s).When bit number and the message amount conflict that allow to send, get that it is little.
When the priority query that sets is many, then need determine the priority query under the message classification, and with message buffering to the corresponding priority level formation.
Step S208: the priority according to buffer queue goes out team's transmission message.
Comprise with the message transmit queue still low postponing message queue and quality services QoS common queue is an example that then message goes out flow process that team sends as shown in Figure 6, comprising:
Step S208a: judge whether message to be sent is arranged in the Low Latency Queueing.
If, execution in step S208c; Otherwise, execution in step S208b.
Step S208b: judge whether message to be sent is arranged in the common queue.
If, execution in step S208c; Otherwise, return and continue execution in step S208a.
Step S208c: send message to be sent.
Return continuation execution in step S208a after sending the message of being with transmission, check whether message to be sent is arranged in the Low Latency Queueing.
When the priority query that sets is many, then need to dispatch each priority query successively according to preferential and order.
That is to say that different with existing implementation, the application can only realize the buffering of message by two formations.Join the team the stage at message,, guarantee that message can go out team directly to the message that surpasses transmitting capacity is carried out discard processing, simultaneously, guarantee low postpone buffer queue preferentially go out team.
The above-mentioned class-based Weighted Fair Queuing method that provides according to the embodiment of the invention, a kind of class-based Weighted Fair Queuing device can be provided, this device can be arranged at the network equipment that is used for message transmissions, for example: router, switch etc.The structure of this device comprises as shown in Figure 7: token distribution module 10, bandwidth monitor module 20, grade determination module 30 and transmission scheduler module 40.
Token distribution module 10 is used for distributing the grade token that characterizes message transmission grade for the message that receives according to the message buffering amount of message classification under the message that receives and the capacity of the classification token bucket that characterizes this message classification permission utilized bandwidth.
Preferable, above-mentioned token distribution module 10 specifically comprises: threshold value is provided with unit 101, first monitoring unit 102 and compares allocation units 103.
Threshold value is provided with unit 101, is used for setting at least one according to the capacity of classification token bucket respectively for each message classification and sends grade threshold.
First monitoring unit 102 is used to monitor the message buffering amount of the message classification under the message that receives.
Compare allocation units 103, be used for the magnitude relationship that comparison message buffering amount that first monitoring unit 102 obtains and threshold value are provided with the transmission grade threshold that unit 101 is provided with, determine the transmission grade of each message of receiving respectively, and distribute the grade token of corresponding transmission grade for each message.
Bandwidth monitor module 20 is used for determining the interface bandwidth operating position in the current dispatching cycle according to the transmitted traffic of the message that passes through interface and the capacity of the interface token bucket that characterizes the interface total bandwidth.
Preferable, above-mentioned bandwidth monitor module 20 specifically comprises: second monitoring unit 201 and grade judging unit 202.
Second monitoring unit 201 is used to monitor the transmitted traffic of passing through the message of interface in the current dispatching cycle.
Grade judging unit 202 is used to judge the transmitted traffic that second monitoring unit 201 is determined and the magnitude relationship of predefined at least one flow grade thresholding, determines to characterize the flow grade of interface bandwidth operating position; Wherein, the flow grade thresholding is set according to the interface token bucket capacity.
Grade determination module 30 is used for the corresponding relation according to the transmission grade of predefined interface bandwidth operating position and permission, determines the transmission grade of the permission corresponding with the interface bandwidth operating position in the current dispatching cycle.
Send scheduler module 40, be used for being buffered in the message transmit queue carrying the message of grade token of the transmission grade of permission.
Preferable, above-mentioned transmission scheduler module 40 specifically comprises: judging unit 401 and scheduling unit 402.
Judging unit 401 is used for the grade token that carries according to the message that receives, judges that whether this message is the message of the transmission grade that allowed in the current dispatching cycle.
Scheduling unit 402 is used for being judged as when being when judging unit 401, with message buffering to the message transmit queue; Otherwise, abandon this message.
Preferable, above-mentioned scheduling unit 402 specifically is used for: according to the priority of the message classification under the message of the permission transmission of determining, the message buffering that will allow to send is to the corresponding priorities formation.
Above-mentioned class-based Weighted Fair Queuing method and device that the embodiment of the invention provides, according to the message buffering amount of each message classification and the permission utilized bandwidth of this message classification, determine the transmission grade of the message that receives, and determine that according to the actual flow of interface the permission of interface sends grade, the transmission control of the message received achieves a butt joint, thereby needn't be at each category setting buffer queue, only need to allow the message buffering to that sends, in two or more buffer queues, thereby realized the merging of buffer queue, make the classification message needn't enter buffer queue separately more separately, the message buffering that only need allow to carry the respective level token enters to send at least one unified message transmit queue and gets final product; When adopting a plurality of priority queries buffering, the message buffering that only need will belong to each classification of same priority gets final product to this priority query, when message goes out team's transmission, only needs each priority query of poll to get final product; Thereby effectively reduced the handover operation of formation poll, the resource waste of having avoided frequent scheduling handover operation to be brought has been saved system resource, has improved the performance of system.
And controlling message by interface token and classification token two-stage token among the application sends, the classification token is related with classification number, realized sending according to the Congestion Level SPCC scheduling and the control message of interface, effectively coordinated the permission traffic volume of interface transmitting capacity and message of all categories, improved bandwidth utilization to greatest extent, only avoided in the prior art problem of distributing allocated bandwidth inequality that bandwidth brings more and can not adjust in real time according to dynamic need for the buffer queue of each classification according to classification and priority.And effectively solved the big time-delay of message in the low latency queue in the prior art, also level and smooth common queue has guaranteed the first in first out of same priority message owing to dispatch the big time-delay situation of unbalanced appearance.
Because the token in the classification token bucket can dynamically update according to the buffering situation of current classification, thereby the control effect when only adopting single or several priority queries can realize a plurality of classification buffer queues of available technology adopting, polling dispatching, and the remaining permission utilized bandwidth of each classification can dynamically update distribution according to priority and active degree in the transmission grade threshold scope of setting, and guarantees that the big data flow of data volume can access transmission timely.
For example: if the Low Latency Queueing that exists 5 users to set, according to former method, then when each scheduling, all need these 5 formations of poll, at this moment, if first formation has message to send within transmitting capacity always, then the message in the Low Latency Queueing thereafter will increase time of delay.This influence is more obvious to the formation influence that is polled at last, and common queue and low latency queue are similar.Adopt the single queue buffering can guarantee that then the message of going into formation earlier sends earlier, can satisfy the low delay requirement of message to a greater extent.
Preferable, can be set to low latency queue and common queue to the different requirement priority query of time delay according to the data flow of video and non-video, determine to enter the message of these two formations according to aforesaid way after, when going out team's transmission, only need these two formations of poll to get final product.
Obviously, those skilled in the art can carry out various changes and modification to the present invention and not break away from the spirit and scope of the present invention.Like this, if of the present invention these are revised and modification belongs within the scope of claim of the present invention and equivalent technologies thereof, then the present invention also is intended to comprise these changes and modification interior.

Claims (12)

1. a class-based Weighted Fair Queuing method is characterized in that, comprising:
According to the message buffering amount of message classification under the message that receives and the capacity of the classification token bucket that characterizes this message classification permission utilized bandwidth, distribute the grade token that characterizes message transmission grade for the message that receives; And
According to the transmitted traffic of the message that passes through interface and the capacity of the interface token bucket that characterizes the interface total bandwidth, determine the interface bandwidth operating position in the current dispatching cycle;
According to the corresponding relation of the transmission grade of predefined interface bandwidth operating position and permission, determine the transmission grade of the permission corresponding with the interface bandwidth operating position in the described current dispatching cycle;
Message with the grade token of the transmission grade correspondence of carrying described permission that receives is buffered in the message transmit queue.
2. the method for claim 1, it is characterized in that, the message buffering amount of message classification and this message classification of sign allow the capacity of the classification token bucket of utilized bandwidth under the message that described basis receives, characterize the grade token that message sends grade for the message that receives distributes, specifically comprise:
Set at least one according to the capacity of described classification token bucket respectively for each message classification and send grade threshold;
The message buffering amount of the message classification under the message that monitoring receives, the magnitude relationship of more described message buffering amount and described transmission grade threshold, determine the transmission grade of each message of receiving respectively, and distribute the grade token of corresponding transmission grade for each message.
3. the method for claim 1 is characterized in that, described basis is determined the interface bandwidth operating position in the current dispatching cycle by the transmitted traffic of the message of interface and the capacity of the interface token bucket that characterizes the interface total bandwidth, specifically comprises:
Pass through the transmitted traffic of the message of interface in the monitoring current dispatching cycle;
Judge the magnitude relationship of described transmitted traffic and predefined at least one flow grade thresholding, determine to characterize the flow grade of interface bandwidth operating position; Described flow grade thresholding is set according to described interface token bucket capacity.
4. method as claimed in claim 3, it is characterized in that, the corresponding relation of described transmission grade according to predefined interface bandwidth operating position and permission is determined the transmission grade of the permission corresponding with the interface bandwidth operating position in the described current dispatching cycle, specifically comprises:
According to each flow grade of predefined interface and the corresponding relation of the transmission grade that is allowed, determine the transmission grade of the pairing permission of present flow rate grade of the interface bandwidth operating position in the sign current dispatching cycle.
5. the method for claim 1 is characterized in that, the capacity summation of the pairing classification token bucket of each message classification is smaller or equal to the capacity of described interface token bucket.
6. as the arbitrary described method of claim 1-5, it is characterized in that, described with message buffering to the message transmit queue, specifically comprise:
The priority of the message classification under the message that sends according to the permission determined, the message buffering that described permission is sent is to the corresponding priorities formation.
7. method as claimed in claim 6 is characterized in that, described message transmit queue comprises: low message queue and the quality services QoS common queue of postponing;
The priority of the message classification under the message that permission that described basis is determined sends, the message buffering that described permission is sent specifically comprises to the corresponding priorities formation:
Whether the affiliated message classification of message of judging the permission transmission of determining belongs to the low message that postpones;
If then the message buffering that described permission is sent is to Low Latency Queueing; Otherwise the message buffering that described permission is sent is to quality services QoS common queue.
8. a class-based Weighted Fair Queuing device is characterized in that, comprising:
The token distribution module is used for distributing the grade token that characterizes message transmission grade for the message that receives according to the message buffering amount of message classification under the message that receives and the capacity of the classification token bucket that characterizes this message classification permission utilized bandwidth;
The bandwidth monitor module is used for determining the interface bandwidth operating position in the current dispatching cycle according to the transmitted traffic of the message that passes through interface and the capacity of the interface token bucket that characterizes the interface total bandwidth;
The grade determination module is used for the corresponding relation according to the transmission grade of predefined interface bandwidth operating position and permission, determines the transmission grade of the permission corresponding with the interface bandwidth operating position in the described current dispatching cycle;
Send scheduler module, be used for the message of the grade token of the transmission grade correspondence of carrying described permission that will receive, be buffered in the message transmit queue.
9. device as claimed in claim 8 is characterized in that, described token distribution module specifically comprises:
Threshold value is provided with the unit, is used for setting at least one according to the capacity of described classification token bucket respectively for each message classification and sends grade threshold;
First monitoring unit is used to monitor the message buffering amount of the message classification under the message that receives;
Allocation units relatively are used for the magnitude relationship of more described message buffering amount and described transmission grade threshold, determine the transmission grade of each message of receiving respectively, and distribute the grade token of corresponding transmission grade for each message.
10. device as claimed in claim 8 is characterized in that, described bandwidth monitor module specifically comprises:
Second monitoring unit is used to monitor the transmitted traffic of passing through the message of interface in the current dispatching cycle;
The grade judging unit is used to judge the magnitude relationship of described transmitted traffic and predefined at least one flow grade thresholding, determines to characterize the flow grade of interface bandwidth operating position; Described flow grade thresholding is set according to described interface token bucket capacity.
11., it is characterized in that described scheduling unit specifically is used for as the arbitrary described device of claim 8-10:
The priority of the message classification under the message that sends according to the permission determined, the message buffering that described permission is sent is to the corresponding priorities formation.
12. a network equipment is characterized in that, comprising: as the arbitrary described class-based Weighted Fair Queuing device of claim 8-11.
CN2010101957290A 2010-06-01 2010-06-01 Weighted fair queue dispatching method and device based on category Expired - Fee Related CN101848167B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2010101957290A CN101848167B (en) 2010-06-01 2010-06-01 Weighted fair queue dispatching method and device based on category

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2010101957290A CN101848167B (en) 2010-06-01 2010-06-01 Weighted fair queue dispatching method and device based on category

Publications (2)

Publication Number Publication Date
CN101848167A true CN101848167A (en) 2010-09-29
CN101848167B CN101848167B (en) 2012-01-25

Family

ID=42772630

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2010101957290A Expired - Fee Related CN101848167B (en) 2010-06-01 2010-06-01 Weighted fair queue dispatching method and device based on category

Country Status (1)

Country Link
CN (1) CN101848167B (en)

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102082693A (en) * 2011-02-15 2011-06-01 中兴通讯股份有限公司 Method and device for monitoring network traffic
CN102437956A (en) * 2011-12-16 2012-05-02 福建星网锐捷网络有限公司 Method, device and equipment for balancing traffic of uplink aggregation port
WO2013000116A1 (en) * 2011-06-28 2013-01-03 中兴通讯股份有限公司 Method and device for leaky bucket speed-limitation
CN103324592A (en) * 2013-06-24 2013-09-25 华为技术有限公司 Data migration control method, data migration method and data migration device
CN103647729A (en) * 2013-11-19 2014-03-19 华为技术有限公司 Delaying request processing method and apparatus based on token bucket
CN104219169A (en) * 2014-09-30 2014-12-17 杭州华三通信技术有限公司 CBQ (Class Based Queuing) scheduling method and device
CN104378602A (en) * 2014-11-26 2015-02-25 福建星网锐捷网络有限公司 Video transmission method and device
CN109246023A (en) * 2018-11-16 2019-01-18 锐捷网络股份有限公司 Flow control methods, the network equipment and storage medium
CN110149282A (en) * 2019-01-30 2019-08-20 华为技术有限公司 Traffic scheduling method and device
CN112838991A (en) * 2021-02-09 2021-05-25 芯河半导体科技(无锡)有限公司 Method for adaptively adjusting token value of small-bandwidth scheduler
CN113067778A (en) * 2021-06-04 2021-07-02 新华三半导体技术有限公司 Flow management method and flow management chip
CN113206799A (en) * 2021-03-30 2021-08-03 新华三信息安全技术有限公司 Message forwarding method and device
CN113489655A (en) * 2021-06-11 2021-10-08 的卢技术有限公司 Intelligent bandwidth allocation method and system based on token bucket
CN113904994A (en) * 2021-09-30 2022-01-07 天翼数字生活科技有限公司 Method for uniformly reporting big data of home gateway to platform
CN116016371A (en) * 2017-12-31 2023-04-25 华为技术有限公司 Method, equipment and system for sending message

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6862265B1 (en) * 2000-04-13 2005-03-01 Advanced Micro Devices, Inc. Weighted fair queuing approximation in a network switch using weighted round robin and token bucket filter
CN101272349A (en) * 2008-05-06 2008-09-24 北京星网锐捷网络技术有限公司 Priority level analog queue control method and device of quality service
CN101478494A (en) * 2009-02-16 2009-07-08 中兴通讯股份有限公司 Data packet processing method and apparatus based on token barrel algorithm

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6862265B1 (en) * 2000-04-13 2005-03-01 Advanced Micro Devices, Inc. Weighted fair queuing approximation in a network switch using weighted round robin and token bucket filter
CN101272349A (en) * 2008-05-06 2008-09-24 北京星网锐捷网络技术有限公司 Priority level analog queue control method and device of quality service
CN101478494A (en) * 2009-02-16 2009-07-08 中兴通讯股份有限公司 Data packet processing method and apparatus based on token barrel algorithm

Cited By (28)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102082693A (en) * 2011-02-15 2011-06-01 中兴通讯股份有限公司 Method and device for monitoring network traffic
CN102082693B (en) * 2011-02-15 2015-05-20 中兴通讯股份有限公司 Method and device for monitoring network traffic
WO2012109911A1 (en) * 2011-02-15 2012-08-23 中兴通讯股份有限公司 Method and apparatus for monitoring network traffic
WO2013000116A1 (en) * 2011-06-28 2013-01-03 中兴通讯股份有限公司 Method and device for leaky bucket speed-limitation
US9160669B2 (en) 2011-06-28 2015-10-13 Zte Corporation Method and device for leaky bucket speed-limitation
CN102437956B (en) * 2011-12-16 2014-07-30 福建星网锐捷网络有限公司 Method, device and equipment for equalizing flow of uplink aggregation port
CN102437956A (en) * 2011-12-16 2012-05-02 福建星网锐捷网络有限公司 Method, device and equipment for balancing traffic of uplink aggregation port
CN103324592A (en) * 2013-06-24 2013-09-25 华为技术有限公司 Data migration control method, data migration method and data migration device
CN103324592B (en) * 2013-06-24 2016-11-23 华为技术有限公司 A kind of Data Migration control method, data migration method and device
CN103647729A (en) * 2013-11-19 2014-03-19 华为技术有限公司 Delaying request processing method and apparatus based on token bucket
US9503385B2 (en) 2013-11-19 2016-11-22 Huawei Technologies Co., Ltd. Delay request processing method and apparatus based on token bucket
CN103647729B (en) * 2013-11-19 2017-01-04 华为技术有限公司 A kind of delay request processing method based on token bucket and device
CN104219169A (en) * 2014-09-30 2014-12-17 杭州华三通信技术有限公司 CBQ (Class Based Queuing) scheduling method and device
CN104219169B (en) * 2014-09-30 2017-11-10 新华三技术有限公司 Class-based queue CBQ dispatching methods and equipment
CN104378602B (en) * 2014-11-26 2017-06-23 福建星网锐捷网络有限公司 Video transmission method and device
CN104378602A (en) * 2014-11-26 2015-02-25 福建星网锐捷网络有限公司 Video transmission method and device
CN116016371A (en) * 2017-12-31 2023-04-25 华为技术有限公司 Method, equipment and system for sending message
CN109246023A (en) * 2018-11-16 2019-01-18 锐捷网络股份有限公司 Flow control methods, the network equipment and storage medium
CN110149282A (en) * 2019-01-30 2019-08-20 华为技术有限公司 Traffic scheduling method and device
CN110149282B (en) * 2019-01-30 2021-02-12 华为技术有限公司 Traffic scheduling method and device
CN112838991A (en) * 2021-02-09 2021-05-25 芯河半导体科技(无锡)有限公司 Method for adaptively adjusting token value of small-bandwidth scheduler
CN113206799A (en) * 2021-03-30 2021-08-03 新华三信息安全技术有限公司 Message forwarding method and device
CN113206799B (en) * 2021-03-30 2022-05-27 新华三信息安全技术有限公司 Message forwarding method and device
CN113067778A (en) * 2021-06-04 2021-07-02 新华三半导体技术有限公司 Flow management method and flow management chip
CN113489655A (en) * 2021-06-11 2021-10-08 的卢技术有限公司 Intelligent bandwidth allocation method and system based on token bucket
CN113489655B (en) * 2021-06-11 2024-09-10 西藏宁算科技集团有限公司 Intelligent bandwidth allocation method and allocation system based on token bucket
CN113904994A (en) * 2021-09-30 2022-01-07 天翼数字生活科技有限公司 Method for uniformly reporting big data of home gateway to platform
CN113904994B (en) * 2021-09-30 2024-03-19 天翼数字生活科技有限公司 Method for uniformly reporting big data of home gateway to platform

Also Published As

Publication number Publication date
CN101848167B (en) 2012-01-25

Similar Documents

Publication Publication Date Title
CN101848167B (en) Weighted fair queue dispatching method and device based on category
US7180855B1 (en) Service interface for QoS-driven HPNA networks
US8532130B2 (en) Service interface for QoS-driven HPNA networks
CN101115013B (en) Provides methods for resource admission control
CN101557340B (en) Method for realizing multilevel queue scheduling in data network and device
US8467295B2 (en) System and methods for distributed quality of service enforcement
RU2005131960A (en) ACCESS PERMISSION MANAGEMENT AND RESOURCE ALLOCATION IN THE COMMUNICATION SYSTEM WITH SUPPORT OF APPLICATION STREAMS WITH AVAILABILITY OF SERVICE QUALITY REQUIREMENTS
CN101436987A (en) Method and apparatus for scheduling and configuring QoS service
RU2010115366A (en) PRIORITY PLANNING AND CONTROL OF ACCESS IN THE COMMUNICATION NETWORK
CN108616458A (en) The system and method for schedule packet transmissions on client device
CN101692648B (en) Queue scheduling method and system
CN113747597B (en) Network data packet scheduling method and system based on mobile 5G network
JP2006506845A (en) How to select a logical link for a packet in a router
US7843825B2 (en) Method and system for packet rate shaping
CN102594669A (en) Data message processing method, device and equipment
CN101212412A (en) MPLS network service scheduling system
CN101616096B (en) Method and device for scheduling queue
CN101061681B (en) Communication time fair transmission management without explicit traffic specifications for wireless networks
Chen et al. Packet scheduling algorithm based on priority adjustment in wireless sensor networks
CN102769566A (en) Method and device for configuring multilevel scheduling system, and method and device for changing configuration of multilevel scheduling system
Yaghmaee et al. A model for differentiated service support in wireless multimedia sensor networks
KR101013764B1 (en) Network congestion control device and method
Yin et al. Critical event-triggered flows tolerance in time-sensitive networks
Omotoye et al. Congestion Management on GSM Networks using Traffic Class Prioritization
CN102201984A (en) Queue scheduling method and device

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
C17 Cessation of patent right
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20120125

Termination date: 20140601