CN100426738C - Realizing method of data stream constraint control - Google Patents
Realizing method of data stream constraint control Download PDFInfo
- Publication number
- CN100426738C CN100426738C CNB2005100336935A CN200510033693A CN100426738C CN 100426738 C CN100426738 C CN 100426738C CN B2005100336935 A CNB2005100336935 A CN B2005100336935A CN 200510033693 A CN200510033693 A CN 200510033693A CN 100426738 C CN100426738 C CN 100426738C
- Authority
- CN
- China
- Prior art keywords
- token
- token bucket
- flow
- bucket
- depth
- 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
Images
Abstract
The present invention relates to a realizing method for the constrained control of a data flow. A token bucket is constructed, and comprises a depth parameter and a bandwidth parameter of the token bucket; the depth of the token bucket is the sum of the depth value of a fixed token bucket and the depth value of a dynamic token bucket; according to a flow distributed to a user, a corresponding token number is added to the token bucket, and meanwhile tokens are subtracted from the token bucket according to the actual flow of the user; before the tokens are subtracted, the flow is illegal when the token number in the token bucket is less than the depth value of the fixed token bucket; when the tokens are added next time, firstly, the depth value of the fixed token bucket is fully replenished, and the user flow can passes through. The present invention can also be realized by adopting a flow plus token way, and the depth value of the dynamic token bucket varies with maximum bandwidth; the depth value of the fixed token bucket can be a token number corresponding to the length of a longest data packet. The present invention can increase the passing rate of the longest data packet, can effectively control an instantaneous burst flow, and reduce the memory space occupied by the token bucket.
Description
Technical field
The present invention relates to communication technical field, relate in particular to a kind of implementation method of data traffic constraint control.
Background technology
CAR (Committed Access Rate) is the commit access rate technology, is used to retrain customer flow, and it is a kind of effective Bandwidth Management mode of IP network.CAR is mainly used in the communications field, carries out flow restriction at the client.
Relying on the user conscious is unpractical by service regulation transmission data flow, user's flow surpasses the service level agreement (SLA) with service provider's signing probably, will tie up other user's bandwidth, cause other legal flow not served normally.In the case, for guaranteeing the normal operation of network, need at the Web portal place customer flow to be limited, with the assurance customer flow not in violation of rules and regulations, and the user can not get surpassing the extra-service of agreement regulation.
CAR can also can carry out flow restriction at a group user at unique user.This a group user must be the user that can be divided into a classification.For example limiting certain Internet user's outlet bandwidth is 512kbps, and the general export bandwidth that also can limit certain all user of sub-district is 512kbps.Usually, at the edge of network interface, can limit bursty traffic by configuration CAR, control IP flow passes in and out network with particular rate, so that the Internet business, provides the network service quality (QoS) of guarantee.
The CAR permission strategy that the definition flow shrinks on route network can be defined in the classification and the traffic policy that enter on the interface, can also take measures when surpassing the limit of fixed-bandwidth distribution when the communication of being managed.CAR can also be provided with priority in the extended access tabulation, so it can provide regulation management to using content, port numbers, source destination address or the like.CAR itself just possesses the ability of supervisory communications amount, so also can prevent and control congested.
At present CAR mainly is that mode with token bucket realizes: controller adds token according to the flow of distributing to the user in token bucket, simultaneously, subtracts token according to the flow of user's reality in token bucket.Perhaps opposite, controller subtracts token according to the flow of distributing to the user in token bucket, simultaneously, adds token according to the flow of user's reality in token bucket.
Add token according to the flow of distributing to the user with controller toward token bucket in, simultaneously, subtracting token according to the flow of user's reality in the token bucket is that example describes, and the token bucket data structure of prior art employing is as follows:
| Parameter name | The token bucket degree of depth | Token number is added in every circulation |
| Scope | N*MAXToken, n is generally 2 | 0～MAXToken |
Illustrate: MAXToken: unit interval maximum stream flow, i.e. bandwidth, unit: token number (common 1 token=1 byte)
For one thing: construct a token bucket, this token bucket can hold A token at most, and the flow B that the permission of switch is passed through in A and a period of time T mates.For 1G token flow, a token bucket takies approximately (60+n-1) bit memory space at most, because 1G=10
9≈ 2
30, need in other words could represent fully with 30 bits.Every circulation is added token number and is equivalent to the 1G flow at most, needs the 30bit binary representation, and bucket is dark to be 2 times of 1G flows, needs the 31bit binary representation, so need the 61bit binary number representation altogether.
Example: set 1 token and represent 1 byte (1Byte), the flow of arranging certain user is no more than 100 megabyte per seconds (100MegaByte/S), so, in one second time (being T=1S), the flow that switch allows to pass through is 100MegaByte/S*1S=100MegaByte, that is to say have the flow of 100MegaByte can pass through switch, every circulation need be added 100MegaByte/1Byte=100M token.
In the above-mentioned example, A equals 200M token, and B equals 100MegaByte, and T equals 1S.A and B are directly proportional.The degree of depth of this token bucket is a 200M token.
Token bucket is represented the degree of depth with data, and the token bucket degree of depth is used for steady flow of balance and burst flow.The token bucket degree of depth of prior art is a fixed value, be generally the maximum stream flow of twice, if the token bucket degree of depth is too big, mean that the ability of controlling burst flow reduces, if promptly the user does not have flow for the previous period, token bucket will accumulate a lot of tokens, the big flow if the user happens suddenly, token bucket all can allow flow pass through, if a plurality of user is this situation, will cause network congestion so.If the token bucket degree of depth is superficial, then token bucket is easy to overflow, the flow that overflows just has been wasted, the user can not get respective service, in addition, if it is less to distribute to user's bandwidth, token in the token bucket is not enough to send long data bag, though user data steadily or the burst, the long data bag all can't pass through.
Adding token is not to carry out at any time, but periodically add, promptly only can add once in a period of time, if packet is a lot of in this time period, the corresponding token that deducts is just many, and token bucket just is easy to sky so, in case token bucket is in dummy status, before adding token operation arriving, user data package all can't be passed through next time.
Subtract token and carry out according to user data, time and data packet length that user data arrives are unpredictable, thus the time that subtracts token at random, quantity at random.The CAR controller needs within a certain period of time token bucket to be added token one time, the token number that obtains in unit interval is exactly the unit interval flow, and just bandwidth has just been controlled flow by the token number that the control unit interval adds, do not have token in the token bucket, user data just can't pass through.
The token number that the short packages correspondence deducts is few, and long data bag corresponding token is just many, and the token number that adds according to the unit of account time can corresponding a lot of short packages, and the number of corresponding long data bag can be not too many.
The general per 1 second tokens bucket of prior art CAR controller once adds token according to " every circulation interpolation token number " to token bucket at every turn.In ensuing this 1 second, deduct corresponding token number according to the customer flow that passes through, in token bucket, do not have till the token.This scheme realizes and operates fairly simplely, but has the shortcoming of following two aspects:
1, token bucket is controlled defectiveness to burst flow in short-term, and lower for the less user's long data bag percent of pass of bandwidth.
The token that the prior art token bucket adds is within a certain period of time fixed, the degree of depth and bandwidth are irrelevant, also fix, the fixing token bucket degree of depth is controlled defectiveness (if the user does not have flow for the previous period to burst flow, token bucket will accumulate a lot of tokens, and the big flow if the user happens suddenly, token bucket all can allow flow pass through, if a plurality of users are this situations, will cause network congestion so).For short packages, the token number that need subtract is less, than be easier to by, the long data bag need deduct more token number, if there are not enough tokens in the token bucket, this long data bag just can't pass through, and the token number that only adds just can pass through greater than the available data packet length.
For adding token cycle relatively shorter application and the lower situation of user's contracted flow, this phenomenon can be more outstanding, the both of these case token bucket can be very not dark, the token of phase interpolation is all few weekly, in other words, behind a long message, to wait a plurality of all after dates just can fill it up with a token that long message is required usually, the long message during this all is dropped.
2, because token bucket is deep, it is bigger to carry out the traffic constraints number of users if desired, just need take more memory space, may need to increase extra cost like this.
Summary of the invention
Technical problem to be solved by this invention is: overcome prior art and carry out data traffic when constraint, token bucket takies that memory space is big, the burst flow control defectiveness in short-term, the deficiency that long data bag percent of pass is lower, a kind of implementation method of data traffic constraint control is provided, improve the percent of pass of long data bag, in short-term burst flow is control effectively, and reduce the memory space that token bucket takies.
The present invention solves the problems of the technologies described above the technical scheme that is adopted to be: the implementation method of this data traffic constraint control, may further comprise the steps: construct a token bucket, comprise token bucket depth parameter and bandwidth parameter, the described token bucket degree of depth is that a fixed token bucket depth value adds a dynamic token bucket depth value; And described fixed token bucket depth value is the pairing token number of length of long data bag; Described dynamic token bucket depth value is the several times of maximum bandwidth, and changes according to the variation of maximum bandwidth; According to the flow of distributing to the user, periodically in token bucket, add corresponding token number, simultaneously, flow according to user's reality subtracts token from token bucket, the fixed token bucket degree of depth is below the dynamic token bucket degree of depth, if the token number in the token bucket during less than described fixed token bucket depth value, flow is illegal before subtracting token; When adding token next time, fixed token bucket depth value is replenished expire earlier, customer flow just can pass through then; Perhaps, according to the flow of distributing to the user, periodically from token bucket, deduct corresponding token number, simultaneously, flow according to user's reality adds token toward token bucket, the fixed token bucket degree of depth is above the dynamic token bucket degree of depth, if the token number in the token bucket during greater than described dynamic token bucket depth value, flow is illegal before adding token; When subtracting token next time, will cut above the token of dynamic token bucket depth value earlier, customer flow just can pass through then.
A token can be represented at least one byte.Periodically adding deduct time interval of token can be for less than 1 second.Described dynamic token bucket depth value is the several times of maximum bandwidth, and changes according to the variation of maximum bandwidth.Be preferably 2 times of maximum bandwidth, and be not under 0 the situation, to be at least 5 pairing token numbers of maximum data packet in bandwidth.Described fixed token bucket depth value is preferably the pairing token number of length of long data bag.
Beneficial effect of the present invention is: the present invention designs the token bucket data structure, increases arrearage mechanism, allows to pay in advance token, and the long data bag can use arrearage bucket token, has improved long data bag percent of pass; And, therefore can carry out shaping output to flow because the degree of depth of token bucket changes according to maximum bandwidth, and take into account the peaceful current stabilization amount of burst, short-term burst and long-period average flow are all controlled, guaranteed the effect of traffic shaping.And the present invention adopts the corresponding a plurality of bytes of token, and the cycle of tokens is shortened, thereby can reduce the memory space that token bucket takies.
Description of drawings
Fig. 1 is an operation principle schematic diagram of the present invention.
Embodiment
With embodiment the present invention is described in further detail with reference to the accompanying drawings below:
The invention provides a kind of implementation method of data traffic constraint control, construct a token bucket, comprise token bucket depth parameter and bandwidth parameter, the token bucket degree of depth is that a fixed token bucket depth value adds a dynamic token bucket depth value, and the fixed token bucket degree of depth is below the dynamic token bucket degree of depth.According to the flow of distributing to the user, periodically in token bucket, add corresponding token number, simultaneously, subtract token from token bucket according to the flow of user's reality, if the token number in the token bucket during less than described fixed token bucket depth value, flow is illegal before subtracting token; When adding token next time, fixed token bucket depth value is replenished expire earlier, customer flow just can pass through then.Dynamic token bucket depth value can adopt 2 times of maximum bandwidth, and changes according to the variation of maximum bandwidth, in bandwidth is not under 0 the situation, to be at least 5 pairing token numbers of maximum data packet.Fixed token bucket depth value can adopt the pairing token number of length of long data bag.
Token bucket Data Structure Design of the present invention is:
| Parameter name | The token bucket degree of depth | Bandwidth parameter |
| Scope | MAXDepth+MAXDataLength, MAXDepth equals 2*MAXBand, is not 0 situation for bandwidth, and MAXDepth is not less than the token number of 5 maximum data packet equivalences. | 0～MAXBand, MAXBand is lower than 10EA usually |
Be described as follows:
MAXDataLength: the length of long data bag, unit: token;
MAXDepth: token bucket depth capacity, unit: token;
MAXBand: maximum bandwidth is that token number, unit: n byte per second are added in every circulation
The present invention adds token and changes per eight signas into once, in addition, 1 token can be equivalent to 4 bytes, dark and the every circulation of token bucket depth capacity bucket is added token number and all is reduced to original 1/32nd so, like this for 1G Bit data flow, a token bucket only need take the memory space that is no more than 30bit, like this with respect to the 61bit of prior art, and has greatly reduced the memory space that token bucket takies.
Required memory space computational process is as follows: because 1G=10
9≈ 2
30, need in other words could represent that with 30 bits fully, the 1G bit is 1 second a flow, owing to the unit interval is 1/8000 second, 1/8000G ≈ 2
17, only need 17 bit representations here.Every circulation is added token number and is equivalent to the 1/8000G flow at most, needs the 17bit binary representation.Because the token bucket depth capacity is 2 times of 1G flows, therefore need the 18bit binary representation.Need the 35bit binary number representation like this, altogether.Because 1 token is represented 4 bytes, be scaled the 4*8 bit again, be reduced to originally 1/32, token number is added in the dark and every circulation of token bucket depth capacity bucket can also save 5bit respectively, so finally token bucket takies the 35-5-5=25bit space.
Like this, per eight signas are added a token, and 1 token is represented 4 bytes, i.e. 4*8=32 bit, and per second adds 32 bits/(1/8000)=256k bits per second like this.Take temperature with 1G bit max-flow, maximum bandwidth is 1G bps, and maximum stream flow is maximum bandwidth/byte bandwidth=(1G)/256k＜10E4 like this.
Suppose that 1 token is equivalent to N byte (Byte), total flow is F bit/s, and the minimum particle size of allocated bandwidth is K bit/s, the long M byte/s of maximum data packet, and total S bar stream need be CAR.So:
MAXBand=F ÷ K, this expression maximum bandwidth parameter, the space size that is used for determining to store " bandwidth parameter ";
MAXDepth=MAXBand * 2; This represents token bucket depth capacity, is used for determining the space size of storage " the token bucket degree of depth ".
These two parameters are used for determining maximum memory space, in physics realization, large or small, press the maximum space static allocation usually, and it is simple to do the token bucket addressing like this, is convenient to management.The actual token bucket is dark to be changed with the bandwidth parameter, dark the lying in the bandwidth parameter of actual bucket.
The following describes token bucket actual grade, the bandwidth of supposing to distribute to a certain data flow is a B minimum particle size, i.e. B * K bit/s,
So, when M * 5≤(B * K) * 2, this token sum of representing twice interpolation is greater than 5 maximum data packet when long, token bucket actual grade=(B * K) * 2; Otherwise, token bucket actual grade=M * 5.
In realization, actual bucket is dark to be dynamic change, because maximum bandwidth can dispose, bucket is dark to be changed according to maximum bandwidth, can control burst flow in short-term well like this.But for avoiding bucket dark too small, stipulated the minimum-depth of token bucket, it is long to be made as 5 maximum data packet here, thereby can guarantee passing through of long data bag, takes into account the control to burst flow simultaneously, steadily flow.
Here why the depth design of token bucket being become the twice of bandwidth and be not less than 5 times of long data bags, is to have different burstiness for the difference that allows each bar to connect according to PCR (peak rate).Here 2 and 5 can become, and 2 and 5 is more rational empirical values.But, be not 5 bags the longest that can happen suddenly continuously at least in every circulation.Because individual arrearage mechanism (DEFICIT) is arranged here, behind the token number of the token number in the bucket smaller or equal to the long equivalence of maximum data packet, just illegal by the flow of CAR, message may be dropped like this.
When the token in the token bucket was slightly less than the token of the long equivalence of maximum data packet, if arrival is a short bag, CAR just can pass through, if a long bag, CAR does not just pass through, and this part bandwidth has just been wasted.When the consequence that causes like this was exactly the linear speed transmission, long bag percent of pass was very low.Linear speed is the physics transfer rate of physical equipment, is exactly 1Gbit=1000Mbit as the linear speed of gigabit Ethernet.The present invention uses arrearage mechanism (DEFICIT), and the long degree of depth of a maximum data packet is brought up in the token bucket bottom.The various types of communication agreement is surrounded by different regulations to long data, is 65535byte as IP, and Ethernet (ETH2) is the longest to be 1518byte or the like.At above-mentioned long bag situation, be that flow is legal behind the CAR, in token bucket, deduct corresponding token.Token in the token bucket of this moment is lower than the long token number of equal value of maximum bag, and difference is exactly the arrearage number, and in this circulation, the token number in the bucket is smaller or equal to the long token number of equal value of maximum data packet, and later flow all is illegal.When token is added in next one circulation, must replenish arrearage.Like this, when linear speed sends, no matter grow the short bag of bag, percent of pass can not reduce, and because arrearage occurs in continuous two cycle periods, flow still is basicly stable.
The bucket of token number correspondence that can maximum data packet length in the token bucket is of equal value is regarded the arrearage bucket as, and remaining barrel deep branch is regarded effective bucket as.As shown in Figure 1, at this moment the residue token has a data block to arrive shown in A in the token bucket, and the controller judgement effectively has token in the bucket, takes out token from effective bucket, finds deficiency, just extracts the difference that needs from the arrearage bucket, shown in B.After this, owing to effectively do not had token in the bucket, in this circulation, it is all illegal that follow-up flow is CAR.Add token to next circulation, add full arrearage bucket earlier, remainder adds in effective bucket, shown in C.Ci Shi token bucket is shown in D so, and long bag has passed through, and bandwidth is waste not, does not also surpass the flow of CAR, and flow also relatively steadily.
Suppose to have in the token bucket 500 tokens, comprise the effectively token (120*4=480byte) in the bucket of 380 arrearage bucket tokens (380*4=1518byte) and 120.To short packages 64 bytes (64/4=16 token), arrearage bucket 500＞16 is arranged, allow to pass through, no arrearage bucket 120＞16 can pass through; If come the long message (1000/4=250 token) of one 1000 byte this moment, if there is not the arrearage bucket, 120＜250, just can't pass through, arrearage bucket 500＞250 has been arranged, allow to pass through.Arrearage can be repaid next time, generally speaking can not exceed predetermined amount of flow.
CAR controller mentioned above is " adding token " mode, can also be with " subtracting token " mode in the practical operation: construct a token bucket, comprise token bucket depth parameter and bandwidth parameter, the token bucket degree of depth is that a fixed token bucket depth value adds a dynamic token bucket depth value, and the fixed token bucket degree of depth is above the dynamic token bucket degree of depth.According to the flow of distributing to the user, periodically from token bucket, deduct corresponding token number, simultaneously, add token according to the past token bucket of the flow of user's reality, if the token number in the token bucket during greater than described dynamic token bucket depth value, flow is illegal before adding token; When subtracting token next time, will cut above the token of dynamic token bucket depth value earlier, customer flow just can pass through then.Operating principle is identical with " adding token " mode, repeats no more here.
The present invention designs the token bucket data structure, increases arrearage mechanism, allows to pay in advance token, and the long data bag can use arrearage bucket token, has improved long data bag percent of pass; And, therefore can carry out shaping output to flow because the degree of depth of token bucket changes according to maximum bandwidth, and take into account the peaceful current stabilization amount of burst, short-term burst and long-period average flow are all controlled, guaranteed the effect of traffic shaping.And the present invention adopts the corresponding a plurality of bytes of token, and the cycle of tokens is shortened, thereby can reduce the memory space that token bucket takies.
Those skilled in the art do not break away from essence of the present invention and spirit, can there be the various deformation scheme to realize the present invention, the above only is the preferable feasible embodiment of the present invention, be not so limit to interest field of the present invention, the equivalent structure that all utilizations specification of the present invention and accompanying drawing content are done changes, and all is contained within the interest field of the present invention.
Claims (4)
1, a kind of implementation method of data traffic constraint control is characterized in that, may further comprise the steps:
Construct a token bucket, comprise token bucket depth parameter and bandwidth parameter, the described token bucket degree of depth is that a fixed token bucket depth value adds a dynamic token bucket depth value, and described fixed token bucket depth value is the pairing token number of length of long data bag; Described dynamic token bucket depth value is the several times of maximum bandwidth, and changes according to the variation of maximum bandwidth;
According to the flow of distributing to the user, periodically in token bucket, add corresponding token number, simultaneously, flow according to user's reality subtracts token from token bucket, the fixed token bucket degree of depth is below the dynamic token bucket degree of depth, if the token number in the token bucket during less than described fixed token bucket depth value, flow is illegal before subtracting token; When adding token next time, fixed token bucket depth value is replenished expire earlier, customer flow just can pass through then;
Perhaps, according to the flow of distributing to the user, periodically from token bucket, deduct corresponding token number, simultaneously, flow according to user's reality adds token toward token bucket, the fixed token bucket degree of depth is above the dynamic token bucket degree of depth, if the token number in the token bucket during greater than described dynamic token bucket depth value, flow is illegal before adding token; When subtracting token next time, will cut above the token of dynamic token bucket depth value earlier, customer flow just can pass through then.
2, the implementation method of data traffic constraint control according to claim 1, it is characterized in that: a token is represented at least one byte.
3, the implementation method of data traffic constraint control according to claim 2, it is characterized in that: the time interval of the token that periodically adds deduct is less than 1 second.
4, the implementation method of controlling according to claim 1,2 or 3 described data traffic constraints, it is characterized in that: described dynamic token bucket depth value is 2 times of maximum bandwidth, and be not under 0 the situation, to be at least 5 pairing token numbers of maximum data packet in bandwidth.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100336935A CN100426738C (en) | 2005-03-16 | 2005-03-16 | Realizing method of data stream constraint control |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005100336935A CN100426738C (en) | 2005-03-16 | 2005-03-16 | Realizing method of data stream constraint control |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1835447A CN1835447A (en) | 2006-09-20 |
| CN100426738C true CN100426738C (en) | 2008-10-15 |
Family
ID=37003061
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2005100336935A Active CN100426738C (en) | 2005-03-16 | 2005-03-16 | Realizing method of data stream constraint control |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100426738C (en) |
Families Citing this family (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101272346B (en) * | 2008-04-29 | 2010-12-08 | 华为技术有限公司 | Method and device for packet flux monitoring |
| CN104038434A (en) * | 2013-03-08 | 2014-09-10 | 中兴通讯股份有限公司 | Traffic policing method and apparatus |
| CN103248580B (en) * | 2013-04-08 | 2016-08-24 | 汉柏科技有限公司 | A kind of method of high accuracy limiting speed of token bucket based on major-minor bucket |
| CN103188160B (en) * | 2013-04-18 | 2016-12-07 | 杭州华三通信技术有限公司 | A kind of flow control methods and retransmission unit |
| CN104243333B (en) * | 2013-06-24 | 2018-04-10 | 阿里巴巴集团控股有限公司 | A kind of flow control methods of address analysis protocol message |
| CN105763475B (en) * | 2014-12-15 | 2019-10-25 | 南京中兴新软件有限责任公司 | Connection control method and device |
| CN105791155B (en) * | 2014-12-24 | 2019-08-06 | 深圳市中兴微电子技术有限公司 | A kind of management method and device of congested flow |
| CN105991467B (en) * | 2015-04-15 | 2019-05-07 | 杭州迪普科技股份有限公司 | A kind of message forwarding method and device |
| CN106487606A (en) * | 2015-08-28 | 2017-03-08 | 阿里巴巴集团控股有限公司 | A kind of dispatching method for network tester and system |
| CN106961392B (en) * | 2016-01-12 | 2020-04-24 | 阿里巴巴集团控股有限公司 | Flow control method and device |
| CN106789720B (en) * | 2016-12-16 | 2020-06-16 | 无锡路通视信网络股份有限公司 | Dynamic token bucket generation method based on system hardware utilization rate |
| TWI691185B (en) * | 2018-01-15 | 2020-04-11 | 思銳科技股份有限公司 | Bandwidth management method for network switch and network system thereof |
| CN110768913A (en) * | 2018-07-27 | 2020-02-07 | 华为技术有限公司 | Flow control method and device |
| CN109861920A (en) * | 2019-01-16 | 2019-06-07 | 深圳市融汇通金科技有限公司 | A kind of method and device of elasticity current limliting |
| CN110275780B (en) * | 2019-06-26 | 2022-08-02 | 北京百度网讯科技有限公司 | Method and apparatus for restricting flow |
| CN115102908B (en) * | 2022-08-25 | 2023-03-10 | 珠海星云智联科技有限公司 | Method for generating network message based on bandwidth control and related device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6046979A (en) * | 1998-05-04 | 2000-04-04 | Cabletron Systems, Inc. | Method and apparatus for controlling the flow of variable-length packets through a multiport switch |
| WO2002021773A2 (en) * | 2000-09-08 | 2002-03-14 | Netscreen Technologies, Inc. | Guaranteed bandwidth sharing in a traffic shaping system |
| US6370116B1 (en) * | 1998-05-26 | 2002-04-09 | Alcatel Canada Inc. | Tolerant CIR monitoring and policing |
| US20030223369A1 (en) * | 2002-06-04 | 2003-12-04 | Eric Anderson | Traffic control at a network node |
-
2005
- 2005-03-16 CN CNB2005100336935A patent/CN100426738C/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6046979A (en) * | 1998-05-04 | 2000-04-04 | Cabletron Systems, Inc. | Method and apparatus for controlling the flow of variable-length packets through a multiport switch |
| US6370116B1 (en) * | 1998-05-26 | 2002-04-09 | Alcatel Canada Inc. | Tolerant CIR monitoring and policing |
| WO2002021773A2 (en) * | 2000-09-08 | 2002-03-14 | Netscreen Technologies, Inc. | Guaranteed bandwidth sharing in a traffic shaping system |
| US20030223369A1 (en) * | 2002-06-04 | 2003-12-04 | Eric Anderson | Traffic control at a network node |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1835447A (en) | 2006-09-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100426738C (en) | Realizing method of data stream constraint control | |
| US6829649B1 (en) | Method an congestion control system to allocate bandwidth of a link to dataflows | |
| US6801500B1 (en) | Method and apparatus for providing reserved rates to multiple flows on a network interface | |
| US7474668B2 (en) | Flexible multilevel output traffic control | |
| CN100593926C (en) | Method and apparatus for implied specification of service quality in network | |
| CN102420776B (en) | Method and system for dynamically regulating portal resource allocation threshold value | |
| EP2761826B1 (en) | Attribution of congestion contributions | |
| CN102413063A (en) | Method and system for dynamically adjusting allocation threshold value of output port resources | |
| CN101803313A (en) | Methods and apparatus for providing congestion information | |
| JP2000236344A (en) | Atm switch and dynamic threshold setting method | |
| IL175840A (en) | Distributed medium access control for broadband access systems | |
| US7609633B2 (en) | Bandwidth policer with compact data structure | |
| US20070058651A1 (en) | Method, system and program product for setting a transmission rate in a network | |
| US8228797B1 (en) | System and method for providing optimum bandwidth utilization | |
| US7836201B2 (en) | Methods and systems for providing efficient provisioning of data flows | |
| Ahmed et al. | Systems approach to modeling the token bucket algorithm in computer networks | |
| CN100450103C (en) | Method and device for flux plastic | |
| EP2388978B1 (en) | Methods and devices for determining network link load | |
| Zhang et al. | Resource allocation and congestion control in distributed sensor networks-a network calculus approach | |
| Abraham et al. | A new approach for asynchronous distributed rate control of elastic sessions in integrated packet networks | |
| CN1697348B (en) | Dynamic bandwidth allocation method and device in multiple operation types, and optical line terminal | |
| US20060088032A1 (en) | Method and system for flow management with scheduling | |
| Jun et al. | Hierarchical fair queuing: single-step approximation of hierarchical-GPS | |
| Fulp et al. | The economic impact of network pricing intervals | |
| Lee et al. | WF/sup 2/QM: a worst-case fair weighted fair queueing with maximum rate control |
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 |