CN105306290B - A kind of prediction technique and device of flow entry scale - Google Patents

A kind of prediction technique and device of flow entry scale Download PDF

Info

Publication number
CN105306290B
CN105306290B CN201410340073.5A CN201410340073A CN105306290B CN 105306290 B CN105306290 B CN 105306290B CN 201410340073 A CN201410340073 A CN 201410340073A CN 105306290 B CN105306290 B CN 105306290B
Authority
CN
China
Prior art keywords
flow entry
moment
flow
interchanger
scale
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
Application number
CN201410340073.5A
Other languages
Chinese (zh)
Other versions
CN105306290A (en
Inventor
王延松
喻敬海
吴少勇
吴春明
鲁昕佳
周伯阳
高文
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.)
Nanjing ZTE New Software Co Ltd
Original Assignee
Nanjing ZTE New Software 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 Nanjing ZTE New Software Co Ltd filed Critical Nanjing ZTE New Software Co Ltd
Priority to CN201410340073.5A priority Critical patent/CN105306290B/en
Publication of CN105306290A publication Critical patent/CN105306290A/en
Application granted granted Critical
Publication of CN105306290B publication Critical patent/CN105306290B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The invention discloses a kind of prediction techniques of flow entry scale, comprising: obtains current tnThe average flow entry scale of the flow table of moment interchanger;Determine current tnThe estimated value of the average failure rate of flow entry and flow entry advanced the speed in flow table described in moment;According to the current tnThe estimated value of the average flow entry scale at moment, the average failure rate of the flow entry and the flow entry advanced the speed, to subsequent time tn+1When flow entry scale predicted.The present invention further simultaneously discloses a kind of prediction meanss of flow entry scale.

Description

A kind of prediction technique and device of flow entry scale
Technical field
The present invention relates to flow entry scales in software defined network (SDN, Software-Defined Networking) Predicting Technique more particularly to a kind of prediction technique and device of flow entry scale.
Background technique
In SDN, control plane is made of multiple controllers, and each controller controls multiple switch, each exchange Machine preserves respective flow table (Flow Table) to determine forwarding strategy.When new stream reaches, interchanger is first and in flow table Flow entry matched, if successful match, this stream is forwarded according to the rule in flow table;If it fails to match, The referred to as mismatch (Mismatch) of flow table, data packet is transmitted to controller by control channel by interchanger at this time, first by controlling Device is handed down to interchanger after determining forwarding strategy again, is ultimately written flow table.
The mismatch of flow table will lead to the generation of control flow, limit forwarded performance, and flow table is by expensive Content adressable memory composition, it is very limited that the number of flow entry is restricted by content adressable memory quantity, if handed over Flow table of changing planes has expired, and will lead to the loss of data packet.In wan environment, the arrival of stream has sudden, this burst flow More stringent requirements are proposed for availability of the presence of amount to flow table;Therefore, flow entry scale is predicted, to prevent due to stream List item number is more than the physical equipments upper limits such as memory and flow table availability is caused to decline, and is current urgent problem to be solved.
Summary of the invention
In view of this, an embodiment of the present invention is intended to provide the prediction techniques and device of a kind of flow entry scale, so that business The dynamic change of flow entry scale can be perceived, is lost because of switch device resource more than data packet caused by its upper limit with reducing Mistake problem, and then improve the utilization rate of flow table.
In order to achieve the above objectives, the technical scheme of the present invention is realized as follows:
The embodiment of the invention provides a kind of prediction techniques of flow entry scale, which comprises
Obtain current tnThe average flow entry scale of the flow table of moment interchanger;
Determine current tnIn flow table described in moment the average failure rate of flow entry and flow entry advance the speed estimate Evaluation;
According to the current tnAverage flow entry scale, the average failure rate of the flow entry and the flow table at moment The estimated value of item advanced the speed, to subsequent time tn+1When flow entry scale predicted.
It is described to obtain current t in above schemenBefore the average flow entry scale of moment interchanger flow table, the method Further comprise: obtaining current tnThe flow entry scale of moment interchanger flow table, and the flow entry scale is saved to described The corresponding flow entry scale queue of interchanger, when the item number for the flow entry scale record that the queue saves is more than the queue When maximum value, an earliest data is deleted according to first-in first-out method FIFO and is recorded.
It is described to obtain current t in above schemenThe average flow entry scale of the flow table of moment interchanger includes: that acquisition is worked as Preceding tnThe flow entry scale of moment interchanger flow table, and current t is obtained by following formulanThe mean flow of moment interchanger flow table List item scale:
As n≤δ,
As n > δ,
Wherein,For tnThe average flow entry scale of the flow table of i-th of interchanger of moment;For tnI-th of moment The flow entry scale of the flow table of interchanger;δ is that the corresponding flow entry scale queue of i-th of interchanger saves flow entry scale record Item number maximum value;tnFor receive i-th of interchanger nth STATS_REPLY message time.
In above scheme, the current t of determinationnThe average failure rate of flow entry includes: to obtain t in flow table described in momentn Moment, unit time tn-tn-1I-th of interchanger flow table in flow entry failure rateAnd according to following Formula determines current tnThe average failure rate of flow entry in flow table described in moment:
As n≤δ,
As n > δ,
Wherein,For tnThe average failure rate of flow entry in the flow table of i-th of interchanger of moment;For tn Moment, unit time tn-tn-1I-th of interchanger flow table in flow entry failure rate;tnTo receive i-th of interchanger Nth STATS_REPLY message time.
In above scheme, the acquisition tnMoment, unit time tn-tn-1I-th of interchanger flow table in flow entry The rate of failureLater, the method further includes rates that the flow entry failsIt saves extremely The corresponding flow entry failure rate queue F of i-th of interchangeri, as queue FiThe item number of the flow entry failure rate record of middle preservation More than the FiMaximum value when, according to FIFO delete FiIn earliest data record.
In above scheme, the current t of determinationnThe estimated value of moment flow entry advanced the speed includes: to obtain tnMoment, Unit time is tn-tn-1I-th of interchanger flow table in the increased rate of flow entryAnd it is true according to following formula Being averaged for flow entry is advanced the speed in the fixed flow table:
As n≤δ,
As n > δ,
Wherein,For tnBeing averaged for flow entry is advanced the speed in the flow table of i-th of interchanger of moment;For tnWhen It carves, unit time tn-tn-1I-th of interchanger flow table in the increased rate of flow entry;tnTo receive i-th of interchanger The time of nth STATS_REPLY message;
Then according to described inAcquisition meets Poisson distributionCumulative probabilityX minimum value Xmin, the minimum value X of the XminThe as estimated value of flow entry advanced the speed;K is Positive integer, γ ∈ (0,1).
In above scheme, the acquisition tnMoment, unit time tn-tn-1I-th of interchanger flow table in flow entry Increased rateLater, the method further includes: will be describedIt saves corresponding to i-th of interchanger Flow entry is advanced the speed queue Gi, as queue GiThe advance the speed item number of record of the flow entry of middle preservation is more than the GiMaximum When value, G is deleted according to FIFOiIn earliest data record.
It is described according to the current t in above schemenThe average mistake of the average flow entry scale, the flow entry at moment The estimated value advanced the speed for imitating rate and the flow entry, to subsequent time tn+1When flow entry scale carry out prediction include:
According to the current tnAverage flow entry scale, the average failure rate of the flow entry and the flow table at moment The estimated value of item advanced the speed, calculates the tn+1When flow entry scale be equal to the current tnThe average flow entry at moment Scale and current tnThe sum for the estimated value of the flow entry at moment advanced the speed, then with current tnFlow entry in flow table described in moment Average failure rate difference.
The embodiment of the invention also provides a kind of prediction meanss of flow entry scale, described device include: obtain module, really Cover half block and prediction module;Wherein,
The acquisition module, for obtaining current tnThe average flow entry scale of the flow table of moment interchanger;
The determining module, for determining current tnThe average failure rate and flow table of flow entry in flow table described in moment The estimated value of item advanced the speed;
The prediction module, for according to the current tnThe average flow entry scale at moment, the flow entry are averaged The estimated value of failure rate and the flow entry advanced the speed, to subsequent time tn+1When flow entry scale predicted.
In above scheme, the acquisition module is further used for obtaining current tnThe flow entry of moment interchanger flow table is advised Mould, and the flow entry scale is saved into flow entry scale queue corresponding to the interchanger, when the stream that the queue saves When the item number of list item scale record is more than the maximum value of the queue, an earliest data is deleted according to FIFO and is recorded.
In above scheme, the acquisition module obtains current tnThe average flow entry scale packet of the flow table of moment interchanger Include: the acquisition module obtains current t by following formulanThe average flow entry scale of moment interchanger flow table:
As n≤δ,
As n > δ,
Wherein,For tnThe average flow entry scale of the flow table of i-th of interchanger of moment;For tnI-th of moment The flow entry scale of the flow table of interchanger;δ is that the corresponding flow entry scale queue of i-th of interchanger saves flow entry scale record Item number maximum value;tnFor receive i-th of interchanger nth STATS_REPLY message time.
In above scheme, the determining module determines current tnThe average failure rate packet of flow entry in flow table described in moment Include: the determining module obtains tnMoment, unit time tn-tn-1I-th of interchanger flow table in flow entry failure speed RateAnd current t is determined according to following formulanThe average failure rate of flow entry in flow table described in moment:
As n≤δ,
As n > δ,
Wherein,For tnThe average failure rate of flow entry in the flow table of i-th of interchanger of moment;For tn Moment, unit time tn-tn-1I-th of interchanger flow table in flow entry failure rate;tnTo receive i-th of interchanger Nth STATS_REPLY message time.
In above scheme, the determining module is further used for the rate that the flow entry failsIt saves To the corresponding flow entry failure rate queue F of i-th of interchangeri, as queue FiThe item of the flow entry failure rate record of middle preservation Number is more than the FiMaximum value when, according to FIFO delete FiIn earliest data record.
In above scheme, the determining module determines current tnThe estimated value of moment flow entry advanced the speed includes: institute It states determining module and obtains tnMoment, unit time tn-tn-1I-th of interchanger flow table in the increased rate of flow entryAnd determine that being averaged for flow entry is advanced the speed in the flow table according to following formula:
As n≤δ,
As n > δ,
Wherein,For tnBeing averaged for flow entry is advanced the speed in the flow table of i-th of interchanger of moment;For tnWhen It carves, unit time tn-tn-1I-th of interchanger flow table in the increased rate of flow entry;tnTo receive i-th of interchanger The time of nth STATS_REPLY message;
Then according to described inAcquisition meets Poisson distributionCumulative probabilityX minimum value Xmin, the minimum value X of the XminThe as estimated value of flow entry advanced the speed;K is Positive integer, γ ∈ (0,1).
In above scheme, the determining module, being further used for will be describedIt saves corresponding to i-th of interchanger Flow entry is advanced the speed queue Gi, as queue GiThe advance the speed item number of record of the flow entry of middle preservation is more than the GiMaximum When value, G is deleted according to FIFOiIn earliest data record.
In above scheme, the prediction module is according to the current tnAverage flow entry scale, the flow entry at moment Average failure rate and the flow entry the estimated value advanced the speed, to subsequent time tn+1When flow entry scale carry out Prediction includes:
The prediction module is according to the current tnAverage flow entry scale, the average failure speed of the flow entry at moment The estimated value of rate and the flow entry advanced the speed, calculates the tn+1When flow entry scale be equal to the current tnMoment Average flow entry scale and current tnThe sum for the estimated value of the flow entry at moment advanced the speed, then with current tnDescribed in moment The difference of the average failure rate of flow entry in flow table.
The prediction technique and device of flow entry scale provided by the embodiment of the present invention, obtain current tnMoment interchanger The average flow entry scale of flow table;Determine current tnThe average failure rate of flow entry and flow entry in flow table described in moment The estimated value advanced the speed;According to the current tnAverage flow entry scale, the average failure rate of the flow entry at moment And the estimated value of the flow entry advanced the speed, to subsequent time tn+1When flow entry scale predicted.In this way, can be with Flow entry is dynamically configured according to prediction, prevents from causing under flow table availability since flow entry number is more than the upper limit of flow table Drop;Decision-making foundation can be provided for the setting of flow entry, reduce data packet packet loss, improve the utilization rate of flow table, and working can It leans on, accuracy height.
Detailed description of the invention
Fig. 1 is the prediction technique flow diagram of one flow entry scale of the embodiment of the present invention;
Fig. 2 is the prediction technique flow diagram of two flow entry scale of the embodiment of the present invention;
Fig. 3 is the composed structure schematic diagram of the prediction meanss of flow entry of embodiment of the present invention scale.
Specific embodiment
In SDN, controller is to undertake management number according to the route test software of open flows (OpenFlow) Protocol Design According to stream, the function of Configuration network equipment.Controller is turned by modifying flow entry, Lai Guanli interchanger stream in interchanger flow table Hair.Interchanger determines the processing mode to message by inquiry flow table item.Exit passageway based on OpenFlow agreement is used Interchanger and controller are connected, so that the two is communicated, referred to as control channel.
Controller modifies the flow table of interchanger by way of sending stream modification (FLOW_MOD) message to interchanger, real Existing control function.When a flow entry is added in flow table, which is not used by within a certain period of time, it will failure, this When can trigger stream and remove (FLOW_REMOVED) message, this message issued controller by interchanger, informs that the flow entry fails. Controller inquires the current status information of interchanger by sending statistics request (STATS_REQUEST) message to interchanger, Interchanger, which sends statistics, replys (STATS_REPLY) message and is replied, STATS_REQUEST message and STATS_REPLY The main part type of message is the STATS type of OpenFlow protocol definition, and content includes exchanger information, single in flow table Statistical information, the collect statistics information of stream, flow table information, physical port information, queuing message of item stream etc..
In embodiments of the present invention, current t is obtainednThe average flow entry scale of the flow table of moment interchanger;Determine current tn The estimated value of the average failure rate of flow entry and flow entry advanced the speed in flow table described in moment;According to described current tnThe estimation of the average flow entry scale, the average failure rate of the flow entry and the flow entry at moment advanced the speed Value, to subsequent time tn+1Flow entry scale predicted.
Fig. 1 is the prediction technique flow diagram of one flow entry scale of the embodiment of the present invention, as shown in Figure 1, the present embodiment The prediction technique of flow entry scale includes:
Step 101: obtaining current tnThe average flow entry scale of the flow table of moment interchanger;
Before this step, the method further includes: controller obtains tnMoment, i-th of interchanger sent nth Flow entry scale when STATS_REPLY messageAnd by the flow entry scaleIt saves to described i-th and exchanges The corresponding flow entry scale queue D of machinei, as the DiThe item number n of the flow entry scale record of preservation is more than the DiMaximum value When δ, an earliest data is deleted according to first-in first-out method (FIFO, First in, First out) and is recorded;Wherein, the stream List item scale be flow table in flow entry item number, the δ be it is preset be greater than 0 integer value, specific value can foundation Actual conditions are set;
The acquisition tnI-th of interchanger of moment sends flow entry scale when nth STATS_REPLY messagePacket Include: i-th of interchanger receives the nth STATS_REQUEST message that controller is sent, and in tnMoment by control channel to The controller sends nth STATS_REPLY message, and controller extracts the flow table for including in the STATS_REPLY message Item scale
It is described to obtain current tnThe average flow entry scale of the flow table of moment interchanger includes: to obtain current tnMoment exchange The flow entry scale of machine flow table, and current t is obtained according to following formulanThe average flow entry scale of the flow table of moment interchanger:
As n≤δ,
As n > δ
Wherein,For tnThe average flow entry scale of the flow table of i-th of interchanger of moment;For tnI-th of moment The flow entry scale of the flow table of interchanger;δ is queue DiSave the maximum value of the item number of flow entry scale record;tnTo receive i-th The time of the nth STATS_REPLY message of a interchanger;
Here, the current tnMoment is the time for receiving the nth STATS_REPLY message of i-th of interchanger.
Step 102: determining current tnThe average failure rate of flow entry and the increase of flow entry in flow table described in moment The estimated value of rate;
Before this step, the method further includes: controller is obtained in tn-1Moment is to tnReceived between moment The number for the FLOW_REMOVED message that i interchanger is sentAnd tn-1Moment is to tnBetween moment controller to The type that i-th of interchanger issues is to increase the number of the FLOW_MOD message of (ADD)That is: the acquisition unit time is tn-tn-1I-th of interchanger flow table in flow entry failure rateAnd the unit time is tn-tn-1I-th The increased rate of flow entry in the flow table of a interchangerAnd it will be describedIt saves corresponding to i-th of interchanger Flow entry failure rate queue Fi, will be describedIt saves to the corresponding flow entry of i-th of interchanger and advances the speed queue Gi, as queue FiThe item number m of the flow entry failure rate record of middle preservation is more than the FiMaximum value α or queue GiMiddle preservation The advance the speed item number s of record of flow entry be more than the GiMaximum value β when, according to first-in first-out method FIFO delete FiOr Gi In earliest data record;Here, α=β=δ is set, to guarantee synchronization, the queue Di, queue FiAnd queue Gi The item number of the record of middle storage is identical, due to tn-1Moment is to tnMoment can save the failure rate and flow entry of a flow entry It advances the speed, and tn-1Moment is to tnMoment will receive a STATS_REPLY message, a record flow entry scale, therefore, M is equal to s and is equal to n in the present embodiment;
The average failure rate of flow entry includes: controller according to described in the determination of following formula in the determination flow table The average failure rate of flow entry in flow table:
As n≤δ,
As n > δ,
Wherein,For tnThe average failure rate of flow entry in the flow table of i-th of interchanger of moment;For tn Moment, unit time tn-tn-1I-th of interchanger flow table in flow entry failure rate;tnTo receive i-th of interchanger Nth STATS_REPLY message time;
The estimated value advanced the speed for determining flow entry in the flow table includes: firstly, controller works as n≤δ according to formula When,As n > δ,Obtain tnI-th of interchanger of moment Being averaged for flow entry is advanced the speed wherein in flow table, is tnFlow entry is flat in the flow table of i-th of interchanger of moment Advance the speed;For tnMoment, unit time tn-tn-1I-th of interchanger flow table in the increased speed of flow entry Rate;tnFor receive i-th of interchanger nth STATS_REPLY message time;
Then according to described inAcquisition meets Poisson distributionCumulative probabilityX minimum value Xmin, the minimum value X of the XminThe as estimated value of flow entry advanced the speed;Its In, γ is preset confidence level, and the value range of γ is (0,1), and k is positive integer;
Here, it should be noted that inevitable time relationship, execution sequence be not interchangeable for step 101 and step 102.
Step 103: according to the current tnThe average flow entry scale at moment, the average failure rate of the flow entry and The estimated value of the flow entry advanced the speed, to subsequent time tn+1When flow entry scale predicted;
This step specifically includes: controller is according to the current tnThe average flow entry scale at moment, the flow entry The estimated value of average failure rate and the flow entry advanced the speed calculates tn+1When flow entry scale be equal to tnMoment Average flow entry scale and tnThe sum for the estimated value of the flow entry at moment advanced the speed, then with tnThe flow table at moment The difference of the average failure rate of item, it may be assumed that according to described according to tnThe average failure rate and t of flow entry in the flow table of average i-th of interchanger of flow entry scale moment of the flow table of i-th of interchanger of momentnI-th of interchanger of moment The estimated value X of flow entry to advance the speed in flow tablemin, obtain tn+1The predicted value of the flow entry scale of i-th of interchanger of moment
Fig. 2 is the prediction technique flow diagram of two flow entry scale of the embodiment of the present invention, as shown in Fig. 2, the present embodiment The prediction technique of flow entry scale includes:
Step 201: obtaining current tnThe flow entry scale of moment interchanger flow table;
Here, the current tnMoment is the time for receiving the nth STATS_REPLY message of i-th of interchanger transmission;
It is described to obtain current tnThe flow entry scale of moment interchanger flow table includes: that i-th of interchanger receives controller hair The nth STATS_REQUEST message sent, and in tnMoment sends nth STATS_ to the controller by control channel REPLY message, controller extract the flow entry scale for including in the STATS_REPLY message
Further, the method also includes: controllers by the flow entry scaleIt saves to described i-th and exchanges The corresponding flow entry scale queue D of machinei, as the DiThe item number n of the flow entry scale record of preservation is more than the DiMaximum value When δ, an earliest data is deleted according to FIFO and is recorded;Wherein, the flow entry scale is the item number of flow entry in flow table, institute State δ be it is preset be greater than 0 integer value, specific value can be set according to actual conditions.
Step 202: obtaining current tnThe average flow entry scale of the flow table of moment interchanger;
This step specifically includes: controller obtains current t according to following formulanThe average flow table of the flow table of moment interchanger Item scale:
As n≤δ,
As n > δ,
Wherein,For tnThe average flow entry scale of the flow table of i-th of interchanger of moment;For tnI-th of moment The flow entry scale of the flow table of interchanger;δ is queue DiSave the maximum value of the item number of flow entry scale record;tnTo receive i-th The time of the nth STATS_REPLY message of a interchanger.
Step 203: obtaining and determine current tnThe increase speed of the failure rate of flow entry and flow entry in flow table described in moment Rate;
This step specifically includes: controller is obtained in tn-1Moment is to tnWhat i-th of the interchanger received between the moment was sent The number of FLOW_REMOVED messageAs tnMoment, unit time tn-tn-1I-th of interchanger flow table The rate of middle flow entry failure, obtains tn-1Moment is to tnThe type that controller issues between moment is the FLOW_ for increasing (ADD) The number of MOD messageAs tnMoment, unit time tn-tn-1I-th of interchanger flow table in flow entry increase The rate added;
Further, the method also includes: controllers will be describedIt saves to the corresponding stream of i-th of interchanger List item failure rate queue Fi, will be describedIt saves to the corresponding flow entry of i-th of interchanger and advances the speed queue Gi, when Queue FiThe item number m of the flow entry failure rate record of middle preservation is more than the FiMaximum value α or queue GiThe stream of middle preservation The advance the speed item number s of record of list item is more than the GiMaximum value β when, according to first-in first-out method delete FiOr GiIn it is earliest One data record;Here, α=β=δ is set, to guarantee synchronization, the queue Di, queue FiAnd queue GiMiddle storage The item number of record is identical, due to tn-1Moment is to tnMoment can save the failure rate an of flow entry and flow entry is advanced the speed, And tn-1Moment is to tnMoment will receive a STATS_REPLY message, a record flow entry scale, therefore, in the present embodiment M is equal to s and is equal to n.
Step 204: determining current tnThe average failure rate of flow entry and the increase of flow entry in flow table described in moment The estimated value of rate;
Here, the average failure rate of flow entry includes: that controller is true according to following formula in the determination flow table The average failure rate of flow entry in the fixed flow table:
As n≤δ,
As n > δ,
Wherein,For tnThe average failure rate of flow entry in the flow table of i-th of interchanger of moment;For tn Moment, unit time tn-tn-1I-th of interchanger flow table in flow entry failure rate;tnTo receive i-th of interchanger Nth STATS_REPLY message time;
The estimated value advanced the speed for determining flow entry in the flow table includes: firstly, controller works as n≤δ according to formula When,As n > δ,Obtain tnI-th of interchanger of moment Being averaged for flow entry is advanced the speed wherein in flow table, is tnFlow entry is flat in the flow table of i-th of interchanger of moment Advance the speed;For tnMoment, unit time tn-tn-1I-th of interchanger flow table in the increased speed of flow entry Rate;tnFor receive i-th of interchanger nth STATS_REPLY message time;
Then according to described inAcquisition meets Poisson distributionCumulative probabilityX minimum value Xmin, the minimum value X of the XminThe as estimated value of flow entry advanced the speed;Its In, γ is preset confidence level, and the value range of γ is (0,1), and k is positive integer;
What needs to be explained here is that the not no inevitable time relationship of step 201-202 and step 203-204, operation order It is interchangeable.
Step 205: according to the current tnThe average flow entry scale at moment, the average failure rate of the flow entry and The estimated value of the flow entry advanced the speed, to subsequent time tn+1Flow entry scale predicted;
This step specifically includes: according to tnThe average flow entry scale of the flow table of i-th of interchanger of momenttnMoment The average failure rate of flow entry in the flow table of i-th of interchangerAnd tnFlow entry in the flow table of i-th of interchanger of moment The estimated value X to advance the speedmin, obtain tn+1The predicted value of the flow entry scale of i-th of interchanger of moment isHere, the subsequent time is tn+1Moment.
Fig. 3 is the composed structure schematic diagram of the prediction meanss of flow entry of embodiment of the present invention scale, the prediction meanss position In on the controller, as shown in figure 3, the composition of the prediction meanss of flow entry scale of the embodiment of the present invention includes: acquisition module 31, determining module 32 and prediction module 33;Wherein,
The acquisition module 31, for obtaining current tnThe average flow entry scale of the flow table of moment interchanger;
The determining module 32, for determining current tnThe average failure rate and stream of flow entry in flow table described in moment The estimated value of list item advanced the speed;
The prediction module 33, for according to the current tnThe average flow entry scale at moment, the flow entry it is flat The estimated value of equal failure rate and the flow entry advanced the speed, to subsequent time tn+1When flow entry scale predicted.
Further, the acquisition module 31, is further used for obtaining current tnThe flow entry of moment interchanger flow table is advised Mould, and the flow entry scale is saved into flow entry scale queue D corresponding to the interchangeri, as the queue DiIt saves The item number n of flow entry scale record is more than the queue DiMaximum value δ when, delete an earliest data according to FIFO and remember Record;Wherein, the flow entry scale be flow table in flow entry item number, the δ be it is preset be greater than 0 integer value, tool Body numerical value can be set according to actual conditions.
Further, the acquisition module 31 obtains current tnThe average flow entry scale packet of the flow table of moment interchanger Include: the acquisition module 31 obtains current t by following formulanThe average flow entry scale of moment interchanger flow table:
As n≤δ,
As n > δ,
Wherein,For tnThe average flow entry scale of the flow table of i-th of interchanger of moment;For tnI-th of moment The flow entry scale of the flow table of interchanger;δ is that the corresponding flow entry scale queue of i-th of interchanger saves flow entry scale record Item number maximum value;tnFor receive i-th of interchanger nth STATS_REPLY message time.
Further, the determining module 32 determines current tnThe average failure rate packet of flow entry in flow table described in moment Include: the determining module 32 obtains tnMoment, unit time tn-tn-1I-th of interchanger flow table in flow entry failure RateAnd current t is determined according to following formulanThe average failure rate of flow entry in flow table described in moment:
As n≤δ,
As n > δ,
Wherein,For tnThe average failure rate of flow entry in the flow table of i-th of interchanger of moment;For tn Moment, unit time tn-tn-1I-th of interchanger flow table in flow entry failure rate;tnTo receive i-th of interchanger Nth STATS_REPLY message time;
Here, the unit time is tn-tn-1I-th of interchanger flow table in flow entry failure rateAs in tn-1Moment is to tnThe number for the FLOW_REMOVED message that i-th of the interchanger received between moment is sent Mesh.
Further, the determining module 32, is further used for the rate that the flow entry failsIt saves To the corresponding flow entry failure rate queue F of i-th of interchangeri, as queue FiThe item of the flow entry failure rate record of middle preservation Number m is more than the FiMaximum value α when, according to FIFO delete FiIn earliest data record.
Further, the determining module 32 determines current tnThe estimated value of moment flow entry advanced the speed includes: institute It states determining module 32 and obtains tnMoment, unit time tn-tn-1I-th of interchanger flow table in the increased rate of flow entryAnd determine that being averaged for flow entry is advanced the speed in the flow table according to following formula:
As n≤δ,
As n > δ,
Wherein,For tnBeing averaged for flow entry is advanced the speed in the flow table of i-th of interchanger of moment;For tnWhen It carves, unit time tn-tn-1I-th of interchanger flow table in the increased rate of flow entry;tnTo receive i-th of interchanger The time of nth STATS_REPLY message;Here, the unit time is tn-tn-1I-th of interchanger flow table in flow table The increased rate of itemAs tn-1Moment is to tnThe type that controller issues between moment is the FLOW_ for increasing (ADD) The number of MOD message;
Then according to described inAcquisition meets Poisson distributionCumulative probabilityX minimum value Xmin, the minimum value X of the XminThe as estimated value of flow entry advanced the speed;K is Positive integer, γ are confidence level, γ ∈ (0,1).
Further, the determining module 32, being further used for will be describedIt saves corresponding to i-th of interchanger Flow entry is advanced the speed queue Gi, as queue GiThe advance the speed item number s of record of the flow entry of middle preservation is more than the GiMaximum When value β, G is deleted according to FIFOiIn earliest data record;Here, α=β=δ is set, it is described to guarantee synchronization Queue Di, queue FiAnd queue GiThe item number of the record of middle storage is identical, due to tn-1Moment is to tnMoment can save a flow table The failure rate and flow entry of item are advanced the speed, and tn-1Moment is to tnMoment will receive a STATS_REPLY message, record Flow entry scale, therefore, m is equal to s and is equal to n in the present embodiment;.
The prediction module 33 is according to the current tnThe average failure of the average flow entry scale, the flow entry at moment The estimated value of rate and the flow entry advanced the speed, to subsequent time tn+1When flow entry scale carry out prediction include:
The prediction module 33 is according to the current tnThe average failure of the average flow entry scale, the flow entry at moment The estimated value of rate and the flow entry advanced the speed, calculates the tn+1When flow entry scale be equal to the current tnWhen The average flow entry scale carved and current tnThe sum for the estimated value of the flow entry at moment advanced the speed, then with current tnMoment institute State the difference of the average failure rate of flow entry in flow table, it may be assumed that according to described according to tnThe flow table of i-th of interchanger of moment is averaged The average failure rate and t of flow entry in the flow table of i-th of interchanger of flow entry scale momentnMoment i-th The estimated value X of flow entry to advance the speed in the flow table of a interchangermin, obtain tn+1The flow entry of i-th of interchanger of moment is advised The predicted value of mould
Acquisition module 31, determining module 32 and prediction module 33 described above can be by the central processing units in controller (CPU, Central Processing Unit) or digital signal processor (DSP, Digital Signal Processor), Or field programmable gate array (FPGA, Field Programmable Gate Array) is realized.
The foregoing is only a preferred embodiment of the present invention, is not intended to limit the scope of the present invention.

Claims (14)

1. a kind of prediction technique of flow entry scale, which is characterized in that the described method includes:
Obtain current tnThe average flow entry scale of the flow table of moment interchanger;
Determine current tnThe estimated value of the average failure rate of flow entry and flow entry advanced the speed in flow table described in moment;
According to the current tnThe average flow entry scale at moment, the average failure rate of the flow entry and the flow entry The estimated value advanced the speed, to subsequent time tn+1When flow entry scale predicted;
It is described according to the current tnAverage flow entry scale, the average failure rate of the flow entry and the flow table at moment The estimated value of item advanced the speed, to subsequent time tn+1When flow entry scale carry out prediction include:
According to the current tnThe average flow entry scale at moment, the average failure rate of the flow entry and the flow entry The estimated value advanced the speed calculates the tn+1When flow entry scale be equal to the current tnThe average flow entry scale at moment With current tnThe sum for the estimated value of the flow entry at moment advanced the speed, then with current tnFlow entry is flat in flow table described in moment The difference of equal failure rate.
2. method according to claim 1, which is characterized in that described to obtain current tnThe average flow table of moment interchanger flow table Before item scale, the method further includes: obtain current tnThe flow entry scale of moment interchanger flow table, and by the stream List item scale saves flow entry scale queue corresponding to the interchanger, when the flow entry scale record that the queue saves When item number is more than the maximum value of the queue, an earliest data is deleted according to first-in first-out method FIFO and is recorded.
3. method according to claim 1, which is characterized in that described to obtain current tnThe mean flow of the flow table of moment interchanger List item scale includes: to obtain current tnThe flow entry scale of moment interchanger flow table, and current t is obtained by following formulanMoment The average flow entry scale of interchanger flow table:
As n≤δ,
As n > δ,
Wherein,For tnThe average flow entry scale of the flow table of i-th of interchanger of moment;For tnI-th of interchanger of moment Flow table flow entry scale;δ is the item number that the corresponding flow entry scale queue of i-th of interchanger saves flow entry scale record Maximum value;tnFor receive i-th of interchanger nth STATS_REPLY message time.
4. method according to claim 1, which is characterized in that the current t of determinationnFlow entry is flat in flow table described in moment Equal failure rate includes: to obtain tnMoment, unit time tn-tn-1I-th of interchanger flow table in flow entry failure speed RateAnd current t is determined according to following formulanThe average failure rate of flow entry in flow table described in moment:
As n≤δ,
As n > δ,
Wherein,For tnThe average failure rate of flow entry in the flow table of i-th of interchanger of moment;For tnMoment, Unit time is tn-tn-1I-th of interchanger flow table in flow entry failure rate;tnFor receive i-th of interchanger n-th The time of STATS_REPLY message.
5. method according to claim 4, which is characterized in that the acquisition tnMoment, unit time tn-tn-1I-th The rate that flow entry fails in the flow table of interchangerLater, the method further includes: the flow entry is lost The rate of effectIt saves to the corresponding flow entry failure rate queue F of i-th of interchangeri, as queue FiThe stream of middle preservation The item number of list item failure rate record is more than the FiMaximum value when, according to FIFO delete FiIn earliest data record.
6. method according to claim 1, which is characterized in that the current t of determinationnMoment flow entry advance the speed estimate Evaluation includes: to obtain tnMoment, unit time tn-tn-1I-th of interchanger flow table in the increased rate of flow entryAnd determine that being averaged for flow entry is advanced the speed in the flow table according to following formula:
As n≤δ,
As n > δ,
Wherein,For tnBeing averaged for flow entry is advanced the speed in the flow table of i-th of interchanger of moment;For tnIt is moment, single The position time is tn-tn-1I-th of interchanger flow table in the increased rate of flow entry;tnFor the nth for receiving i-th of interchanger The time of STATS_REPLY message;
Then according to described inAcquisition meets Poisson distributionCumulative probabilityX minimum value Xmin, the minimum value X of the XminThe as estimated value of flow entry advanced the speed;K is Positive integer, γ ∈ (0,1).
7. method according to claim 6, which is characterized in that the acquisition tnMoment, unit time tn-tn-1I-th The increased rate of flow entry in the flow table of interchangerLater, the method further includes: will be describedIt saves It advances the speed queue G to the corresponding flow entry of i-th of interchangeri, as queue GiThe flow entry of middle preservation is advanced the speed the item of record Number is more than the GiMaximum value when, according to FIFO delete GiIn earliest data record.
8. a kind of prediction meanss of flow entry scale, which is characterized in that described device includes: to obtain module, determining module and pre- Survey module;Wherein,
The acquisition module, for obtaining current tnThe average flow entry scale of the flow table of moment interchanger;
The determining module, for determining current tnThe average failure rate of flow entry and flow entry in flow table described in moment The estimated value advanced the speed;
The prediction module, for according to the current tnAverage flow entry scale, the average failure speed of the flow entry at moment The estimated value of rate and the flow entry advanced the speed, to subsequent time tn+1When flow entry scale predicted;
The prediction module is according to the current tnThe average flow entry scale at moment, the average failure rate of the flow entry and The estimated value of the flow entry advanced the speed, to subsequent time tn+1When flow entry scale carry out prediction include:
The prediction module is according to the current tnThe average flow entry scale at moment, the average failure rate of the flow entry and The estimated value of the flow entry advanced the speed, calculates the tn+1When flow entry scale be equal to the current tnMoment puts down Equal flow entry scale and current tnThe sum for the estimated value of the flow entry at moment advanced the speed, then with current tnFlow table described in moment The difference of the average failure rate of middle flow entry.
9. device according to claim 8, which is characterized in that the acquisition module is further used for obtaining current tnMoment hands over The flow entry scale for flow table of changing planes, and the flow entry scale is saved into flow entry scale queue corresponding to the interchanger, When the item number for the flow entry scale record that the queue saves is more than the maximum value of the queue, deleted earliest according to FIFO One data record.
10. device according to claim 9, which is characterized in that the acquisition module obtains current tnThe flow table of moment interchanger Average flow entry scale to include: the acquisition module obtain current t by following formulanThe mean flow of moment interchanger flow table List item scale:
As n≤δ,
As n > δ,
Wherein,For tnThe average flow entry scale of the flow table of i-th of interchanger of moment;For tnI-th of interchanger of moment Flow table flow entry scale;δ is the item number that the corresponding flow entry scale queue of i-th of interchanger saves flow entry scale record Maximum value;tnFor receive i-th of interchanger nth STATS_REPLY message time.
11. device according to claim 8, which is characterized in that the determining module determines current tnIt is flowed in flow table described in moment The average failure rate of list item includes: that the determining module obtains tnMoment, unit time tn-tn-1I-th interchanger The rate that flow entry fails in flow tableAnd current t is determined according to following formulanFlow entry in flow table described in moment Average failure rate:
As n≤δ,
As n > δ,
Wherein,For tnThe average failure rate of flow entry in the flow table of i-th of interchanger of moment;For tnMoment, Unit time is tn-tn-1I-th of interchanger flow table in flow entry failure rate;tnFor receive i-th of interchanger n-th The time of STATS_REPLY message.
12. device according to claim 11, which is characterized in that the determining module is further used for the flow entry The rate of failureIt saves to the corresponding flow entry failure rate queue F of i-th of interchangeri, as queue FiMiddle preservation The item number of flow entry failure rate record is more than the FiMaximum value when, according to FIFO delete FiIn earliest data note Record.
13. device according to claim 8, which is characterized in that the determining module determines current tnThe increase of moment flow entry The estimated value of rate includes: that the determining module obtains tnMoment, unit time tn-tn-1I-th of interchanger flow table in The increased rate of flow entryAnd determine that being averaged for flow entry is advanced the speed in the flow table according to following formula:
As n≤δ,
As n > δ,
Wherein,For tnBeing averaged for flow entry is advanced the speed in the flow table of i-th of interchanger of moment;For tnIt is moment, single The position time is tn-tn-1I-th of interchanger flow table in the increased rate of flow entry;tnFor the nth for receiving i-th of interchanger The time of STATS_REPLY message;
Then according to described inAcquisition meets Poisson distributionCumulative probabilityX minimum value Xmin, the minimum value X of the XminThe as estimated value of flow entry advanced the speed;K is Positive integer, γ ∈ (0,1).
14. 3 described device according to claim 1, which is characterized in that the determining module, being further used for will be described It saves to the corresponding flow entry of i-th of interchanger and advances the speed queue Gi, as queue GiThe flow entry of middle preservation is advanced the speed record Item number be more than the GiMaximum value when, according to FIFO delete GiIn earliest data record.
CN201410340073.5A 2014-07-16 2014-07-16 A kind of prediction technique and device of flow entry scale Active CN105306290B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410340073.5A CN105306290B (en) 2014-07-16 2014-07-16 A kind of prediction technique and device of flow entry scale

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410340073.5A CN105306290B (en) 2014-07-16 2014-07-16 A kind of prediction technique and device of flow entry scale

Publications (2)

Publication Number Publication Date
CN105306290A CN105306290A (en) 2016-02-03
CN105306290B true CN105306290B (en) 2019-10-25

Family

ID=55203070

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410340073.5A Active CN105306290B (en) 2014-07-16 2014-07-16 A kind of prediction technique and device of flow entry scale

Country Status (1)

Country Link
CN (1) CN105306290B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108512769B (en) * 2017-02-28 2021-04-09 华为技术有限公司 Method and related equipment for sending and receiving flow table items in software defined network

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102780641A (en) * 2012-08-17 2012-11-14 北京傲天动联技术有限公司 Flow table aging method and device of quick forwarding engine, and switch
CN102868578A (en) * 2012-10-11 2013-01-09 盛科网络(苏州)有限公司 Test method and test system of openflow switch table body capacity
CN102882746A (en) * 2012-10-11 2013-01-16 盛科网络(苏州)有限公司 Measuring method and measuring system of flow table adding rate in Openflow switchboard system
US8385210B1 (en) * 2008-12-18 2013-02-26 Cisco Technology, Inc. System and method for detection and delay control in a network environment
CN103888313A (en) * 2014-03-11 2014-06-25 浙江大学 Method for forecasting optimal timeout of flow table item

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8385210B1 (en) * 2008-12-18 2013-02-26 Cisco Technology, Inc. System and method for detection and delay control in a network environment
CN102780641A (en) * 2012-08-17 2012-11-14 北京傲天动联技术有限公司 Flow table aging method and device of quick forwarding engine, and switch
CN102868578A (en) * 2012-10-11 2013-01-09 盛科网络(苏州)有限公司 Test method and test system of openflow switch table body capacity
CN102882746A (en) * 2012-10-11 2013-01-16 盛科网络(苏州)有限公司 Measuring method and measuring system of flow table adding rate in Openflow switchboard system
CN103888313A (en) * 2014-03-11 2014-06-25 浙江大学 Method for forecasting optimal timeout of flow table item

Also Published As

Publication number Publication date
CN105306290A (en) 2016-02-03

Similar Documents

Publication Publication Date Title
Kaul et al. Age of information: Updates with priority
US9185047B2 (en) Hierarchical profiled scheduling and shaping
CN102726009B (en) Data flow exchange control method and relative device
CN104038966A (en) Data flow scheduling method and apparatus under long term evolution network
Sidhu et al. Packet scheduling of GOOSE messages in IEC 61850 based substation intelligent electronic devices (IEDs)
Hu et al. Adjusting switching granularity of load balancing for heterogeneous datacenter traffic
CN105700940B (en) A kind of statistical multiplex method of scheduler and scheduler
CN108632863A (en) Flow method for early warning, device and server
EP3895398B1 (en) Communication system with de-jitter buffer for reducing jitter
CN104718734A (en) Congestion notification in a network
US9203759B2 (en) Switch state reporting
CN105306290B (en) A kind of prediction technique and device of flow entry scale
EP3280105A1 (en) A system and a method for a cellular communication network
Zhang et al. ATFQ: a fair and efficient packet scheduling method in multi-resource environments
CN106789743A (en) The adaptive bandwidth adjusting method and device of a kind of SDN data transmission sets
CN102684805B (en) For the method and apparatus controlling clock-signal generator
Vitale et al. An analytical approach to performance analysis of coupled processor systems
Jha et al. New Queuing Technique for Improving Computer Networks QoS
KR101681613B1 (en) Apparatus and method for scheduling resources in distributed parallel data transmission system
CN114338545A (en) Traffic shaping method and device
CN110932974A (en) Access switch equipment, control center and power data link transmission method
CN103107955B (en) Packet Transport Network array dispatching method and device
Schulz et al. End-to-end latency distribution in future mobile communication networks
Peng et al. Bit-level power-law queueing theory with applications in lte networks
Stoltidis et al. Active Queue Management in Disaggregated 5G and Beyond Cellular Networks Using Machine Learning

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
TA01 Transfer of patent application right
TA01 Transfer of patent application right

Effective date of registration: 20190923

Address after: Yuhuatai District of Nanjing City, Jiangsu province 210012 Bauhinia Road No. 68

Applicant after: Nanjing Zhongxing New Software Co., Ltd.

Address before: 518057 Nanshan District Guangdong high tech Industrial Park, South Road, science and technology, ZTE building, Ministry of Justice

Applicant before: ZTE Communications Co., Ltd.

GR01 Patent grant
GR01 Patent grant