CN106856459A - A kind of method for dispatching message and device - Google Patents
A kind of method for dispatching message and device Download PDFInfo
- Publication number
- CN106856459A CN106856459A CN201510897368.7A CN201510897368A CN106856459A CN 106856459 A CN106856459 A CN 106856459A CN 201510897368 A CN201510897368 A CN 201510897368A CN 106856459 A CN106856459 A CN 106856459A
- Authority
- CN
- China
- Prior art keywords
- equipment
- core
- credit
- value
- cnt
- 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
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/50—Queue scheduling
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/50—Queue scheduling
- H04L47/62—Queue scheduling characterised by scheduling criteria
- H04L47/622—Queue service order
Abstract
The invention discloses a kind of method for dispatching message and device, the device includes:Credit value acquiring unit is updated, for for each equipment, when sending current message, obtaining the credit value of renewal of the equipment;Remaining credit value determining unit, the remaining credit value for determining all devices;Dispatching message unit, for remaining credit value and the credit value of renewal of the equipment according to all devices, determines whether equipment sends current message.By the present invention, each equipment individually carries out the scheduling of message, can give full play of the performance of each equipment, and the efficiency and success rate of the message transmission of raising improve the performance of system, while not increasing extra hardware device, reduces cost.
Description
Technical field
The present invention relates to communication network areas, more particularly to a kind of method for dispatching message and device.
Background technology
The current polycaryon processor extensive use on web-transporting device, the hardware performance of polycaryon processor
It is that, as being added to for core number is linearly increasing, can the performance and software that give full play of multinuclear hardware set
Meter method has direct relation.
In prevailing system, the characteristics of due to data stream scheduling, data stream scheduling is all according to polycaryon processor
Core serially disposed, each core divides different functions on processor, and pipeline system completes data and adjusts
Degree function.Fig. 1 is the schematic diagram of polycaryon processor conventional serial treatment in the prior art, as shown in figure 1,
By the most basic function of network processes " Packet reassembling ", " reception flow restriction ", " message decryption ", " message
Encryption ", " message fragment ", are deployed on the different core of polycaryon processor respectively.
The benefit for serializing implementation in the prior art is extremely simple scheme, but brings asking more than comparing simultaneously
Topic.On functional perspective:1st, the differentiation based on function, polycaryon processor needs a number of Core
Number;2nd, software operation deployment scheme needs to redesign after another processor is changed into.At performance angle
On degree:If the 1, " message fragment " is relatively more, then Core0 and Core6 will be extremely busy, and
Other Core may be different degrees of it is idle, cause system bottleneck point to appear in Core0 and Core6
On, make whole system performance undesirable;If the 2, MTU of the message for the receiving and MTU for sending message
Length it is different, then will appear from again " Core0 is busier, or Core6 is busier ",
Cause system bottleneck point either to also result in overall performance on Core0 or on Core6, equally to pay no attention to
Think;If the 3, message needs the comparing of decryption many, then the bottleneck point that Core1 is system occurs again,
Make whole system performance undesirable, if same encrypted message is relatively more, then Core5 occurs again to be
The bottleneck point of system, makes whole system performance undesirable;4th, when the number of queues used in system is more
Wait, occur that Core3 turns into the bottleneck point of system again, influence the performance boost of whole system, similarly, when
When the number of queues used in system is few, Core3 load reductions, but the load of Core4 relative can be improved,
New bottleneck point is caused to appear in Core4.
To sum up analyze, it can be seen that multiple core serial connections, in message transmission process, in various situations
The different bottleneck point of lower appearance, it is impossible to give full play of the performance of polycaryon processor, causes report
The success rate that text sends is relatively low, reduces the performance of system.
The content of the invention
In view of multiple core serial connections in the prior art, in message transmission process, occur in all cases
Different bottleneck points, it is impossible to give full play of the performance of polycaryon processor, cause message to send
Success rate it is relatively low, reduce the problem of the performance of system, it is proposed that the present invention is to provide on one kind overcomes
State problem or the method for dispatching message and device that solve the above problems at least in part.
The present invention provides a kind of message scheduling device, is scheduled for the message to multiple equipment, the device
Including:
Credit value acquiring unit is updated, for sending during current message, the credit value of the renewal of the equipment is obtained,
Wherein, the equipment current message to be sent;
Remaining credit value determining unit, the remaining credit value for determining all devices;
Dispatching message unit, for the remaining credit value according to all devices and the credit value of the renewal of the equipment,
Determine whether the equipment sends current message.
Wherein, dispatching message unit includes:
Addition module, for the remaining credit value of all devices to be added with the credit value of the renewal of the equipment, makees
It is the current credit of the equipment;
Judge module, for judging current credit whether not less than predetermined threshold;
Sending module, for being yes according to the result of judge module, then sends current message.
Preferably, predetermined threshold is not less than zero.
Wherein, remaining credit value determining unit includes:
Acquisition module, for obtaining the remaining credit value of each equipment;
Determining module, for the remaining credit value of each equipment to be added, as the remaining credit of all devices
Value.
Preferably, credit value Core [i] .New_Credit for updating is determined using following formula:
Core [i] .New_Credit={ (Core [i] .New_Cnt+Core [i] .Cnt_Lost_RW)
* T_Credit_R*Core [i] .Weight_R/SUM (Core [i] .Weight_R) }/T_Cnt_R, wherein,
Core [i] .New_Credi represent during Core [i] .New_Cnt should increased credit value,
Core [i] .New_Cnt represents Counter Values of the equipment i in the dispatching message time twice,
Core [i] .Cnt_Lost_RW represents the current last credit values that calculate of equipment i because integer arithmetic loses
Counter Value, Core [i] .Weight_R represents that the credit of equipment i updates weight, and T_Cnt_R represents unit
Time T inside counting device value changes amount, T_Credit_R represents credit value variable quantity in the unit interval, and i is represented
The sequence number of equipment.
Preferably, the Counter Value that this credit value integer arithmetic loses is determined using following formula
Core[i].Cnt_Lost_New:
Core [i] .Cnt_Lost_New=
(Core [i] .New_Cnt+Core [i] .Cnt_Lost_RW) %T_Cnt_R
Wherein, Core [i] .Cnt_Lost_New represents the counting of the current credit integer arithmetic loss of equipment i
Device value, Core [i] .Cnt_Lost_RW represents that the Cnt_Lost_RW of equipment i, Core [i] .New_Cnt are represented
The Counter Value of dispatching message time twice, T_Cnt_R represents unit interval T inside counting device value changes amount.
Wherein, device also includes:
Parallel processing element, for the multiple equipment to be connected in parallel.
Wherein, device also includes:
First parameter processing unit, for placing the first parameter, the first parameter is shared and non-multiple equipment
The parameter of real-time change.
Preferably, the first parameter includes at least one of:Unit interval inside counting device value changes amount, unit
Credit value variable quantity in time, credit max value.
Preferably, the first parameter processing unit is arranged in internal memory, is alignd with cacheline.
Also, the first parameter is read-only parameter.
Wherein, device also includes:
Multiple second parameter processing units, each equipment one the second parameter processing unit of correspondence;
Second parameter processing unit, for placing the second parameter, the second parameter is total and real multiple equipment
The parameter of Shi Bianhua.
Preferably, the second parameter includes at least one of:Last message sends the Counter Value on opportunity,
Last time calculates the Counter Value that credit value is lost due to integer arithmetic, and the last time is sent completely rear credit value
Residue has value of symbol.
Preferably, the second parameter processing unit is arranged in internal memory, is alignd with cacheline.
Also, the second parameter is read-write parameter.
Wherein, the same server of the Data flow direction of multiple equipment
Wherein, equipment is one below:Core, processor, veneer.
Present invention also offers a kind of method for dispatching message, including:
When sending current message, the credit value of renewal of the equipment is obtained;
Determine the remaining credit value of all devices;
The credit value of remaining credit value and the renewal of the equipment according to all devices, determines whether the equipment sends
Current message.
Wherein, the remaining credit value according to all devices and the credit value of the renewal of the equipment, determine that the equipment is
The no treatment for sending current message includes:
The remaining credit value of all devices is added with the credit value of the renewal of the equipment, as the current of the equipment
Credit value;
Judge current credit whether not less than predetermined threshold;
If it is judged that being yes, then current message is sent.
Preferably, predetermined threshold is not less than zero.
The present invention has the beneficial effect that:
By means of the terminal of the embodiment of the present invention, multiple core serial connections in the prior art are solved, in message
In transmission process, the different bottleneck point for occurring in all cases, it is impossible to make the property of polycaryon processor
Can give full play of, the success rate for causing message to send is relatively low, reduces the problem of the performance of system, this
In invention, each equipment individually carries out the scheduling of message, can give full play of the performance of each equipment, carries
Efficiency and success rate that message high sends, improve the performance of system, while not increasing extra hardware
Equipment, reduces cost.
Described above is only the general introduction of technical solution of the present invention, in order to better understand technology of the invention
Means, and being practiced according to the content of specification, and in order to allow above and other objects of the present invention,
Feature and advantage can become apparent, below especially exemplified by specific embodiment of the invention.
Brief description of the drawings
By reading the detailed description of hereafter preferred embodiment, various other advantages and benefit are for ability
Domain those of ordinary skill will be clear understanding.Accompanying drawing is only used for showing the purpose of preferred embodiment, and simultaneously
It is not considered as limitation of the present invention.And in whole accompanying drawing, identical is denoted by the same reference numerals
Part.In the accompanying drawings:
Fig. 1 is the structural representation of the message scheduling device of the embodiment of the present invention;
Fig. 2 is the detailed construction schematic diagram of the message scheduling device of the embodiment of the present invention;
Fig. 3 is the schematic diagram of the parameter of the embodiment of the present invention first and the second parameter
Fig. 4 is the flow chart of the method for dispatching message of the embodiment of the present invention;
Fig. 5 is the implementation method schematic diagram of embodiment of the present invention dispatching message in multiple equipment;
Fig. 6 be the embodiment of the present invention polycaryon processor in dispatching message implementation method schematic diagram;
Fig. 7 be the embodiment of the present invention multiprocessor veneer in dispatching message implementation method schematic diagram;
Fig. 8 be the embodiment of the present invention multiple class blade veneers in dispatching message implementation method schematic diagram;
Fig. 9 be the embodiment of the present invention multiple network packaged equipments in dispatching message implementation method schematic diagram.
Specific embodiment
The exemplary embodiment of the disclosure is more fully described below with reference to accompanying drawings.Although being shown in accompanying drawing
The exemplary embodiment of the disclosure, it being understood, however, that may be realized in various forms the disclosure without should be by
Embodiments set forth here is limited.Conversely, there is provided these embodiments are able to be best understood from this
It is open, and can by the scope of the present disclosure it is complete convey to those skilled in the art.
Below in conjunction with accompanying drawing and embodiment, the present invention will be described in further detail.It should be appreciated that this
The described specific embodiment in place is only used to explain the present invention, does not limit of the invention.Cannot give full play to
The multinuclear performance of multi-core CPU.
Device embodiment
Embodiments in accordance with the present invention, there is provided a kind of message scheduling device, Fig. 1 is the embodiment of the present invention
The structural representation of message scheduling device, as shown in figure 1, message scheduling device according to embodiments of the present invention
Including:Credit value acquiring unit 12, remaining credit value determining unit 14 and dispatching message unit 16 are updated,
The modules to the embodiment of the present invention are described in detail below.
Credit value acquiring unit 12 is updated, for sending during current message, the credit value of renewal of the equipment is obtained,
Wherein, the equipment current message to be sent;
Remaining credit value determining unit 14, the remaining credit value for determining all devices;
Dispatching message unit 16, for the remaining credit value according to all devices and the credit of the renewal of the equipment
Value, determines whether the equipment sends current message.
Fig. 2 is the detailed construction schematic diagram of the message scheduling device of the embodiment of the present invention, as shown in Fig. 2 report
Literary scheduling unit 16 includes addition module 160, judge module 162 and sending module 164, remaining credit value
Determining unit 14 includes the first acquisition module 140, the second acquisition module 142 and determining module 144, wherein,
Dispatching message unit 16 includes:
Addition module 160, for the remaining credit value of all devices to be done and the credit with the renewal of the equipment
Value is added, used as the current credit of the equipment;
Judge module 162, for judging current credit whether not less than predetermined threshold, it is preferable that predetermined
Threshold value is not less than zero, for example, the threshold value can select the integers such as 0,1,2,3, it is also possible to select decimal,
For example, 2.1,3.5 etc., the threshold value can rule of thumb be chosen, it is also possible to sent according to message in practice and imitated
Rate highest statistical value is chosen.
Sending module 164, for being yes according to the result of judge module, then sends current message.
Remaining credit value determines that Unit 14 include:
Acquisition module 140, obtains the remaining credit value of each equipment;
Determining module 142, for the remaining credit value of each equipment to be added, as the residue of all devices
Credit value.
Message scheduling device and multiple equipment (equipment 1, equipment 2 ..., equipment N) annexation
Also as shown in Fig. 2 wherein, multiple equipment (equipment 1, equipment 2 ..., equipment N) data flow
Same server is flowed to, for example, equipment can be core, processor, veneer etc..
The embodiment of the present invention provides a kind of credit value calculating method of renewal, it is possible to use following formula determine more
New credit value Core [i] .New_Credit:
Core [i] .New_Credit={ (Core [i] .New_Cnt+Core [i] .Cnt_Lost_RW)
* T_Credit_R*Core [i] .Weight_R/SUM (Core [i] .Weight_R) }/T_Cnt_R, wherein,
Core [i] .New_Credi represent during Core [i] .New_Cnt should increased credit value,
Core [i] .New_Cnt represents Counter Values of the equipment i in the dispatching message time twice,
Core [i] .Cnt_Lost_RW represents that the last of equipment i calculates the meter that credit value is lost due to integer arithmetic
Number device value, Core [i] .Weight_R represents that the credit of equipment i updates weight, and T_Cnt_R represents the unit interval
T inside counting device value changes amounts, T_Credit_R represents credit value variable quantity in the unit interval, and i represents equipment
Sequence number.
The embodiment of the present invention also provides a kind of Counter Value computational methods of credit value integer arithmetic loss, can be with
Determine the Counter Value that this credit value integer arithmetic loses using following formula
Core[i].Cnt_Lost_New:
Core [i] .Cnt_Lost_New=
(Core [i] .New_Cnt+Core [i] .Cnt_Lost_RW) %T_Cnt_R
Wherein, Core [i] .Cnt_Lost_New represents the counting of this credit value integer arithmetic loss of equipment i
Device value, Core [i] .Cnt_Lost_RW represents that the Cnt_Lost_RW of equipment i, Core [i] .New_Cnt are represented
The Counter Value of dispatching message time twice, T_Cnt_R represents unit interval T inside counting device value changes amount.
The message scheduling device also includes parallel processing element, for multiple equipment to be connected in parallel.
The message scheduling device also includes the first parameter processing unit and multiple second parameter processing units:
First parameter processing unit, for placing the first parameter, the first parameter is shared and non-multiple equipment
The read-only parameter of real-time change, for example, the first parameter includes at least one of:Unit interval inside counting device
Value changes amount, credit value variable quantity in the unit interval, credit max value.First parameter processing unit can set
Put in internal memory, alignd with cacheline.
Multiple second parameter processing units, each equipment one the second parameter processing unit of correspondence, the second parameter
Processing unit is used to place the second parameter, and the second parameter is that multiple equipment has and the read-write of real-time change is joined
Number, for example, the second parameter includes at least one of:Last message sends the Counter Value on opportunity, on
The Counter Value that credit value is lost due to integer arithmetic is once calculated, the last time is sent completely rear credit value and remains
It is remaining to have value of symbol.Second parameter processing unit is arranged in internal memory, is alignd with cacheline.
Because the first parameter processing unit and multiple second parameter processing units are arranged in internal memory, here not
Draw again signal, the internal memory 12 is connected with N number of equipment, also with parallel processing element 10, update credit
Value acquiring unit 12, remaining credit value determining unit 14 and dispatching message unit 16 are respectively connected with.
Fig. 3 is the schematic diagram of the parameter of the embodiment of the present invention first and the second parameter, as shown in figure 3, a class/
Level characteristics parameter (the first parameter i.e. mentioned above) is shared by multiple equipment, and each equipment has oneself
Two classes/level characteristics parameter (the second parameter i.e. mentioned above), some equipment can also have three classes of oneself/
Level characteristics parameter (can also be named as the 3rd parameter), and the 3rd parameter can be arranged as required to, and such as
Three parameters can be the distinctive feature of certain equipment, rather than the total feature of all devices, and the 3rd parameter also may be used
To be arranged in internal memory, alignd with cacheline.That is, the implementation method is by one class/level characteristics
Parameter, is distributed in each equipment, preferably read for local device such/level characteristics parameter value, the portion
The target component of the shaping current limliting that multiple equipment joint/collaboration that point parameter is mainly non-volatility is completed, each equipment
Between be required to communication, for the interaction of two classes/level characteristics parameter.Two class feature parameters are the easy variable element of this equipment,
Can be:The historical data of real-time update sends information and is currently needed for the information of real-time update, and passes through
Communication interface it is updated come other processors transmission historical information and be currently needed for the information of real-time update,
And other more real-time/non-real time parameters
The present invention is connected in parallel by by multiple equipment, and each equipment individually carries out the scheduling of message, Neng Gouchong
Distribution shoots the performance of each equipment, and the efficiency and success rate of the message transmission of raising improve the property of system
Can, while not increasing extra hardware device, reduce cost.By the embodiment of the present invention, Ke Yi
The side of more preferable combined in parallel (collaboration) scheduling between polycaryon processor, multiprocessor veneer and many physical equipments
Method, is characterized in that polycaryon processor core, multiprocessor veneer or multiple physical equipments can be given full play of
Maximum performance, while the harsh functional requirement of various comparings in communication network can be adapted to preferably.
Embodiment of the method
A kind of embodiments in accordance with the present invention, there is provided multi-frequency band radio-frequency signaling method, Fig. 4 is this hair
The flow chart of the method for dispatching message of bright embodiment, as shown in figure 4, message according to embodiments of the present invention is adjusted
Degree method includes following treatment:
Step 401, when sending current message, obtains the credit value of renewal of the equipment, wherein, the equipment is pending
Send current message;
Step 402, determines the remaining credit value of all devices;
Step 403, the credit value of remaining credit value and the renewal of the equipment according to all devices, it is determined that this sets
It is standby whether to send current message.
In step 403, the remaining credit value according to all devices and the credit value of the renewal of the equipment, it is determined that
The treatment whether equipment sends current message includes:The remaining credit value of all devices is done and is set with this
The standby credit value for updating is added, used as the current credit of the equipment;Judge whether current credit is not less than
Predetermined threshold;If it is judged that being yes, then current message is sent, wherein, predetermined threshold can be not less than
Zero.
Fig. 5 is the implementation method schematic diagram of embodiment of the present invention combined in parallel shaping current limliting in multiple equipment,
The embodiment of the present invention is illustrated in conjunction with Fig. 5.For the sake of simplicity it is assumed that N is 3, three equipment are had,
Equipment 1, equipment 2 and equipment 3, predetermined threshold is set to 0.The length of current message 1 to be sent of equipment 1
It is 1000, the length of current message 2 to be sent of equipment 2 is 500, the current message 3 to be sent of equipment 3
Length is 100;Remaining credit value after the last scheduling of equipment 1 is 200, after the last scheduling of equipment 2
Remaining credit value be -700, the remaining credit value after the last scheduling of equipment 3 is 300;Equipment 1 updates
Credit value be 200, equipment 2 update credit value be 100, equipment 1 update credit value be 300.
The current credit of computing device 1, equipment 2 and equipment 3 respectively, judges current message 1, currently
Whether message 2, current message 3 can send:
The current credit of equipment 1 be the remaining credit value of equipment 1, equipment 2 and equipment 3 is done and, with
The credit value that equipment 1 updates is added, i.e. 200+ (- 700)+300+200=0, and predetermined threshold is 0, equipment
1 current credit is equal to 0, meets the condition not less than predetermined threshold, then length is 1000 by equipment 1
Current message 1 be sent to network side, now the remaining credit value of equipment 1 be 0-1000=-1000;
The current credit of equipment 2 be the remaining credit value of equipment 1, equipment 2 and equipment 3 is done and, with
The credit value that equipment 2 updates is added, i.e. 200+ (- 700)+300+100=-100, and predetermined threshold is 0,
The current credit of equipment 1 is equal to -100, less than 0, is unsatisfactory for the condition not less than predetermined threshold, then set
Standby 2 will be unable to the current message 2 that length is 500 is sent to network side, now to wait the renewal of credit value,
The calculating of a new round is initiated again, method is same as mentioned above, repeats no more, until current credit is more than
Equal to 0, current message 2 could be sent;
The current credit of equipment 3 be the remaining credit value of equipment 1, equipment 2 and equipment 3 is done and, with
The credit value that equipment 3 updates is added, i.e. 200+ (- 700)+300+300=100, and predetermined threshold is 0, if
Standby 3 current credit is equal to 100, meets the condition not less than predetermined threshold, then length is by equipment 3
100 current message 3 is sent to network side, and now the remaining credit value of equipment 3 is 100-100=0.
It should be noted that the credit value for updating is calculated can use above-mentioned formula Core [i] .New_Credit:
Core [i] .New_Credit={ (Core [i] .New_Cnt+Core [i] .Cnt_Lost_RW)
* T_Credit_R*Core [i] .Weight_R/SUM (Core [i] .Weight_R) }/T_Cnt_R, by obtaining
Corresponding first parameter, the second parameter come determine update credit value.
As can be seen that in the prior art, being connected in series between equipment, each equipment only possesses message transmission
Partial function, for example, the function of equipment 1 be " Packet reassembling ", equipment 2 function be " receiving stream
Amount limitation ", the function of equipment 3 are " message decryption ", the function of equipment 4 is " message encryption ", equipment 5
Function be " message fragment ", the scheduling of each message, each equipment carries out different treatment, but at this
In inventive embodiments, it is connected in parallel between equipment, each equipment possesses " Packet reassembling ", " receives flow
Limitation ", " message decryption ", functional, the tune of each message of " message encryption " and " message fragment "
Degree, it is only necessary to which one of equipment is processed, for example, equipment 1 is responsible for the scheduling of message 1, equipment 2
It is responsible for the scheduling of message 2, that is to say, that the scheduling of message 1, equipment 2 or other equipment are not involved in,
Only equipment 1 processes the scheduling of message 1, and the scheduling of message 2, equipment 1 or other equipment are not involved in,
Only equipment 2 processes the scheduling of message 2, and other are similarly.
The technical scheme to the embodiment of the present invention is described in further detail below in conjunction with the accompanying drawings, and the present invention is real
Apply example offer on the communication network being made up of polycaryon processor, multiprocessor veneer or multiple physical equipment
Parallel distributed joint (collaboration) shaping current limliting dispatching method.It should be noted that for convenience of description originally
Inventive embodiments, sets forth in polycaryon processor in the embodiment of the present invention, multiprocessor veneer and multiple
Using the function of the method shaping current limliting in network transmission on physical equipment, but this is not limited to, meets many
The equipment of the same server feature of Data flow direction of individual equipment is still within the scope of the present invention.
The dispatching message of the polycaryon processor of embodiment 1
One extraordinary feature of polycaryon processor is exactly, if necessary to performance higher, then can select
The more processors of check figure can just reach the purpose, but on condition that software operation can make full use of this
Point, as the various limitations of serial approach mentioned in the prior art, if so software being capable of complete parallel
And increase/the reduction of performance can be reached by the core number of addition/reduction polycaryon processor, then for
Decisive role is played in giving full play to for multinuclear performance, example as mentioned in the background art, in network processes
Middle parallelization implement in relatively it is difficult be core3 and core4 business because the partial service is related to close
The real-time update of bond parameter (critical resource), and under multinuclear multiple unconfined operation same internal memories of cores
Key parameter (critical resource) will occur that hydraulic performance decline is serious and the inconsistent phenomenon of data will occur,
Ultimately result in data stream scheduling failure.The implementation method that this method is provided can be very good to solve the problem, make
The performance of device for multi-core gives full play to, and the basic linear increase of number with core, while tune will not be produced
Spend the problem of failure.
Fig. 6 is the dispatching message of the polycaryon processor of the embodiment of the present invention, i.e., join parallel in polycaryon processor
The application of (collaboration) shaping current limliting is closed, as shown in fig. 6, scheduling parameter is categorized into the shared parameter of multinuclear (i.e.
First parameter mentioned above) and the exclusive parameter of each core (the second parameter i.e. mentioned above or the 3rd ginseng
Number), the shared parameter of multinuclear has non-volatility characteristic, and the exclusive parameter of each core has real-time variable characteristic, and
The information in the exclusive parameter of other cores is added in scheduling judgement by doing add operation.For example, in the reality
Apply in scheme, " target shaping current limliting configuration parameter " is divided into " one class/level characteristics parameter ", will be " every
The data of individual core real-time update, send historical information and be currently needed for the information of real-time update " it is divided into " two
Class/level characteristics parameter ".On " three classes/level characteristics parameter ", can be designed according to specific functional requirement,
Do not illustrated in present embodiment, but be whether " three classes/level characteristics parameter " still in the patent
In protection domain.
Specifically, one class/level characteristics parameter:Target shaping current limliting configuration parameter.The partial parameters are common for multinuclear
The parameter enjoyed, the parameter is in most cases read-only parameter, is designated Para_R, can in internal memory
Only to retain portion, preferably cache line alignment, the parameter includes:Unit interval T inside counting device value
Variable quantity T_Cnt_R, credit value variable quantity T_Credit_R in the unit interval, the accumulative maximum letter in shaping port
With value Credit_Max_R;(credit value has direct corresponding relation, conventional correspondence pass with transmissible byte
System, one data of Byte of a credit value correspondence).
Two classes/level characteristics parameter:The data of each core real-time update, send historical information and be currently needed in real time
The information of renewal.The partial parameters are the exclusive parameters of each core, and compulsive requirement keeps cache line alignment,
And it is preferably that the partial parameters are to read and write parameter in real time in cache line continuous region of memorys, mark
It is Para_RW, including:Last message sends the Counter Value Cnt_RW on opportunity, and the last time calculates letter
The Counter Value Cnt_Lost_RW lost due to integer arithmetic with value, the last time is sent completely rear credit value
Residue has value of symbol Credit_Left_RW.
Credit value more new strategy:T_Cnt_R values reflect the stream of shaping current limliting with the size of T_Credit_R values
Speed, each core can update the credit value of this core according to the two parameters in the case of multinuclear, can use
State formula and describe credit value more new strategy:
Core [i] .New_Credit={ (Core [i] .New_Cnt+Core [i] .Cnt_Lost_RW)
* T_Credit_R*Core [i] .Weight_R/SUM (Core [i] .Weight_R) }/T_Cnt_R, wherein,
Core [i] .New_Credi represent during Core [i] .New_Cnt should increased credit value,
Core [i] .New_Cnt represents the Counter Value of dispatching message time twice, Core [i] .Cnt_Lost_RW tables
Show that the last of equipment i calculates the Counter Value that credit value is lost due to integer arithmetic,
Core [i] .Weight_R represents that the credit of equipment i updates weight, and T_Cnt_R represents the inside counting of unit interval T
Device value changes amount, T_Credit_R values represent credit value variable quantity in the unit interval, and i represents the sequence number of equipment.
Wherein, when all of Core [i] .Weight_R is identical, actually each core is with the speed of 1/N
Credit value is updated, when the Weight_R non-zeros of only one of which core, then only one of which core updates credit value.
For example, situation can be the same from Core [i] .Weight_R, it is possibility to have other strategies.
Each core need to know when scheduling judges that other cores are finished is remaining after last time is dispatched doing
Credit value Credit_Left_RW judges whether present port allows to send, and method is as follows:
By doing addition by other cores and the Credit_Left_RW accumulating operations of this core, all cores are obtained whole
Remaining credit value in history is dispatched on individual port, and (signed number can be positive and negative or 0), be labeled as
Credit_Left_Sum, is calculated using " credit value more new strategy " method above-mentioned
Core [i] .New_Credit, updates Credit_Left_Sum_New=Core [i] .New_Credit+
Credit_Left_Sum.Circular and dispatching method hereinbefore for example, no longer go to live in the household of one's in-laws on getting married here
State.
It is connected in parallel by by multiple cores, takes full advantage of the performance advantage of polycaryon processor, shaping queue
The limitation of number, and dynamic flexible deployment can be required according to equipment performance, improve web-transporting device
Overall performance and flexibility.
The dispatching message of the multiprocessor veneer of embodiment 2
In Wireless Communication Equipment, in order to complete the processor active task of complexity, there are multiple processors on most veneers,
For example 3 data processings of 120o cells can be simultaneously completed using 3 processors, 3 cells are as one
Individual omnidirectional 360o base stations are after common network interface needs shaping current limliting (such as according to 3 cells
300Mbit/s), it is sent in transmission network equipment.Traditional way can have two kinds:The first,
The flow for limiting each cell carries out shaping current limliting, such 3 according to the flow no more than 100Mbit/s
Cell flow integrally certainly no more than 300Mbit/s, but so occur, such as:Cell 1 has
The burst message of 120Mbit/s, cell 2 has 50Mbit/s burst messages, and cell 3 has 100Mbit/s, 3
Individual cell is added together 270Mbit/s and is also not reaching to 300Mbit/s, is just understanding this 360o, omnidirectional's base
The whole flows stood can send, but if according to current configuration, cell 1 will have 20Mbit/s
The possibility packet loss problem of flow;Another solution is extra hardware to be added, by 3 numbers of cell
Solved according to the shaping current limliting that 300Mbit/s is after being pooled on this hardware, but so need to increase hardware
Cost.
In the present invention, by will be connected in parallel between multiple processors, by the embodiment of the present invention, so that it may
Same shaping current-limiting function can be just completed with without adding additional hardware, the utilization of the network bandwidth had both been improved
Rate can reduce the cost of the network equipment again.
Fig. 7 be the embodiment of the present invention multiprocessor veneer in dispatching message implementation method schematic diagram, that is, exist
The application of combined in parallel (collaboration) shaping current limliting on multiprocessor veneer, as shown in fig. 7, in the implementation
By one class/level characteristics parameter in mode, it is distributed in the processor that each does data dispatch, for being originally located in
Reason device preferably reads and can mutually access/communicate between such parameter value, processor, for interacting two class features
Parameter.Two classes/level parameter is the data is activation historical information of the real-time update inside each processor and currently needs
Want the information of real-time update, and by communication interface it is updated come other processors transmission historical information
And it is currently needed for the information of real-time update, and other more real-time/non-real time parameters.According to data stream scheduling
Accurate situation determine two class feature parameters interaction frequent degree.
(gone up by the way that scheduling parameter is categorized into multiple processors joint (collaboration) shaping current limliting configuration parameters
The first parameter described in text) with per each and every one processor real-time mutability scheduling parameter (i.e. mentioned above second
Parameter), the interaction of the real-time mutability scheduling parameter between processor can be each processor to other lists
Plate is wide/multicast, or using reading between processor by the way of.
Circular and dispatching method have explained above, and repeat no more here.
More than 3 dispatching message of class blade veneer of embodiment
The specific embodiment is to expand to veneer/class blade from processor rank with the difference of embodiment 2
Come on veneer, same application scenarios have also extended to the equipment of similar blade service structure, such applied environment
Extend to be similar in the applied environment of blade framework.
In the present invention, multiple blade veneers are connected in parallel, by the embodiment of the present invention, can be joined parallel
(collaboration) shaping current limliting network data flow is closed, the equipment for being similar to blade server structure can be more preferable
The utilization network bandwidth reduce network equipment cost etc. simultaneously and also have many benefits.
Fig. 8 be the embodiment of the present invention multiple class blade veneers in dispatching message implementation method schematic diagram, i.e.,
The application of combined in parallel (collaboration) shaping current limliting in multiple class blade veneers, as shown in figure 8, the implementation
Be distributed in one class/level characteristics parameter on each class blade veneer by mode, more preferable for local class blade veneer
Reading such parameter value, multiple class blade veneers that the parameter is also mainly non-volatility need common completion
Can be communicated between target shaping current limliting, each class blade veneer, for interacting two classes/level characteristics parameter.Two classes
Characterisitic parameter is that the historical data of real-time update variable in real time on each veneer sends information and is currently needed for
The information of real-time update, and other more non real-time parameters in real time.Communication mode between processor can be adopted
Do not limit in any way.
Scheduling parameter is categorized into multi-blade veneer joint (collaboration) shaping current limliting configuration parameter and each blade
The real-time mutability scheduling parameter of veneer;The interaction of the real-time mutability scheduling parameter between blade veneer, Ke Yishi
Each veneer is wide to other veneers/multicast, or using reading between veneer by the way of.
Circular and dispatching method have explained above, and repeat no more here.
If having used multiprocessor or polycaryon processor in class blade veneer, example 1 and reality are may be referred to
Example 2.
By the equipment of class blade server framework, combined in parallel/collaboration is completed using the similar blade veneer of multiple
Data stream scheduling provides good solution, can be good at solving similar blade server framework and
The parallel computation application of row application.
Embodiment 4
In transmission network equipment, it is made up of countless multiple routers and gateway device, the data flow of multiple equipment
A destination server can be flowed to, because the disposal ability problem of destination server sets, it is necessary to be directed to multiple
Standby overall flow carries out the utilization server that shaping could be more efficient, and the present invention is provided in this aspect problem
Good solution.
Fig. 9 be the embodiment of the present invention multiple network packaged equipments in dispatching message implementation method schematic diagram,
The application of combined in parallel (collaboration) shaping current limliting i.e. in multiple network packaged equipments, as shown in figure 9,
In SDN equipment (SDN Switch) network, each SDN transmission equipment is required for entering line number with controller
According to stream interaction, in order to controller not by from multiple SDN transmission equipments data burst impact, it is necessary to
According to the disposal ability (ability of such as maximum treatment 500Mbit/s) of controller (SDN Controller),
Carry out combined in parallel (collaboration) shaping current limliting to come to the data of multiple SDN equipment (SDN Switch).
In the present invention, multiple SDN Switch parallel connections, by the embodiment of the present invention, can be parallel
Joint/synergistic data stream is dispatched, and improves the performance of SDN and big data network.
Circular and dispatching method have explained above, and repeat no more here.
To sum up, by the embodiment of the present invention, can be set in polycaryon processor, multiprocessor veneer and many physics
The method that preferably combined in parallel (collaboration) is dispatched between standby, is characterized in that multinuclear treatment can be given full play of
The maximum performance of device core, multiprocessor veneer or multiple physical equipments, while communication network can be adapted to preferably
The harsh functional requirement of various comparings in network.
Obviously, those skilled in the art can carry out various changes and modification without deviating from this hair to the present invention
Bright spirit and scope.So, if it is of the invention these modification and modification belong to the claims in the present invention and
Within the scope of its equivalent technologies, then the present invention is also intended to comprising these changes and modification.
Algorithm and display be not intrinsic with any certain computer, virtual system or miscellaneous equipment provided herein
It is related.Various general-purpose systems can also be used together with based on teaching in this.As described above, structure
It is obvious to make the structure required by this kind of system.Additionally, the present invention is not also directed to any certain programmed
Language.It is understood that, it is possible to use various programming languages realize the content of invention described herein, and
The description done to language-specific above is to disclose preferred forms of the invention.
In specification mentioned herein, numerous specific details are set forth.It is to be appreciated, however, that this hair
Bright embodiment can be put into practice in the case of without these details.In some instances, not in detail
Known method, structure and technology are shown, so as not to obscure the understanding of this description.
Similarly, it will be appreciated that in order to simplify the disclosure and help understand one in each inventive aspect or
Multiple, in above to the description of exemplary embodiment of the invention, each feature of the invention is sometimes by one
Rise and be grouped into single embodiment, figure or descriptions thereof.However, should not be by the method for the disclosure
It is construed to reflect following intention:I.e. the present invention for required protection requirement ratio institute in each claim is clear and definite
The more features of feature of record.More precisely, as the following claims reflect, hair
Bright aspect is all features less than single embodiment disclosed above.Therefore, it then follows specific embodiment
Claims be thus expressly incorporated in the specific embodiment, wherein each claim conduct in itself
Separate embodiments of the invention.
Those skilled in the art are appreciated that can be carried out certainly to the module in the client in embodiment
Adaptively change and they are arranged in one or more clients different from the embodiment.Can be with
The block combiner in embodiment an into module, and multiple submodule or son can be divided into addition
Unit or sub-component.At least some in except such feature and/or process or unit exclude each other it
Outward, can be using any combinations to public in this specification (including adjoint claim, summary and accompanying drawing)
The all features and all processes or unit of so disclosed any method or client opened carry out group
Close.Unless expressly stated otherwise, it is public in this specification (including adjoint claim, summary and accompanying drawing)
Each feature opened can be replaced by the alternative features for providing identical, equivalent or similar purpose.
Although additionally, it will be appreciated by those of skill in the art that some embodiments described herein include other
Some included features are rather than further feature, but the combination meaning of the feature of different embodiments in embodiment
Taste and is within the scope of the present invention and is formed different embodiments.For example, in following claim
In book, the one of any of embodiment required for protection mode can use in any combination.
All parts embodiment of the invention can realize with hardware, or with one or more processor
The software module of upper operation is realized, or is realized with combinations thereof.It will be understood by those of skill in the art that
Can be realized using microprocessor or digital signal processor (DSP) in practice according to of the invention real
Apply some or all functions of some or all parts being loaded with the client of sequence network address of example.
The present invention be also implemented as some or all equipment for performing method as described herein or
Person's program of device (for example, computer program and computer program product).It is such to realize journey of the invention
Sequence can be stored on a computer-readable medium, or can have the form of one or more signal.This
The signal of sample can be downloaded from internet website and obtained, or be provided on carrier signal, or with any
Other forms are provided.
It should be noted that above-described embodiment the present invention will be described rather than limiting the invention, and
And those skilled in the art can design replacement implementation without departing from the scope of the appended claims
Example.In the claims, any reference symbol being located between bracket should not be configured to claim
Limitation.Word "comprising" does not exclude the presence of element or step not listed in the claims.Before element
Word "a" or "an" do not exclude the presence of element as multiple.If the present invention can be by means of including
The hardware of dry different elements and realized by means of properly programmed computer.If listing equipment for drying
In unit claim, several in these devices can be embodied by same hardware branch.
The use of word first, second, and third does not indicate that any order.These words can be explained and run after fame
Claim.
Claims (22)
1. a kind of message scheduling device, is scheduled for the message to multiple equipment, it is characterised in that
Methods described includes:
Credit value acquiring unit is updated, for sending during current message, the credit value of renewal of the equipment is obtained, its
In, the equipment current message to be sent;
Remaining credit value determining unit, the remaining credit value for determining all devices;
Dispatching message unit, for remaining credit value and the letter of the renewal of the equipment according to all devices
With value, determine whether the equipment sends current message.
2. device as claimed in claim 1, it is characterised in that the dispatching message unit includes:
Addition module, for by the credit value phase of the remaining credit value of all devices and the renewal of the equipment
Plus, as the current credit of the equipment;
Judge module, for judging the current credit whether not less than predetermined threshold;
Sending module, for being yes according to the result of judge module, then sends the current message.
3. device as claimed in claim 2, it is characterised in that the predetermined threshold is not less than zero.
4. device as claimed in claim 1, it is characterised in that the remaining credit value determining unit bag
Include:
Acquisition module, for obtaining the remaining credit value of each equipment;
Determining module, for the remaining credit value of each equipment to be added, as the remaining credit of all devices
Value.
5. device as claimed in claim 1, it is characterised in that the letter for determining to update using following formula
With value Core [i] .New_Credit:
Core [i] .New_Credit={ (Core [i] .New_Cnt+Core [i] .Cnt_Lost_RW) * T_Credit_R
* Core [i] .Weight_R/SUM (Core [i] .Weight_R) }/T_Cnt_R, wherein, Core [i] .New_Credi
Represent during Core [i] .New_Cnt should increased credit value, Core [i] .New_Cnt represents equipment i
In the counter difference values of dispatching message time twice, Core [i] .Cnt_Lost_RW represents current upper of equipment i
The Counter Value that credit value is lost due to integer arithmetic is once calculated, Core [i] .Weight_R represents equipment i
Credit update weight, SUM (Core [i] .Weight_R) represents the weight summation of equipment, and T_Cnt_R represents
Unit interval T inside counting device value changes amount, T_Credit_R represents credit value variable quantity in the unit interval, i
The sequence number of expression equipment.
6. device as claimed in claim 5, it is characterised in that determine this credit using following formula
Counter Value Core [i] .Cnt_Lost_New of value integer arithmetic loss:
Core [i] .Cnt_Lost_New=(Core [i] .New_Cnt+Core [i] .Cnt_Lost_RW) %T_Cnt_R
Wherein, Core [i] .Cnt_Lost_New represents the counting of the current credit integer arithmetic loss of equipment i
Device value, Core [i] .Cnt_Lost_RW represents that the Cnt_Lost_RW of equipment i, Core [i] .New_Cnt are represented
Counter difference values of dispatching message time twice, T_Cnt_R represents unit interval T inside counting device value changes
Amount, % represents modulo operation.
7. device as claimed in claim 1, it is characterised in that described device also includes:
Parallel processing element, for the multiple equipment to be connected in parallel.
8. device as claimed in claim 1, it is characterised in that described device also includes:
First parameter processing unit, for placing the first parameter, first parameter is common for the multiple equipment
Enjoy and non real-time change parameter.
9. device as claimed in claim 8, it is characterised in that first parameter include it is following at least
One of:Unit interval inside counting device value changes amount, credit value variable quantity in the unit interval, credit max value,
The flow velocity of dispatching message, the maximum of counter difference values.
10. device as claimed in claim 8, it is characterised in that first parameter processing unit is set
In internal memory, alignd with cacheline.
11. devices as claimed in claim 8, it is characterised in that first parameter is read-only parameter.
12. devices as claimed in claim 1, it is characterised in that described device also includes:
Multiple second parameter processing units, each equipment one the second parameter processing unit of correspondence;
Second parameter processing unit, for placing the second parameter, second parameter sets for the multiple
The parameter of standby total and real-time change.
13. devices as claimed in claim 12, it is characterised in that second parameter include it is following extremely
It is one of few:Last message sends the Counter Value on opportunity, and the last time calculates credit value because integer arithmetic is damaged
The Counter Value for losing, the last time is sent completely the credit that rear credit value residue has value of symbol, consumed
Value aggregate-value.
14. devices as claimed in claim 12, it is characterised in that second parameter processing unit sets
Put in internal memory, alignd with cacheline.
15. devices as claimed in claim 12, it is characterised in that second parameter is read-write parameter.
16. device as any one of claim 1 to 15, it is characterised in that the multiple to set
The same server of standby Data flow direction.
17. device as any one of claim 1 to 15, it is characterised in that the equipment is
One below:Core, processor, veneer.
A kind of 18. message transmitting methods, it is characterised in that including:
When sending current message, the credit value of renewal of the equipment is obtained, wherein, the equipment is to be sent described to be worked as
Preceding message;
Determine the remaining credit value of all devices;
The credit value of remaining credit value and the renewal of the equipment according to all devices, it is determined that described set
It is standby whether to send current message.
19. methods as claimed in claim 18, it is characterised in that described according to all devices
Remaining credit value and the credit value of the renewal of the equipment, determine whether the equipment sends the treatment of current message
Including:
The remaining credit value of all devices is added with the credit value of the renewal of the equipment, is set as described
Standby current credit;
Judge the current credit whether not less than predetermined threshold;
If it is judged that being yes, then the current message is sent.
20. methods as claimed in claim 19, it is characterised in that the predetermined threshold is not less than zero.
21. methods as claimed in claim 18, it is characterised in that determine what is updated using following formula
Credit value Core [i] .New_Credit:
Core [i] .New_Credit={ (Core [i] .New_Cnt+Core [i] .Cnt_Lost_RW) * T_Credit_R
* Core [i] .Weight_R/SUM (Core [i] .Weight_R) }/T_Cnt_R, wherein, Core [i] .New_Credi
Represent during Core [i] .New_Cnt should increased credit value, Core [i] .New_Cnt represents and report twice
The counter difference values of literary scheduling time, Core [i] .Cnt_Lost_RW represents the current last calculating of equipment i
The Counter Value that credit value is lost due to integer arithmetic, Core [i] .Weight_R represents the credit of equipment i more
New weight, T_Cnt_R represents unit interval T inside counting device value changes amount, and T_Credit_R values represent unit
Credit value variable quantity in time, i represents the sequence number of equipment.
22. methods as claimed in claim 21, it is characterised in that determine that this is believed using following formula
Counter Value Core [i] .Cnt_Lost_New lost with value integer arithmetic:
Core [i] .Cnt_Lost_New=(Core [i] .New_Cnt+Core [i] .Cnt_Lost_RW) %T_Cnt_R
Wherein, Core [i] .Cnt_Lost_New:The counting of the current credit integer arithmetic loss of expression equipment i
Device value, Core [i] .Cnt_Lost_RW represents that the Cnt_Lost_RW of equipment i, Core [i] .New_Cnt are represented
Counter difference values of dispatching message time twice, T_Cnt_R represents unit interval T inside counting device value changes
Amount, % represents modulo operation.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510897368.7A CN106856459B (en) | 2015-12-08 | 2015-12-08 | Message scheduling method and device |
PCT/CN2016/107269 WO2017097117A1 (en) | 2015-12-08 | 2016-11-25 | Message scheduling method and apparatus |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510897368.7A CN106856459B (en) | 2015-12-08 | 2015-12-08 | Message scheduling method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106856459A true CN106856459A (en) | 2017-06-16 |
CN106856459B CN106856459B (en) | 2021-09-17 |
Family
ID=59012691
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510897368.7A Active CN106856459B (en) | 2015-12-08 | 2015-12-08 | Message scheduling method and device |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN106856459B (en) |
WO (1) | WO2017097117A1 (en) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10243866B2 (en) | 2016-07-01 | 2019-03-26 | International Business Machines Corporation | Controlling packet data transmissions via data transmission media |
CN112866079B (en) * | 2020-12-31 | 2022-10-28 | 广州技象科技有限公司 | Method, device and equipment for transmitting regular messages based on Internet of things and storage medium |
CN117014384A (en) * | 2022-04-29 | 2023-11-07 | 华为技术有限公司 | Message transmission method and message forwarding equipment |
CN114584519B (en) * | 2022-05-05 | 2022-09-16 | 飞狐信息技术(天津)有限公司 | Message middleware and current limiting method thereof |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101227410A (en) * | 2008-02-03 | 2008-07-23 | 杭州华三通信技术有限公司 | Flow monitoring method and flow monitoring equipment |
US7606154B1 (en) * | 2004-04-01 | 2009-10-20 | Juniper Networks, Inc. | Fair bandwidth allocation based on configurable service classes |
CN101834786A (en) * | 2010-04-15 | 2010-09-15 | 华为技术有限公司 | Queue scheduling method and device |
CN102340443A (en) * | 2010-07-21 | 2012-02-01 | 中兴通讯股份有限公司 | Traffic scheduling device and method for packet switched network |
-
2015
- 2015-12-08 CN CN201510897368.7A patent/CN106856459B/en active Active
-
2016
- 2016-11-25 WO PCT/CN2016/107269 patent/WO2017097117A1/en active Application Filing
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7606154B1 (en) * | 2004-04-01 | 2009-10-20 | Juniper Networks, Inc. | Fair bandwidth allocation based on configurable service classes |
CN101227410A (en) * | 2008-02-03 | 2008-07-23 | 杭州华三通信技术有限公司 | Flow monitoring method and flow monitoring equipment |
CN101834786A (en) * | 2010-04-15 | 2010-09-15 | 华为技术有限公司 | Queue scheduling method and device |
CN102340443A (en) * | 2010-07-21 | 2012-02-01 | 中兴通讯股份有限公司 | Traffic scheduling device and method for packet switched network |
Also Published As
Publication number | Publication date |
---|---|
WO2017097117A1 (en) | 2017-06-15 |
CN106856459B (en) | 2021-09-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107959636B (en) | Method and device for sending BGP (Border gateway protocol) message | |
CN106856459A (en) | A kind of method for dispatching message and device | |
CN110602156A (en) | Load balancing scheduling method and device | |
CN101217528B (en) | A link aggregation method and device | |
CN101616083A (en) | A kind of message forwarding method and device | |
CN110750341A (en) | Task scheduling method, device, system, terminal equipment and storage medium | |
US10983828B2 (en) | Method, apparatus and computer program product for scheduling dedicated processing resources | |
CN104125166B (en) | A kind of array dispatching method and computing system | |
CN101072176A (en) | Report processing method and system | |
CN103036792A (en) | Transmitting and scheduling method for maximizing minimal equity multiple data streams | |
CN113411262A (en) | Method and device for setting large receiving and unloading function | |
CN105978821B (en) | The method and device that network congestion avoids | |
CN109417533A (en) | Send the method and forwarding device of data | |
Ke et al. | Aggregation on the fly: Reducing traffic for big data in the cloud | |
CN111475250A (en) | Network optimization method and device in cloud environment | |
CN110535697B (en) | Method for balancing block chain cross-domain broadcast network load | |
CN101374109B (en) | Method and apparatus for scheduling packets | |
CN110019205A (en) | A kind of data storage, restoring method, device and computer equipment | |
US9954771B1 (en) | Packet distribution with prefetch in a parallel processing network device | |
CN110061924A (en) | A kind of message forwarding method, device and Related product | |
CN111555984B (en) | Method and device for data transmission, smart home equipment and storage medium | |
WO2019024759A1 (en) | Data communication method and data communication network | |
CN115834054B (en) | Multistage key progression management method and device | |
CN110740100A (en) | transmission rate determination method, device, network equipment and storage medium | |
CN116016375A (en) | Chip cache scheduling method and system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |