CN106302224A - A kind of task optimization bootstrap technique under cloud service environment - Google Patents
A kind of task optimization bootstrap technique under cloud service environment Download PDFInfo
- Publication number
- CN106302224A CN106302224A CN201610899764.8A CN201610899764A CN106302224A CN 106302224 A CN106302224 A CN 106302224A CN 201610899764 A CN201610899764 A CN 201610899764A CN 106302224 A CN106302224 A CN 106302224A
- Authority
- CN
- China
- Prior art keywords
- task
- optimization
- virtual machine
- pilot model
- solution
- 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.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/12—Avoiding congestion; Recovering from congestion
- H04L47/125—Avoiding congestion; Recovering from congestion by balancing the load, e.g. traffic engineering
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/70—Admission control; Resource allocation
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The calculating scheduling of resource that the present invention is directed in cloud service system is difficult to unified coordinated management problem with transfer resource, task optimization bootstrap technique under a kind of cloud service environment is provided, set up task optimization pilot model by employing and carry out the through engineering approaches process of task optimization guiding, and the solution that optimizes that through engineering approaches processes acquisition is contrasted with the optimization solution obtained by task optimization pilot model, and revise task optimization pilot model in good time, it is achieved the dynamic equilibrium planning of the calculating scheduling of resource of the cloud service under mobile environment and transfer resource.
Description
Technical field
The present invention relates to field of cloud computer technology, particularly relate to traffic scheduling, and communication network.
Background technology
The product that cloud computing is merged as several information development, it includes Distributed Calculation, parallel computation, grid meter
Calculate;Network technology includes the network storage, load balancing and Intel Virtualization Technology etc., and cloud computing is a kind of novel computation schema.Its
By technology such as virtualizations, it is achieved storage resource, the calculating IT resource such as resource and software service and bottom hardware physical equipment
Decoupling combination, it is achieved that Lower level logical basic resource, software and the integration of program application, and ensure that these resources excellent
Change.Simultaneously, it is achieved that user services on-demand dynamic adjustment, and the application model of cloud computing is as shown in Figure 1.
Under cloud environment, efficient resource management system mechanism it is advantageously ensured that reliability of service, effectiveness.One efficiently
Resource management mechanism, be first the user application demand that receives, then the service condition of monitoring node resource, then by money
Source scheduling mechanism is the virtual resource that its distribution is optimal.In one resource management system based on SLA at least includes five aspects
Hold, i.e. resource deployment, SLA attribute fault monitoring, the mapping of SLA parameter, scheduling of resource, virtual machine (vm) migration etc..
In sum: the complexity of resource, isomerism, multiformity and the dynamic of user's request under the cloud environment, because of
And user's request is all uncertain with the distribution of resource, it is therefore necessary to realize the task optimization under cloud computing environment and guide,
The efficient dispatching of lifting business.
Summary of the invention
The technical problem to be solved is: sets up task optimization pilot model by employing and carries out task optimization
The through engineering approaches guided processes, and the optimization that through engineering approaches processes acquisition is solved and the optimization obtained by task optimization pilot model
Solution contrasts, and in good time revise task optimization pilot model, it is achieved the calculating scheduling of resource of the cloud service under mobile environment with
The dynamic equilibrium planning of transfer resource.
The present invention solves that the technical scheme that above-mentioned technical problem is used comprises the following steps, as shown in Figure 2:
A, set up task optimization pilot model;
B, carry out task optimization guiding through engineering approaches process;
C, the solution that optimizes through engineering approaches process obtained are contrasted with the optimization solution obtained by task optimization pilot model,
And revise task optimization pilot model in good time.
In described step A, set up task optimization pilot model, particularly as follows: Tset={ t1,t2,...,tnFor being used for bearing
Carrying set task time of equilibrium treatment, n is number of tasks, VMjFor the virtual machine j, j ∈ [1, m] that process for task, BkzFor void
Bandwidth between plan machine k and z, k and z and j has same physical meaning, but k ≠ z, m are the virtual machine number processed for task
Mesh, yikzFor decision variable, if task i is from virtual machine VMzIt is transferred to VMkThen yikz=1, otherwise then yikz=0, xikBecome for decision-making
Amount, if task i is assigned to virtual machine VMkThen xik=1, otherwise then xik=0, DikIt is assigned to virtual machine VM for task ikProduce
Data volume, mkFor virtual machine VMzMemory size, ckFor virtual machine VMkProcessor number, DTkzFor virtual machine VMkWith VMz
Between data exchange capacity, virtual machine VMkTask process the time beThe process of cloud service system
Time isEach task is the switching time between virtual machineTask optimization draws
Guided mode type is:
In described step B, the through engineering approaches that task optimization guides processes particularly as follows: (1). judge the overload capacity of each virtual machine;
(2). select VMset={ vm1,...,vmmAs standby virtual machine set, and meet following condition, VM be not in overload condition and
VM is not in the physical machine of free time;(3). obtain at Tset={ t1,t2,...,tnShift from the VM being in overload condition in }
Set of tasks;(4) task that minimizes processes time and each task switching time between virtual machine, and concrete sub-step is, and: a. is
Each particleGenerate one and initialize population array, andThere is in n-dimensional space random site and speed;B. willContinuous position vector be reversed to dispersion vectorC. based onObtain Dik,mk,ck,Bkz, then obtain adaptation
Property function;D. for each particle, TP is calculatedk, TP and TSW;E. by the adaptability teaching value of each particleOptimal with individuality
Fitness function valueContrast, ifIt is better thanThen makeAnd make optimal location piEqual to current location
F. for each particle, calculateAnd according to ruleWithThe position of more new particle and speed,For grain
Sub-position,For particle rapidity, whereinFor global optimum, W, C1,r1,C2,r2For weight coefficient;G. space will be exportedIn optimal particle position as between virtual machine task switching Best Times;If h.It is better thanThen go to sub-step
Rapid b, on the contrary then stop calculating.
In described step C, particularly as follows: in timestamp q, through engineering approaches is processed the optimization solution that obtains and passes through task optimization
The optimization solution that pilot model obtains contrasts, if the optimization solution set of the two is completely the same, then using this optimization solution as the time
Optimization solution in stamp q;If the optimization solution set of the two is inconsistent, then through engineering approaches is processed the optimization solution obtained as timestamp q
In final optimization pass solution set, and the adjustable parameter revising the cloud computing system in task optimization pilot model makes it obtain engineering
The optimization solution set that change processes, and using revised task optimization pilot model as the task optimization of next timestamp q+1
Pilot model;The Business Processing of cloud computing system and transmission time are Q={1,2 ..., q, q+1}, by obtain in time Q
Q+1 optimizes solution set and carries out statistical average process, and this statistical average processes q+1 the optimization optimizing solution set and obtaining
Solve set and revise task optimization pilot model, and this model is processed and transmission as the task in cloud computing system next time
Priori task optimization pilot model.
Accompanying drawing explanation
The application model schematic diagram of Fig. 1 cloud computing
Fig. 2 task optimization boot flow schematic diagram
Detailed description of the invention
For reaching above-mentioned purpose, technical scheme is as follows:
The first step, sets up task optimization pilot model, particularly as follows: Tset={ t1,t2,...,tnFor being used for load balancing
Set task time processed, n is number of tasks, VMjFor the virtual machine j, j ∈ [1, m] that process for task, BkzFor virtual machine k
And the bandwidth between z, k and z and j has same physical meaning, but k ≠ z, m are the virtual machine number processed for task, yikz
For decision variable, if task i is from virtual machine VMzIt is transferred to VMkThen yikz=1, otherwise then yikz=0, xikFor decision variable, if appointing
Business i is assigned to virtual machine VMkThen xik=1, otherwise then xik=0, DikIt is assigned to virtual machine VM for task ikThe data produced
Amount, mkFor virtual machine VMzMemory size, ckFor virtual machine VMkProcessor number, DTkzFor virtual machine VMkWith VMzBetween
Data exchange capacity, virtual machine VMkTask process the time beThe process time of cloud service system isEach task is the switching time between virtual machineTask optimization pilot model
For:
Second step, the through engineering approaches carrying out task optimization guiding processes, and concretely comprises the following steps: the through engineering approaches that task optimization guides processes
Particularly as follows: (1). judge the overload capacity of each virtual machine;(2). select VMset={ vm1,...,vmmAs standby virtual machine set,
And meeting following condition, VM is not in overload condition and VM is not in the physical machine of free time;(3). obtain at Tset={ t1,t2,...,
tnCarry out the set of tasks shifted from the VM being in overload condition in };(4) task that minimizes processes time and each task at virtual machine
Between switching time, concrete sub-step is: a. is each particleGenerate one and initialize population array, andTie up at n
Space has random site and speed;B. willContinuous position vector be reversed to dispersion vectorC. based on
Obtain Dik,mk,ck,Bkz, then obtain fitness function;D. for each particle, TP is calculatedk, TP and TSW;E. by each grain
The adaptability teaching value of sonWith individual optimum adaptation functional valueContrast, ifIt is better thanThen makeAnd order
Optimal location piEqual to current locationF. for each particle, calculateAnd according to
RuleWithMore
The position of new particle and speed,For particle position,For particle rapidity, whereinFor global optimum, W, C1,
r1,C2,r2For weight coefficient;G. space will be exportedIn optimal particle position as between virtual machine task switching optimal
Time;If h.It is better thanThen go to sub-step b, otherwise then stop calculating.
3rd step, is processed through engineering approaches the solution that optimizes obtained and carries out with the optimization solution obtained by task optimization pilot model
Contrast, and in good time revise task optimization pilot model, particularly as follows: in timestamp q, through engineering approaches is processed the optimization solution that obtains with
The optimization solution obtained by task optimization pilot model is contrasted, if the optimization solution set of the two is completely the same, then this is excellent
Dissolve as the optimization solution in timestamp q;If the optimization solution set of the two is inconsistent, then through engineering approaches is processed the optimization solution obtained
As the final optimization pass solution set in timestamp q, and revise the adjustable parameter of cloud computing system in task optimization pilot model
It is made to obtain the optimization solution set that through engineering approaches processes, and using revised task optimization pilot model as next timestamp q+
The task optimization pilot model of 1;The Business Processing of cloud computing system and transmission time are Q={1,2 ..., q, q+1}, will time
Between q+1 optimization obtaining in Q solve set and carry out statistical average process, and process this statistical average to optimize and solve set and acquisition
Q+1 optimization solve set and revise task optimization pilot model, and using this model as the task in cloud computing system next time
Process and the priori task optimization pilot model of transmission.
The present invention proposes the task optimization bootstrap technique under a kind of cloud service environment, sets up task optimization by employing and draws
Guided mode type and the through engineering approaches carrying out task optimization guiding process, and it is excellent with by task that through engineering approaches is processed the optimization solution obtained
The optimization solution changing pilot model acquisition contrasts, and revises task optimization pilot model in good time, it is achieved the cloud under mobile environment
The calculating scheduling of resource of service is planned with the dynamic equilibrium of transfer resource.
Claims (4)
1. the task optimization bootstrap technique under cloud service environment, by setting up task optimization pilot model and to carry out task excellent
Change the through engineering approaches guided to process, it is achieved the cloud service scheduling of resource under mobile environment and the planning energy of the dynamic equilibrium between transfer resource
Power, comprises the steps:
A, set up task optimization pilot model;
B, carry out task optimization guiding through engineering approaches process;
C, the solution that optimizes through engineering approaches process obtained are contrasted with the optimization solution obtained by task optimization pilot model, and fit
Shi Xiuzheng task optimization pilot model.
Method the most according to claim 1, is characterized in that for described step A: particularly as follows: Tset={ t1,t2,...,tnIs
Being used for set task time of load balance process, n is number of tasks, VMjFor process for task virtual machine j, j ∈ [1,
M], BkzFor the bandwidth between virtual machine k and z, k and z and j has same physical meaning, but k ≠ z, m are to process for task
Virtual machine number, yikzFor decision variable, if task i is from virtual machine VMzIt is transferred to VMkThen yikz=1, otherwise then yikz=0, xik
For decision variable, if task i is assigned to virtual machine VMkThen xik=1, otherwise then xik=0, DikIt is assigned to virtual for task i
Machine VMkThe data volume produced, mkFor virtual machine VMzMemory size, ckFor virtual machine VMkProcessor number, DTkzFor virtual
Machine VMkWith VMzBetween data exchange capacity, virtual machine VMkTask process the time beCloud service system
System the process time beEach task is the switching time between virtual machineAppoint
Business optimizes pilot model:
。
Method the most according to claim 1, is characterized in that for described step B: the through engineering approaches that task optimization guides processes concrete
For: (1). judge the overload capacity of each virtual machine;(2). select VMset={ vm1,...,vmmAs standby virtual machine set, and meet
Following condition, VM is not in overload condition and VM is not in the physical machine of free time;(3). obtain at Tset={ t1,t2,...,tn}
The interior set of tasks carrying out from the VM being in overload condition shifting;(4) task that minimizes processes time and each task between virtual machine
Switching time, concrete sub-step is: a. is each particleGenerate one and initialize population array, andTie up at n
Space has random site and speed;B. willContinuous position vector be reversed to dispersion vectorC. based on
Obtain Dik,mk,ck,Bkz, then obtain fitness function;D. for each particle, TP is calculatedk, TP and TSW;E. by each grain
The adaptability teaching value of sonWith individual optimum adaptation functional valueContrast, ifIt is better thanThen makeAnd order
Optimal location piEqual to current locationF. for each particle, calculateAnd according to
RuleWithMore
The position of new particle and speed,For particle position,For particle rapidity, whereinFor global optimum, W, C1,
r1,C2,r2For weight coefficient;G. space will be exportedIn optimal particle position as task switching between virtual machine
The good time;If h.It is better thanThen go to sub-step b, otherwise then stop calculating.
Method the most according to claim 1, is characterized in that for described step C: particularly as follows: in timestamp q, by through engineering approaches
Process the solution that optimizes obtained to contrast with the optimization solution obtained by task optimization pilot model, if the optimization solution set of the two
Completely the same, then using this optimization solution as the optimization solution in timestamp q;If the optimization solution set of the two is inconsistent, then by engineering
The optimization solution that change process obtains is as the final optimization pass solution set in timestamp q, and revises the cloud in task optimization pilot model
The adjustable parameter of calculating system makes it obtain the optimization solution set that through engineering approaches processes, and by revised task optimization pilot model
Task optimization pilot model as next timestamp q+1;The Business Processing of cloud computing system and transmission time are Q={1,
2 ..., q, q+1}, q+1 obtained in time Q optimization is solved set and carries out statistical average process, and by this statistical average
Process q+1 the optimization solution set optimizing solution set and obtain and revise task optimization pilot model, and using this model as next
Task in secondary cloud computing system processes and the priori task optimization pilot model of transmission.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610899764.8A CN106302224A (en) | 2016-10-15 | 2016-10-15 | A kind of task optimization bootstrap technique under cloud service environment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610899764.8A CN106302224A (en) | 2016-10-15 | 2016-10-15 | A kind of task optimization bootstrap technique under cloud service environment |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106302224A true CN106302224A (en) | 2017-01-04 |
Family
ID=57718114
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610899764.8A Pending CN106302224A (en) | 2016-10-15 | 2016-10-15 | A kind of task optimization bootstrap technique under cloud service environment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106302224A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112764879A (en) * | 2021-01-14 | 2021-05-07 | 深圳市科思科技股份有限公司 | Load balancing method, electronic device and computer readable storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130218618A1 (en) * | 2010-03-24 | 2013-08-22 | International Business Machines Corporation | Dynamically optimized distribuited cloud computing-based business process management (bpm) system |
CN104331321A (en) * | 2014-10-09 | 2015-02-04 | 冷明 | Cloud computing task scheduling method based on tabu search and load balancing |
CN104516785A (en) * | 2014-12-19 | 2015-04-15 | 上海电机学院 | Cloud computing resource scheduling system and method |
CN105005501A (en) * | 2015-07-02 | 2015-10-28 | 中国地质大学(武汉) | Cloud data center oriented second-order task optimization scheduling method |
-
2016
- 2016-10-15 CN CN201610899764.8A patent/CN106302224A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130218618A1 (en) * | 2010-03-24 | 2013-08-22 | International Business Machines Corporation | Dynamically optimized distribuited cloud computing-based business process management (bpm) system |
CN104331321A (en) * | 2014-10-09 | 2015-02-04 | 冷明 | Cloud computing task scheduling method based on tabu search and load balancing |
CN104516785A (en) * | 2014-12-19 | 2015-04-15 | 上海电机学院 | Cloud computing resource scheduling system and method |
CN105005501A (en) * | 2015-07-02 | 2015-10-28 | 中国地质大学(武汉) | Cloud data center oriented second-order task optimization scheduling method |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112764879A (en) * | 2021-01-14 | 2021-05-07 | 深圳市科思科技股份有限公司 | Load balancing method, electronic device and computer readable storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Cheng et al. | Network function virtualization in dynamic networks: A stochastic perspective | |
Chekired et al. | Industrial IoT data scheduling based on hierarchical fog computing: A key for enabling smart factory | |
Rahbari et al. | Task offloading in mobile fog computing by classification and regression tree | |
Aujla et al. | MEnSuS: An efficient scheme for energy management with sustainability of cloud data centers in edge–cloud environment | |
CN112153700B (en) | Network slice resource management method and equipment | |
Etemadi et al. | A cost-efficient auto-scaling mechanism for IoT applications in fog computing environment: a deep learning-based approach | |
CN110401964A (en) | A kind of user oriented is Poewr control method of the center network based on deep learning | |
CN104468688A (en) | Method and apparatus for network virtualization | |
Mechalikh et al. | PureEdgeSim: A simulation framework for performance evaluation of cloud, edge and mist computing environments | |
CN104104621B (en) | A kind of virtual network resource dynamic self-adapting adjusting method based on Nonlinear Dimension Reduction | |
Kim et al. | Modems: Optimizing edge computing migrations for user mobility | |
Hussain et al. | Fog computing for big data analytics in IoT aided smart grid networks | |
CN109446385A (en) | A kind of method of equipment map that establishing Internet resources and the application method of the equipment map | |
Li et al. | Research on QoS service composition based on coevolutionary genetic algorithm | |
Bouzidi et al. | Dynamic clustering of software defined network switches and controller placement using deep reinforcement learning | |
Li et al. | Data analytics for fog computing by distributed online learning with asynchronous update | |
Ryabko et al. | Graph model of Fog Computing system | |
Zhang | Storage optimization algorithm design of cloud computing edge node based on artificial intelligence technology | |
Ryabko et al. | Graph theory methods for fog computing: A pseudo-random task graph model for evaluating mobile cloud, fog and edge computing systems | |
Saravanan et al. | Improving map reduce task scheduling and micro-partitioning mechanism for mobile cloud multimedia services | |
Zhang et al. | Offloading demand prediction-driven latency-aware resource reservation in edge networks | |
Zhang et al. | A data stream prediction strategy for elastic stream computing systems | |
CN106302224A (en) | A kind of task optimization bootstrap technique under cloud service environment | |
Subrahmanyam et al. | Optimizing horizontal scalability in cloud computing using simulated annealing for Internet of Things | |
US9503367B2 (en) | Risk mitigation in data center networks using virtual machine sharing |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170104 |
|
WD01 | Invention patent application deemed withdrawn after publication |