CN109948848A - Research-on-research flows down the Cost Optimization dispatching method of deadline constraint in a kind of cloud - Google Patents
Research-on-research flows down the Cost Optimization dispatching method of deadline constraint in a kind of cloud Download PDFInfo
- Publication number
- CN109948848A CN109948848A CN201910205996.2A CN201910205996A CN109948848A CN 109948848 A CN109948848 A CN 109948848A CN 201910205996 A CN201910205996 A CN 201910205996A CN 109948848 A CN109948848 A CN 109948848A
- Authority
- CN
- China
- Prior art keywords
- task
- virtual machine
- workflow
- deadline
- ant
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- Economics (AREA)
- Development Economics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Human Resources & Organizations (AREA)
- General Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Game Theory and Decision Science (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Operations Research (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Educational Administration (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention dispatches the cost issues that leased virtual machine is spent for scientific workflow in the cloud of deadline constraint, propose based on ant colony optimization system and combine the Cost Optimization dispatching algorithm R-ACS of the upward weight of task with traditional probability, it is intended to reduce the expense of workflow task lease virtual machine in entire cloud.Algorithm considers the characteristic mutually constrained between task in scientific workflow, it is ranked up using execution sequence of the upward weight of conventional probability to task, is then carried out under deadline constraint using ant colony optimization system to optimize scheduling of the lease expenses between task-resource of target.The expense leasing virtual machine and spending can be effectively reduced in method proposed by the present invention.
Description
Technical field
The invention belongs to the scientific workflow off periods in cloud computing and the big field of dispatching algorithm two more particularly to a kind of cloud
Limit the Cost Optimization dispatching method of constraint.
Background technique
Scientific workflow is the set of tasks for handling particular order, it has also become standardization and structuring complexity scientific process
Important normal form.With the continuous complication of scientific system, feature is mainly shown as data-intensive and computation-intensive,
The system environments of higher performance is needed to execute a large amount of task.Scientific workflow scheduling is used according to certain resource in cloud
Resource is distributed and managed to rule.It is a N-P difficult problem, is without determining multinomial to seek optimal solution.
For workflow schedule, extensive research is obtained in traditional local system, in cluster, network;But in traditional local
Using not only very expensive in system, but also inconvenient extended resources.And cloud computing as it is a kind of by Basis of Computer Engineering facility and
Software uses the mode for being supplied to user as needed as service, be adapted to execute scientific workflow, it can be achieved that payable at sight use and
And use is very flexible.
Cloud computing can provide the computing capability of its needs for user, that is, can lease to the unlimited resource of user.
It is worth noting that, nowadays the charging mode of business cloud computing service supplier is generally hour as minimum charge unit (such as
Amazon EC2), that is to say, that no matter we lease resource 59 minutes or one second, all with one hour for unit
Carry out collecting for lease expenses.Therefore, how in the reasonable time (deadline) with lower expense come execute task at
For a main problem of workflow in cloud.For scientific workflow scheduling in cloud, good dispatching method or strategy,
The mapping of task-resource can be efficiently completed and effectively reduce the expense of lease resource.
Summary of the invention
It is an object of the invention to: leased virtual machine, which is dispatched, for scientific workflow in the cloud of deadline constraint is spent
The cost issues taken propose the Cost Optimization dispatching method that research-on-research in a kind of cloud flows down deadline constraint, it is intended to subtract
The expense of workflow task lease virtual machine in few entire cloud.
In order to achieve the above object, the technical scheme adopted by the invention is that: research-on-research flows down the off period in a kind of cloud
Limit the Cost Optimization dispatching method of constraint, the specific steps are as follows:
Step 1: user submits demand: user submits workflow and relevant resource requirement and the cut-off of entire workflow schedule
Time limit;
Step 2: related notion being defined: including between the transmission data this paper objective function, constraint condition, task
Communication overhead, task operation starting time, the definition for terminating runing time etc.;
Step 3: tasks ranking in workflow in cloud: according to the weight of each task of the upward weight computing of conventional probability and to appoint
Business descending arrangement obtains task sequence SortedT;
Step 4: distributing sub- deadline: being task distribution son cut-off according to the resulting upward weight of task probability of step 3
Time limit;
Step 5: carrying out task schedule using ant group algorithm: son cut-off being met to sorted task using ant group algorithm
Cost minimization is set to be dispatched to virtual machine under time limit, so that final execution cost is minimum.
Further, research-on-research flows down the Cost Optimization dispatching method that deadline constrains in cloud proposed by the present invention,
The workflow that user is submitted in the step 1 is described with oriented no circulation figure G={ T, E }, and the node T of figure indicates work
Task-set T={ t in stream1,t2,...,tnShare n task, in the directed edge E expression workflow of figure between task it is mutual according to
Rely set of relations E={ ei,j|ti∈T∩tj∈ T }, wherein ei,jExpression task tiFor task tjForerunner (alternatively task tjFor
Task tiIt is subsequent), i.e., only work as task tiTask t could be executed after executionj, and the weight on directed edge indicates task
Between the data set Data={ data that transmitsi,j|ti∈T∩tj∈ T }, wherein datai,jExpression task tiWith task tjBetween transmit
Data.
Further, the present invention is based on scientific workflow Cost Optimization dispatching method, the steps in the cloud of ant group algorithm
Objective function defined in rapid 2 is to lease all virtual machine expense summations, and constraint condition is that need to meet user its deadline to determine
The deadline of justice.
s.t.WFE≤D (2)
Wherein LFlIndicate lease virtual machine vmlEnd time, LSlIndicate lease virtual machine vmlAt the beginning of, ClTable
Show and lease required cost in the virtual machine unit time, in constraint condition, WFE indicates that entire workflow deadline, D indicate user
The deadline of definition.
Further, the present invention is based on scientific workflow Cost Optimization dispatching method, the steps in the cloud of ant group algorithm
Rapid 5 be the process of an iteration using the process that ant group algorithm is scheduled sorted task, and specific iterative step is as follows:
Step 5.1: initialization with Ant colony algorithm relevant parameter: relevant parameter includes initial ant colony pheromones τ0, heuristic function
Matrix ηn×m, pheromones heuristic factor α, heuristic greedy method β, the number of iterations Itmax, ant colony quantity AntNmaxDeng;
Step 5.2: iteration starts: judging whether to have reached maximum number of iterations ItmaxIf not reaching, then follow the steps
5.3, otherwise iteration terminates, and goes to step 5.9;
Step 5.3: calculating transition probability: ant antkTask and calculating task t are successively taken from task sequence SortedTi
It distributes to virtual machine vmlProbability;
Step 5.4: selecting virtual machine: according to probability selection formula and meeting sub- deadline schedule virtual for task
Machine.
Step 5.5: judging ant antkWhether completion is searched for, and is gone to step 5.7 if search is completed, is otherwise gone to step
Rapid 5.3.
Step 5.6: local updating pheromones
Step 5.7: judge whether that all ants are complete search, goes to step 5.9 if being fully completed search, it is no
Then go to step 5.3.
Step 5.8: the overall situation updates pheromones and records local optimum scheduling scheme
Step 5.9: output global optimum's scheduling scheme.
Further, the present invention is based on scientific workflow Cost Optimization dispatching method, the steps in the cloud of ant group algorithm
In rapid 5.3, heuristic function matrix η is definedn×m, every a line of matrix represents a task, and each column represent a virtual machine,
In ηi,lValue represents task tiIt distributes to virtual machine vmlHeuristic function value.
We define the reciprocal as inspiration item, that is, task t of execution costiIn virtual machine vmlConsumed by upper execution
Expense, defined formula is as shown in formula 4, defines ECi,lFor task tiIn virtual machine vmlOn execution cost, defined formula such as formula 5
It is shown.
Transition probability formula are as follows:
Wherein, τi,lIndicate the pheromone concentration on path (i, l);α is pheromones heuristic factor, indicates ant in path
On the importance of pheromones that leaves, the bigger influence for illustrating pheromones to the selection of other ants below path is bigger;β is expectation
Heuristic factor, reflects the importance of heuristic function, β is bigger illustrate ant in moving process heuristic information by attention degree more
It is high.
Optimization Scheduling provided by the invention has the following advantages and beneficial effects: the present invention in view of scientific workflow
The characteristic mutually constrained between middle task is ranked up using execution sequence of the upward weight of conventional probability to task, is then used
Ant group algorithm carries out under deadline constraint to optimize scheduling of the lease expenses between task-resource of target, and the method can have
Imitate the expense of reduction task lease virtual machine.
Detailed description of the invention
Fig. 1 is cloud workflow schedule system model figure
Fig. 2 is the Cost Optimization dispatching method process that research-on-research flows down deadline constraint in cloud provided by the invention
Figure;
Fig. 3 is cloud workflow instance structure chart;
Specific embodiment
In order to make those skilled in the art more fully understand the technical problem in the application, technical solution and technical effect,
The Cost Optimization of deadline constraint is flowed down to research-on-research in a kind of cloud of the present invention with reference to the accompanying drawings and detailed description
Dispatching method is described in further detail.
User submits workflow and correlation to the detailed process of Optimal Scheduling model proposed by the present invention as shown in Figure 1:
Resource requirement and deadline;After Workflow Management System (WMS) receives the demand of user, automatically execute resource acquisition,
Task schedule and workflow execution.In more detail, the resource capacity estimation analysis workflow structure calculating money required with determination
Source amount, source obtaining module are held consultation with external resource configuration system, and acquisition meets required determining resource quantity, once
It is assigned with resource appropriate, executing manager will coordinate with workflow-deployment module, complete between workflow task and resource
Mapping, and guide and run task on executing manager.
If Fig. 2 is method specific implementation step of the invention:
Step 1: user submits workflow G={ T, E } and entire workflow schedule deadline D;Specific example such as Fig. 3 institute
Show, one task of each vertex representation in example, shares 13 tasks in this example;Directed line segment represents between task mutually about
The relationship of beam, such as from task t2Direction task t8Directed line segment indicate task t2For task t8Forerunner;Have in figure on directed edge
Digital representation task between the data transmitted, such as from task t2Number 3 on to the directed edge of task 8 is represented from task t2Transmission
Data volume to task 8 is unit 3.
Step 2: present invention assumes that shared m kind virtual machine VMs={ vm1,vm2,...,vmmAnd all in same data
Under the heart, then all virtual machine bandwidth (namely bw) are identical.For task tiWith task tjBetween transmit data communication open
Pin, defined formula are as follows.
Because there is the characteristic mutually constrained between task in workflow, for task tiFor, only when all predecessor tasks
Terminate to execute and transmission data have been transferred to behind this, can start to execute, task tiStarting Executing Time definition such as
Formula 8, wherein due to t1Without forerunner, then Starting Executing Time is 0.
When task tiIt distributes to virtual machine vmlWhen being executed, the time is executed by the processing capacity of virtual machine and is appointed
The size of business i determines, can be can be calculated by formula 9.Wherein ETi,lExpression task tiIt is assigned to virtual machine vmlThe execution time,
size(ti) indicate task tiData volume size, p (vml) indicate virtual machine vmlProcessing capacity.
As task tiDistribute to virtual machine vmlExecute terminate at the time of i.e. task tiEnd execute the time, by holding
The row time started codetermines with the time required to executing on a virtual machine, and defined formula is as shown in formula 10.Particularly, workflow
In the last one task tendJust entire workflow execution is represented at the end of execution to be terminated, that is, the deadline of workflow
For task tendEnd execute the time, defined formula is as shown in formula 11.
EFTi=ESTi+ETi,l (10)
WFE=ExFTend (11)
Step 3: the present invention according in HEFT by the upward weight R of the probability of task iiIt is defined as from itself to tendKey
Path length, defined formula is as shown in formula 12, wherein TTi,jIt can be calculated by formula 1,Expression task tiAverage execution
Time, defined formula are as shown in formula 13.According to the upward weight R of probabilityiSize tasks all in workflow are carried out it is preferential
The big minispread of grade, obtains task ranking queue SortedT.
Step 4: the initial task of workflow is t1, the critical path of workflow known to formula 10 is task t1Probability
Upward weight R1, SubDiExpression task tiSub- deadline, when task distributes and run it to virtual machine and terminate to execute
Between need to meet corresponding sub- deadline, defined formula is as shown in formula 14.Wherein, D is the total deadline provided.
Step 5.1: initialization with Ant colony algorithm relevant parameter: relevant parameter includes initial ant colony pheromones τ0, heuristic function
Matrix ηn×m, pheromones heuristic factor α, heuristic greedy method β, the number of iterations Itmax, ant colony quantity AntNmaxDeng;
Step 5.2: iteration starts: judging whether to have reached maximum number of iterations ItmaxIf not reaching, then follow the steps
5.3, otherwise iteration terminates, and goes to step 5.10;
Step 5.3: calculating transition probability: ant antkTask and calculating task t are successively taken from task sequence SortedTi
It distributes to virtual machine vmlProbability;
Transition probability formula are as follows:
Wherein, τi,lIndicate the pheromone concentration on path (i, l);α is pheromones heuristic factor, indicates ant in path
On the importance of pheromones that leaves, the bigger influence for illustrating pheromones to the selection of other ants below path is bigger;β is expectation
Heuristic factor, reflects the importance of heuristic function, β is bigger illustrate ant in moving process heuristic information by attention degree more
It is high.
Define heuristic function matrix ηn×m, every a line of matrix represents a task, and each column represent a virtual machine,
In ηi,lValue represents task tiIt distributes to virtual machine vmlHeuristic function value.
We define the reciprocal as inspiration item, that is, task t of execution costiIn virtual machine vmlConsumed by upper execution
Expense, defined formula is as shown in formula 17, defines ECi,lFor task tiIn virtual machine vmlOn execution cost, defined formula such as formula
Shown in 18.
Step 5.4: selecting virtual machine: according to probability selection formula and meeting sub- deadline schedule virtual for task
Machine.
Step 5.5: judging ant antkWhether completion is searched for, and is gone to step 5.7 if search is completed, is otherwise gone to step
Rapid 5.3.
Step 5.6: local updating pheromones: if an ant completes the scheduling to task, carrying out local updating information
Element goes to step 5.3, otherwise goes to step 5.4.Local information element more new formula are as follows:
Wherein Δ τk i,lFor the sum of the pheromones Increment Matrix that kth ant on path (i, l) is left, Q indicates pheromones
Intensity constant, TLCkAll expenses of virtual machine are used for kth ant.
Step 5.7: judge whether that all ants are complete search, goes to step 5.9 if being fully completed search, it is no
Then go to step 5.3.
Step 5.8: the overall situation updates pheromones and records local optimum scheduling scheme;
All information element more new formula are as follows:
τi,l(t+1)=(1- ρ) τi,l(t)+Δτi,l (20)
Wherein ρ indicates pheromones volatility coefficient;(1- ρ) indicates the residual factor of pheromones, and pheromones are unlimited in order to prevent
Accumulation, value range are limited between 0~1;Δτi,lIndicate the pheromones Increment Matrix that all ants are left on path (i, l)
The sum of, it can be calculated by formula 21.
Step 5.9: output global optimum's scheduling scheme.
Above example is only to illustrate the present invention rather than limits technical solution described in the invention, for those skilled in the art
Member it should be understood that research-on-research in cloud disclosed in foregoing invention flow down deadline constraint Cost Optimization dispatching method,
Do not depart from big name far under the premise of, improvement can also be made on this basis, these improvement are also considered as protection of the invention
Range.
Claims (1)
1. it is an object of the invention to: leased virtual machine, which is dispatched, for scientific workflow in the cloud of deadline constraint is spent
Cost issues, propose the Cost Optimization dispatching method that research-on-research in a kind of cloud flows down deadline constraint, it is intended to reduce
The expense of workflow task lease virtual machine in entire cloud.
In order to achieve the above object, the technical scheme adopted by the invention is that: research-on-research flows down deadline about in a kind of cloud
The Cost Optimization dispatching method of beam, the specific steps are as follows:
Step 1: user submits demand: user submits workflow and relevant resource requirement and entire workflow schedule off period
Limit;
Step 2: related notion being defined: including between the logical of the transmission data this paper objective function, constraint condition, task
Believe the definition of expense, task operation starting time, end runing time etc.;
Step 3: tasks ranking in workflow in cloud: being dropped according to the weight of each task of the upward weight computing of conventional probability and to task
Sequence arrangement obtains task sequence SortedT;
Step 4: distributing sub- deadline: being that task distributes sub- deadline according to the resulting upward weight of task probability of step 3;
Step 5: carrying out task schedule using ant group algorithm: sub- deadline being met to sorted task using ant group algorithm
Under so that cost minimization is dispatched to virtual machine so that final execution cost is minimum.
Further, research-on-research flows down the Cost Optimization dispatching method that deadline constrains in cloud proposed by the present invention, described
Step 1 in the workflow submitted of user described with oriented no circulation figure G={ T, E }, the node T of figure is indicated in workflow
Task-set T={ t1,t2,...,tnN task is shared, the directed edge E of figure indicates the pass that interdepends in workflow between task
Assembly E={ ei,j|ti∈T∩tj∈ T }, wherein ei,jExpression task tiFor task tjForerunner (alternatively task tjFor task
tiIt is subsequent), i.e., only work as task tiTask t could be executed after executionj, and passed between the weight expression task on directed edge
Data set Data={ the data passedi,j|ti∈T∩tj∈ T }, wherein datai,jExpression task tiWith task tjBetween the number that transmits
According to.
Further, the present invention is based on scientific workflow Cost Optimization dispatching method in the cloud of ant group algorithm, in the step 2
The objective function of definition is to lease all virtual machine expense summations, and constraint condition is that its deadline need to meet user-defined section
The only time limit.
s.t.WFE≤D (2)
Wherein LFlIndicate lease virtual machine vmlEnd time, LSlIndicate lease virtual machine vmlAt the beginning of, ClIt indicates to rent
It rents required cost in the virtual machine unit time, in constraint condition, WFE indicates that entire workflow deadline, D indicate user's definition
Deadline.
Further, the present invention is based on scientific workflow Cost Optimization dispatching method, the steps 5 in the cloud of ant group algorithm to make
It is the process of an iteration with the process that ant group algorithm is scheduled sorted task, specific iterative step is as follows:
Step 5.1: initialization with Ant colony algorithm relevant parameter: relevant parameter includes initial ant colony pheromones τ0, heuristic function matrix
ηn×m, pheromones heuristic factor α, heuristic greedy method β, the number of iterations Itmax, ant colony quantity AntNmaxDeng;
Step 5.2: iteration starts: judging whether to have reached maximum number of iterations ItmaxIf not reaching, 5.3 are thened follow the steps,
Otherwise iteration terminates, and goes to step 5.8;
Step 5.3: calculating transition probability: ant antkTask and calculating task t are successively taken from task sequence SortedTiDistribution
To virtual machine vmlProbability;
Step 5.5: selecting virtual machine: according to probability selection formula and meeting sub- deadline scheduling virtual machine for task;
Step 5.6: judging ant antkWhether completion is searched for, and is gone to step 5.7 if search is completed, is otherwise gone to step
5.3;
Step 5.7: local updating pheromones;
Step 5.8: judging whether that all ants are complete search, go to step 5.9 if being fully completed search, otherwise turn
To step 5.3;
Step 5.9: the overall situation updates pheromones and records local optimum scheduling scheme;
Step 5.10: output global optimum's scheduling scheme.
Further, the present invention is based on scientific workflow Cost Optimization dispatching method, the steps 5.3 in the cloud of ant group algorithm
In, define heuristic function matrix ηn×m, every a line of matrix represents a task, and each column represent a virtual machine, therein
ηi,lValue represents task tiIt distributes to virtual machine vmlHeuristic function value.
We define the reciprocal as inspiration item, that is, task t of execution costiIn virtual machine vmlTake consumed by upper execution
With defined formula is as shown in formula 4, defines ECi,lFor task tiIn virtual machine vmlOn execution cost, defined formula such as 5 institute of formula
Show.
Transition probability formula are as follows:
Wherein, τi,lIndicate the pheromone concentration on path (i, l);α is pheromones heuristic factor, indicates that ant stays on path
Under pheromones importance, it is bigger illustrate pheromones to other ants below selection path influence it is bigger;β is that expectation inspires
The factor reflects the importance of heuristic function, and β is bigger to illustrate that ant heuristic information in moving process is higher by attention degree.
Optimization Scheduling provided by the invention has the following advantages and beneficial effects: the present invention in view of appointing in scientific workflow
The characteristic mutually constrained between business is ranked up using execution sequence of the upward weight of conventional probability to task, then uses ant colony
Algorithm carries out under deadline constraint to optimize scheduling of the lease expenses between task-resource of target, and the method can effectively drop
The expense of low task lease virtual machine.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910205996.2A CN109948848A (en) | 2019-03-19 | 2019-03-19 | Research-on-research flows down the Cost Optimization dispatching method of deadline constraint in a kind of cloud |
PCT/CN2019/128110 WO2020186872A1 (en) | 2019-03-19 | 2019-12-25 | Expense optimization scheduling method for deadline constraint under cloud scientific workflow |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910205996.2A CN109948848A (en) | 2019-03-19 | 2019-03-19 | Research-on-research flows down the Cost Optimization dispatching method of deadline constraint in a kind of cloud |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109948848A true CN109948848A (en) | 2019-06-28 |
Family
ID=67009032
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910205996.2A Pending CN109948848A (en) | 2019-03-19 | 2019-03-19 | Research-on-research flows down the Cost Optimization dispatching method of deadline constraint in a kind of cloud |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN109948848A (en) |
WO (1) | WO2020186872A1 (en) |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110908772A (en) * | 2019-11-14 | 2020-03-24 | 北京理工大学 | Energy-saving scheduling method for improving reliability of multiple workflows |
WO2020186872A1 (en) * | 2019-03-19 | 2020-09-24 | 中国石油大学(华东) | Expense optimization scheduling method for deadline constraint under cloud scientific workflow |
CN111882234A (en) * | 2020-08-03 | 2020-11-03 | 浪潮云信息技术股份公司 | Scientific workflow task management method and device |
CN111913800A (en) * | 2020-07-15 | 2020-11-10 | 东北大学秦皇岛分校 | Resource allocation method for optimizing cost of micro-service in cloud based on L-ACO |
CN113127206A (en) * | 2021-04-30 | 2021-07-16 | 东北大学秦皇岛分校 | Cloud environment task scheduling method based on improved ant colony algorithm |
CN113127205A (en) * | 2021-04-30 | 2021-07-16 | 东北大学秦皇岛分校 | Workflow scheduling method meeting deadline constraint and optimizing cost in cloud |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103473122A (en) * | 2013-08-21 | 2013-12-25 | 上海交通大学 | Workflow system resource scheduling method in cloud computing environment |
CN103970609A (en) * | 2014-04-24 | 2014-08-06 | 南京信息工程大学 | Cloud data center task scheduling method based on improved ant colony algorithm |
CN104965755A (en) * | 2015-05-04 | 2015-10-07 | 东南大学 | Cloud service workflow scheduling method based on budget constraint |
CN106055395A (en) * | 2016-05-18 | 2016-10-26 | 中南大学 | Method for constraining workflow scheduling in cloud environment based on ant colony optimization algorithm through deadline |
CN106447173A (en) * | 2016-09-05 | 2017-02-22 | 北京理工大学 | Cloud workflow scheduling method supporting any flow structure |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9779374B2 (en) * | 2013-09-25 | 2017-10-03 | Sap Se | System and method for task assignment in workflows |
CN106844049A (en) * | 2017-01-13 | 2017-06-13 | 广西电网有限责任公司电力科学研究院 | A kind of method for allocating tasks of distributed computing system |
CN108182109B (en) * | 2017-12-28 | 2021-08-31 | 福州大学 | Workflow scheduling and data distribution method in cloud environment |
CN109948848A (en) * | 2019-03-19 | 2019-06-28 | 中国石油大学(华东) | Research-on-research flows down the Cost Optimization dispatching method of deadline constraint in a kind of cloud |
-
2019
- 2019-03-19 CN CN201910205996.2A patent/CN109948848A/en active Pending
- 2019-12-25 WO PCT/CN2019/128110 patent/WO2020186872A1/en active Application Filing
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103473122A (en) * | 2013-08-21 | 2013-12-25 | 上海交通大学 | Workflow system resource scheduling method in cloud computing environment |
CN103970609A (en) * | 2014-04-24 | 2014-08-06 | 南京信息工程大学 | Cloud data center task scheduling method based on improved ant colony algorithm |
CN104965755A (en) * | 2015-05-04 | 2015-10-07 | 东南大学 | Cloud service workflow scheduling method based on budget constraint |
CN106055395A (en) * | 2016-05-18 | 2016-10-26 | 中南大学 | Method for constraining workflow scheduling in cloud environment based on ant colony optimization algorithm through deadline |
CN106447173A (en) * | 2016-09-05 | 2017-02-22 | 北京理工大学 | Cloud workflow scheduling method supporting any flow structure |
Non-Patent Citations (1)
Title |
---|
陈彦橦等: "云科学工作流截止期限约束代价优化调度算法", 《计算机科学与探索》 * |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020186872A1 (en) * | 2019-03-19 | 2020-09-24 | 中国石油大学(华东) | Expense optimization scheduling method for deadline constraint under cloud scientific workflow |
CN110908772A (en) * | 2019-11-14 | 2020-03-24 | 北京理工大学 | Energy-saving scheduling method for improving reliability of multiple workflows |
CN110908772B (en) * | 2019-11-14 | 2022-09-06 | 北京理工大学 | Energy-saving scheduling method for improving reliability of multiple workflows |
CN111913800A (en) * | 2020-07-15 | 2020-11-10 | 东北大学秦皇岛分校 | Resource allocation method for optimizing cost of micro-service in cloud based on L-ACO |
CN111913800B (en) * | 2020-07-15 | 2022-07-19 | 东北大学秦皇岛分校 | Resource allocation method for optimizing cost of micro-service in cloud based on L-ACO |
CN111882234A (en) * | 2020-08-03 | 2020-11-03 | 浪潮云信息技术股份公司 | Scientific workflow task management method and device |
CN113127206A (en) * | 2021-04-30 | 2021-07-16 | 东北大学秦皇岛分校 | Cloud environment task scheduling method based on improved ant colony algorithm |
CN113127205A (en) * | 2021-04-30 | 2021-07-16 | 东北大学秦皇岛分校 | Workflow scheduling method meeting deadline constraint and optimizing cost in cloud |
CN113127206B (en) * | 2021-04-30 | 2022-03-11 | 东北大学秦皇岛分校 | Cloud environment task scheduling method based on improved ant colony algorithm |
CN113127205B (en) * | 2021-04-30 | 2022-05-17 | 东北大学秦皇岛分校 | Workflow scheduling method meeting deadline constraint and optimizing cost in cloud |
Also Published As
Publication number | Publication date |
---|---|
WO2020186872A1 (en) | 2020-09-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109948848A (en) | Research-on-research flows down the Cost Optimization dispatching method of deadline constraint in a kind of cloud | |
CN104317658B (en) | A kind of loaded self-adaptive method for scheduling task based on MapReduce | |
CN103699446B (en) | Quantum-behaved particle swarm optimization (QPSO) algorithm based multi-objective dynamic workflow scheduling method | |
CN106155791B (en) | A kind of workflow task dispatching method under distributed environment | |
Toosi et al. | Revenue maximization with optimal capacity control in infrastructure as a service cloud markets | |
Chen et al. | A set-based discrete PSO for cloud workflow scheduling with user-defined QoS constraints | |
CN101237469B (en) | Method for optimizing multi-QoS grid workflow based on ant group algorithm | |
CN109934416A (en) | Research-on-research flows down the time optimization scheduling method of expense budget constraint in a kind of cloud | |
Kessaci et al. | A pareto-based genetic algorithm for optimized assignment of vm requests on a cloud brokering environment | |
US8316010B2 (en) | Systems and methods for SLA-aware scheduling in cloud computing | |
CN103491024B (en) | Job scheduling method and device for streaming data | |
CN106371924B (en) | A kind of method for scheduling task minimizing MapReduce cluster energy consumption | |
CN109491761A (en) | Cloud computing multiple target method for scheduling task based on EDA-GA hybrid algorithm | |
CN112685138B (en) | Multi-workflow scheduling method based on multi-population hybrid intelligent optimization in cloud environment | |
Niño-Mora | Resource allocation and routing in parallel multi-server queues with abandonments for cloud profit maximization | |
CN108737462A (en) | A kind of cloud computation data center method for scheduling task based on graph theory | |
Zhou et al. | Concurrent workflow budget-and deadline-constrained scheduling in heterogeneous distributed environments | |
Li et al. | Endpoint-flexible coflow scheduling across geo-distributed datacenters | |
CN109710372A (en) | A kind of computation-intensive cloud workflow schedule method based on cat owl searching algorithm | |
CN110111006A (en) | Scientific workflow Cost Optimization dispatching method in a kind of cloud based on chaos Ant ColonySystem | |
Ye et al. | SHWS: Stochastic hybrid workflows dynamic scheduling in cloud container services | |
CN108304253A (en) | Map method for scheduling task based on cache perception and data locality | |
Li et al. | A QoS-based scheduling algorithm for instance-intensive workflows in cloud environment | |
Bessai et al. | Bi-criteria strategies for business processes scheduling in cloud environments with fairness metrics | |
CN109976873A (en) | The scheduling scheme acquisition methods and dispatching method of containerization distributed computing framework |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190628 |