CN107341041A - Cloud task Multi-dimensional constraint backfill dispatching method based on Priority Queues - Google Patents
Cloud task Multi-dimensional constraint backfill dispatching method based on Priority Queues Download PDFInfo
- Publication number
- CN107341041A CN107341041A CN201710500564.5A CN201710500564A CN107341041A CN 107341041 A CN107341041 A CN 107341041A CN 201710500564 A CN201710500564 A CN 201710500564A CN 107341041 A CN107341041 A CN 107341041A
- Authority
- CN
- China
- Prior art keywords
- task
- virtual machine
- backfill
- priority queues
- queue
- 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
-
- 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
- G06F9/4806—Task transfer initiation or dispatching
-
- 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
- G06F9/45533—Hypervisors; Virtual machine monitors
-
- 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
- G06F9/45533—Hypervisors; Virtual machine monitors
- G06F9/45558—Hypervisor-specific management and integration aspects
- G06F2009/4557—Distribution of virtual machine instances; Migration and load balancing
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a kind of cloud task Multi-dimensional constraint based on Priority Queues to backfill dispatching method, it the method overcome in cloud computing and to consider to backfill that performance is bad, and resources of virtual machine dutycycle becomes the defects of big in tradition backfill dispatching algorithm and its improvement backfill dispatching algorithm caused by task index is excessively single；Estimated the two dimensions of execution duration for considering independent task in application processor core quantity and batch tasks are weighed, the Priority Queues based on minimum Binary Heap is constructed using the value for backfilling weight coefficient and backfills the application task that resource is less and the expected time is longer come preferential, and increases the constraintss such as EMS memory occupation and bandwidth occupancy.The present invention improves the utilization rate of virtual machine computing resource in the unit interval, and reduce batch tasks is finally completed the time, reduces task average waiting delay.
Description
Technical field
The present invention relates to field of cloud computer technology, specifically applied to resource management techniques in cloud computing platform, especially
It is related to a kind of cloud task Multi-dimensional constraint backfill dispatching method based on Priority Queues.
Background technology
Nowadays, we are in the epoch of information explosion, and the development of computer technology is maked rapid progress.Instantly trend is user
Amount increases severely, and traditional data center is powerless to meet existing demand.The extensive of cloud computing this new technology is brought therewith
Using his birth and the product of information technology revolution.Cloud computing is generally all the composition towards isomery, and in face of magnanimity
Task requests, the data center of cloud computing how in limited computing resource with the most short time come dispatch deal task, and
And shorten task and be finally completed the time, reduce the average waiting delay of batch tasks, these sides of increase server resource utilization rate
Face is just particularly important.
The more existing method based on cloud computing task scheduling also divides many kinds, and some are based on ant group algorithm, annealing is calculated
The heuritic approaches such as method, genetic algorithm, such as CN103345657A are disclosed under a kind of cloud computing environment based on heredity and ant colony
Method for scheduling task, the advantage for combining genetic algorithm and ant group algorithm is iterated solution.It is public for another CN103870317A
Method for scheduling task and system in a kind of cloud computing are opened, the bacterial foraging algorithm used in this method calculates best effort section
Point, distribution is scheduled to task and resource.The model of these heuritic approach high abstractions in reality towards high timeliness, quickly
Can expose that iteration is very slow in the cloud computing task scheduling processing of response, in face of a large amount of different tasks be easily ensnared into it is local most
Excellent solution, make system response slack-off, reduce Consumer's Experience.
The simple priority weighting for considering task of some dispatching party rules, without many-sided consideration task and virtual machine etc.
The relation of computing resource.As CN103324525A discloses the dispatching method under a kind of cloud computing environment, this method passes through calculating
Each task priority index and task wait exponent pair task to be ranked up scheduling.CN103914754A discloses a kind of work
The method for scheduling task of stream, the average scheduled time and data transfer that this method runs each task on all processors put down
The equal time is combined as weights sequence, then travels through calculate node and the maximum of weights is tied into the optimal virtual machine of performance
On.The carrying of the computing resource of every machine and the resource requirement of independent task are not accounted for so, it is easy to bind task
It is idle that larger virtual machine is produced in virtual machine, resources of virtual machine utilization rate can not be increased to greatest extent.
Also some dispatching methods go to consider task scheduling from the performance triggering of calculate node merely.Such as CN101986272A
Disclose the method for scheduling task under a kind of cloud computing environment.This method does not consider the performance of node in task distribution, only considers
The health degree of node.It is also not processed after the task of distribution exceedes the threshold value of its setting, task is withdrawn to sub-distribution again again.
Task rate of violation is higher caused by so, and the average delay of task can also increase.A kind of and use disclosed in CN105740077A
The method for allocating tasks of cloud computing is the load that each calculate node is established in data center, by task be sequentially allocated load from
On high to Low node, it can so cause node load excessive, instead result in integrated scheduling performance impairment.
Separately there are some dispatching methods to be based purely on task response-time to model to obtain optimal scheduling mode, such as
CN103841208A discloses a kind of cloud computing method for scheduling task optimized based on the response time, and this method is according to task point
The transmission time of piece, task calculate target letter in the average handling time of each node and the total processing time of task burst
Number, is modeled using task response-time as target.But so it is easy to make the less priority of task scheduling of execution time, and holds
Row time longer task response-time becomes big.The scheduling relation between different length task can not be balanced.
In addition, some traditional BACKFILLs and existing follow-on method based on backfill scheduling are when choosing backfill task
This single index of the processor core number of wait task application is only considered, without the resource of application and independent task is pre-
Meter performs duration this index and is combined, this single to consider mode ability only in the case of independent task is performed similar in duration
Effect, the aggravation of virtual machine free time hunger intensity is will result in when independent task execution duration difference in size is larger, is led
It is elongated that cause task is finally completed the time, the increase of task average waiting delay.And the existing task specific under cloud computing environment
There is not one kind in dispatching method also efficiently, and the improvement side based on backfill scheduling of comprehensive consideration multidimensional performance Index Constraints
Method.
The content of the invention
The defects of present invention is for existing modified backfill dispatching method and deficiency, it is proposed that a kind of based on Priority Queues
Cloud task Multi-dimensional constraint backfills dispatching method, and this method comprehensive consideration task submits processor resource and the estimated execution of independent task
The two dimensions of time, which combine, calculates backfill weight coefficient, then weighs the backfill for being submitted to each task after virtual machine
Weight coefficient is built into priority of task queue using the data structure of minimum Binary Heap, and the wait task of batch is carried out into backfill scheduling
Processing, the constraint of the resources such as internal memory, bandwidth is added when scanning backfills simultaneously, to improve the success rate of backfill efficiency and backfill.From
And reach and preengage virtual machine idling-resource backfill in time to greatest extent, shortening task is finally completed the time, and reduction task is put down
Wait the QoS indexes such as time delay.
The technical solution adopted in the present invention is：A kind of cloud task Multi-dimensional constraint backfill dispatching party based on Priority Queues
Method, this method comprise the following steps：
Step 1：The task in task queue will be waited according to assignment instructions length and the processor core quantity ratio of application
The size of value carrys out descending arrangement as the task processing time factor, is stored in common queue；The processing time factor is used for characterizing solely
The expected processing time of vertical task；
Step 2：Virtual machine in data center is calculated into the size descending of disposal ability according to it to arrange, and in data
The heart establishes the numbering of virtual machine and the corresponding relation of its calculating processing ability value；
Step 3：Task in common queue in step 1 is performed into the virtual fleet that duration is tied to step 2 on the estimation
Row, execution duration is longer, and the virtual machine performance of binding is better, if the estimated duration that performs of being assigned to current virtual machine of the task is less than
Equal to next virtual machine then binds the task to current virtual machine, otherwise binds the task to current virtual machine, completes virtual
The preliminary binding of machine；
Step 4：The batch tasks entered in step 3 on every virtual machine, according to the processor core of each task application
Quantity square divided by the backfill weight coefficient that is worth to of ratio of assignment instructions length minimum Binary Heap is established as key, form
Priority of task queue；It is described backfill weight coefficient be characterized as each task application processor core quantity square divided by task
The ratio of command length；
Step 5：First task in minimum Binary Heap is ejected, the task in being currently running in virtual machine is scanned, if together
When meet after the constraints i.e. task backfill of total processor core calculation of virtual machine, total memory size and total bandwidth size not
Disposal ability, internal memory and the bandwidth of virtual machine can be exceeded, then backfilled；If not satisfied, Binary Heap queue is then pressed into again
Again float at end；
Step 6：Step 5 is repeated, is finished until by all task schedulings in Binary Heap.
The calculating disposal ability of described virtual machine passes through MIPS (Million Instructions Per Second, list
Word length fixed point instruction averagely performs speed) characterize.
Beneficial effects of the present invention：The time is performed according to expectation by the task to not submitting to be preferentially assigned to from big to small
Most fast virtual machine, complete preliminary binding, it is possible to reduce the longest finishing time of task；To enter virtual machine after task according to
BFV (BackFill Values, backfill weight coefficient) establishes the priority of task queue of minimum Binary Heap, using this data structure,
Can be efficiently by BFV values minimum, being most badly in need of backfill of the task quickly shifts root node onto.Preferentially the less tasks of BFV are entered
Row backfill can fully reduce the free time of virtual machine, improve resource utilization, while can reduce task average waiting delay.
Brief description of the drawings
Fig. 1 is that the cloud task Multi-dimensional constraint based on Priority Queues backfills patch bay composition；
Fig. 2 is that the cloud task Multi-dimensional constraint based on Priority Queues backfills dispatching method flow chart；
Fig. 3 is the minimum Binary Heap structure chart that the embodiment of the present invention selectes task queue after virtual machine.
Embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, it is right below in conjunction with drawings and Examples
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
Fig. 1 is the cloud task backfill patch bay composition of the present invention, and whole framework is by wait task queue and data central. set
Into, and (DataCenterBroker) acts on behalf of by data center in data center, main frame and the virtual robot arm established on main frame
Into.The role of wherein data center agency is exactly cloud task dispatcher.
Data center is initialized first, data center first turns on some main frames；Then task, data are submitted
Next center can establish the virtual machine of respective numbers on main frame；Some tasks are now run on virtual machine, in data
There are some to wait task queue outside the heart, we are stored with a common queue.
Fig. 2 is the detailed process of whole cloud task Multi-dimensional constraint backfill dispatching method：
Step 1：First against being introduced into the task waiting list of data center according to the T in formula 1ratioValue is entered to task
Row descending is arranged, and the priority of task for making the expected time longer is arranged in before queue.
Step 2：Establish concordance list successively to the virtual machine on All hosts in data center, key is every virtual machine
id；Value is MIPS, i.e. the disposal ability of the virtual machine processor core.Then the MIPS values of virtual machine are directed to according to descending
Arrangement, the virtual machine for making performance best are arranged in the foremost of concordance list.Then the concordance list counted is submitted in data
The heart is acted on behalf of (DataCenterBroker) and preserved.
Step 3：Task in common task queue in step 1 is successively toward the virtual machine in the concordance list in step 2
Upper binding, need to calculate current virtual machine while binding and next virtual machine is accumulative performs the time, when selecting accumulative perform
Between most short virtual machine bound.
Step 4：According to the batch tasks being submitted in step 3 on every virtual machine foundation minimum is carried out according to BFV value
Binary Heap forms Priority Queues.
The batch tasks on every virtual machine are carried out to build heap operation after entering data center after selected virtual machine, made here
It is minimum Binary Heap data structure as shown in Figure 3, minimum Binary Heap is a kind of complete binary tree of stacking order.Such as Fig. 3
What middle T1~T9 nodes stored respectively is the BFV values of task.Wherein T1 is less than T2/T3, and T2 is less than T4/T5, and T3 is less than T6/T7,
T4 is less than T8.T1 is that BFV values are minimum in whole queue as root node, wherein the index value of each left child node is father node
2 times, the index value of right child node is 2 times+1 of father node.The Priority Queues of minimum Binary Heap composition is constructed successively
Step 5：Idling-resource in virtual machine is scanned, first task in Binary Heap is ejected successively, scans in virtual machine
Task in being currently running, carry out task backfill.
Step 5-1：Because the root node in minimum Binary Heap is the minimum task of BFV values, therefore we scan currently again
If idle computing resources in virtual machine be present, first task of Priority Queues is ejected, if then judging to incite somebody to action according to formula 3
Whether task backfill meets resource constraint.If satisfied, then backfilled successively.Wherein PE is processor core quantity,
RAM and BW is the EMS memory occupation amount and bandwidth usage amount of task respectively.
Step 5-2：If being found in step 5-1, the task of ejection is unsatisfactory for resource constraint, by the task again
The tail of the queue of Binary Heap is inserted, backfill then is attempted into the child node ejection before below root node, after backfilling successfully, because from team
The task of tail insertion is likely less than his father node, so according to the definition requirement of Binary Heap, the task is floated up to suitably
Position.
Step 5-3：According to both the above sub-step, by the task of the submission of all virtual machines, all backfill scheduling finishes successively,
Finally wait tasks carrying to finish, destroy virtual machine, whole task scheduling flow finishes.
Technological means disclosed in the present invention program is not limited only to the technological means disclosed in above-mentioned embodiment, in addition to
Formed technical scheme is combined by above technical characteristic.
Claims (2)
- A kind of 1. cloud task Multi-dimensional constraint backfill dispatching method based on Priority Queues, it is characterised in that comprise the following steps,Step 1：The task in task queue will be waited according to assignment instructions length and the processor core quantity ratio of application Size carrys out descending arrangement as the task processing time factor, is stored in common queue；The processing time factor is used for characterizing independent appoint The expected processing time of business；Step 2：Virtual machine in data center is calculated into the size descending of disposal ability according to it to arrange, and built in data center The corresponding relation of the numbering of vertical virtual machine and its calculating processing ability value；Step 3：Task in common queue in step 1 is performed into the virtual machine queue that duration is tied to step 2 on the estimation, held Row duration is longer, and the virtual machine performance of binding is better, if the estimated duration that performs of being assigned to current virtual machine of the task is less than or equal to Next virtual machine then binds the task to current virtual machine, otherwise binds the task to current virtual machine, completes virtual machine Preliminary binding；Step 4：The batch tasks entered in step 3 on every virtual machine, according to the processor core quantity of each task application Square divided by the backfill weight coefficient that is worth to of ratio of assignment instructions length minimum Binary Heap is established as key, form task Priority Queues；It is described backfill weight coefficient be characterized as each task application processor core quantity square divided by assignment instructions The ratio of length；Step 5：First task in minimum Binary Heap is ejected, scans the task in being currently running in virtual machine, if simultaneously full The constraints of total processor core calculation of sufficient virtual machine, total memory size and total bandwidth size is that will not surpass after task backfill Go out disposal ability, internal memory and the bandwidth of virtual machine, then backfilled；If not satisfied, the end of Binary Heap queue is then pressed into again Again float；Step 6：Step 5 is repeated, is finished until by all task schedulings in Binary Heap.
- 2. the cloud task Multi-dimensional constraint backfill dispatching method according to claim 1 based on Priority Queues, it is characterised in that The calculating disposal ability of virtual machine described in step 2 is characterized by MIPS.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710500564.5A CN107341041B (en) | 2017-06-27 | 2017-06-27 | Cloud task multidimensional constraint backfill scheduling method based on priority queue |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710500564.5A CN107341041B (en) | 2017-06-27 | 2017-06-27 | Cloud task multidimensional constraint backfill scheduling method based on priority queue |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107341041A true CN107341041A (en) | 2017-11-10 |
| CN107341041B CN107341041B (en) | 2020-08-11 |
Family
ID=60220835
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710500564.5A Active CN107341041B (en) | 2017-06-27 | 2017-06-27 | Cloud task multidimensional constraint backfill scheduling method based on priority queue |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107341041B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111080298A (en) * | 2019-12-26 | 2020-04-28 | 电子科技大学 | Block generation and transaction verification method suitable for energy source block chain |
| CN111782355A (en) * | 2020-06-03 | 2020-10-16 | 上海交通大学 | Cloud computing task scheduling method and system based on mixed load |
| WO2021073111A1 (en) * | 2019-10-15 | 2021-04-22 | 平安科技（深圳）有限公司 | Distributed storage file reading and writing method, device and platform, and readable storage medium |
| CN113132265A (en) * | 2021-04-16 | 2021-07-16 | 武汉光迅信息技术有限公司 | Multi-stage scheduling method and device for multi-path Ethernet |
| CN113192322A (en) * | 2021-03-19 | 2021-07-30 | 东北大学 | Expressway traffic flow counting method based on cloud edge cooperation |
| CN113238841A (en) * | 2021-05-11 | 2021-08-10 | 湖北碳聚生物科技有限责任公司 | Task scheduling method based on cloud computing technology |
| CN117193963A (en) * | 2023-08-03 | 2023-12-08 | 北京大学 | Function feature-based server non-aware computing scheduling method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101819540A (en) * | 2009-02-27 | 2010-09-01 | 国际商业机器公司 | Method and system for scheduling task in cluster |
| US20120304275A1 (en) * | 2011-05-24 | 2012-11-29 | International Business Machines Corporation | Hierarchical rule development and binding for web application server firewall |
| CN103761147A (en) * | 2014-01-15 | 2014-04-30 | 清华大学 | Method and system for managing calculation examples in cloud platforms |
| CN104731642A (en) * | 2013-12-23 | 2015-06-24 | 国际商业机器公司 | Backfill scheduling method and system for embarrassingly parallel jobs |
-
2017
- 2017-06-27 CN CN201710500564.5A patent/CN107341041B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101819540A (en) * | 2009-02-27 | 2010-09-01 | 国际商业机器公司 | Method and system for scheduling task in cluster |
| US20120304275A1 (en) * | 2011-05-24 | 2012-11-29 | International Business Machines Corporation | Hierarchical rule development and binding for web application server firewall |
| CN104731642A (en) * | 2013-12-23 | 2015-06-24 | 国际商业机器公司 | Backfill scheduling method and system for embarrassingly parallel jobs |
| CN103761147A (en) * | 2014-01-15 | 2014-04-30 | 清华大学 | Method and system for managing calculation examples in cloud platforms |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021073111A1 (en) * | 2019-10-15 | 2021-04-22 | 平安科技（深圳）有限公司 | Distributed storage file reading and writing method, device and platform, and readable storage medium |
| CN111080298A (en) * | 2019-12-26 | 2020-04-28 | 电子科技大学 | Block generation and transaction verification method suitable for energy source block chain |
| CN111080298B (en) * | 2019-12-26 | 2023-12-29 | 电子科技大学 | Block generation and transaction verification method suitable for energy block chain |
| CN111782355A (en) * | 2020-06-03 | 2020-10-16 | 上海交通大学 | Cloud computing task scheduling method and system based on mixed load |
| CN111782355B (en) * | 2020-06-03 | 2024-05-28 | 上海交通大学 | Cloud computing task scheduling method and system based on mixed load |
| CN113192322A (en) * | 2021-03-19 | 2021-07-30 | 东北大学 | Expressway traffic flow counting method based on cloud edge cooperation |
| CN113192322B (en) * | 2021-03-19 | 2022-11-25 | 东北大学 | Expressway traffic flow counting method based on cloud edge cooperation |
| CN113132265A (en) * | 2021-04-16 | 2021-07-16 | 武汉光迅信息技术有限公司 | Multi-stage scheduling method and device for multi-path Ethernet |
| CN113132265B (en) * | 2021-04-16 | 2022-05-10 | 武汉光迅信息技术有限公司 | Multi-stage scheduling method and device for multi-path Ethernet |
| CN113238841A (en) * | 2021-05-11 | 2021-08-10 | 湖北碳聚生物科技有限责任公司 | Task scheduling method based on cloud computing technology |
| CN113238841B (en) * | 2021-05-11 | 2023-12-12 | 湖北碳聚生物科技有限责任公司 | Task scheduling method based on cloud computing technology |
| CN117193963A (en) * | 2023-08-03 | 2023-12-08 | 北京大学 | Function feature-based server non-aware computing scheduling method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107341041B (en) | 2020-08-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107341041A (en) | Cloud task Multi-dimensional constraint backfill dispatching method based on Priority Queues | |
| CN109561148B (en) | Distributed task scheduling method based on directed acyclic graph in edge computing network | |
| CN108182115B (en) | Virtual machine load balancing method in cloud environment | |
| CN111694656B (en) | Cluster resource scheduling method and system based on multi-agent deep reinforcement learning | |
| CN107357652B (en) | Cloud computing task scheduling method based on segmentation ordering and standard deviation adjustment factor | |
| CN104333569A (en) | Cloud task scheduling algorithm based on user satisfaction | |
| CN103473134A (en) | Dependent task scheduling method of heterogeneous multi-core processor | |
| CN103336723A (en) | Scheduling method for adapting to shortage of processor cores and based on critical paths | |
| CN111813506A (en) | Resource sensing calculation migration method, device and medium based on particle swarm algorithm | |
| CN107122233A (en) | A kind of adaptive real-time scheduling methods of many VCPU towards TSN business | |
| CN111611062A (en) | Cloud-edge collaborative hierarchical computing method and cloud-edge collaborative hierarchical computing system | |
| CN107832143A (en) | A kind for the treatment of method and apparatus of physical machine resource | |
| CN109005211B (en) | Micro-cloud deployment and user task scheduling method in wireless metropolitan area network environment | |
| CN115237568A (en) | Mixed weight task scheduling method and system for edge heterogeneous equipment | |
| CN104461722B (en) | A kind of job scheduling method for cloud computing system | |
| CN113722112B (en) | Service resource load balancing processing method and system | |
| CN109871270B (en) | Scheduling scheme generation method and device | |
| CN110888745A (en) | MEC node selection method considering task transmission arrival time | |
| CN114579282A (en) | Task scheduling method and device | |
| CN110084507A (en) | The scientific workflow method for optimizing scheduling of perception is classified under cloud computing environment | |
| CN112685162A (en) | High-efficiency scheduling method, system and medium for heterogeneous computing resources of edge server | |
| CN116244050B (en) | High-response-ratio-based pneumatic characteristic example scheduling method | |
| CN102073548A (en) | Method for executing task, and system thereof | |
| CN115098240B (en) | Multiprocessor application scheduling method and system and storage medium | |
| CN102736955B (en) | Computational grid task scheduling method based on reliability and non-cooperation game |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |