CN110851263A - Green cloud task scheduling method for heterogeneous cloud data center - Google Patents
Green cloud task scheduling method for heterogeneous cloud data center Download PDFInfo
- Publication number
- CN110851263A CN110851263A CN201911117023.XA CN201911117023A CN110851263A CN 110851263 A CN110851263 A CN 110851263A CN 201911117023 A CN201911117023 A CN 201911117023A CN 110851263 A CN110851263 A CN 110851263A
- Authority
- CN
- China
- Prior art keywords
- cloud
- minimum
- task
- matrix
- completion time
- 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
- 238000000034 method Methods 0.000 title claims abstract description 26
- 239000011159 matrix material Substances 0.000 claims abstract description 48
- 238000005265 energy consumption Methods 0.000 claims abstract description 13
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 10
- 238000004364 calculation method Methods 0.000 abstract description 2
- 238000005516 engineering process Methods 0.000 description 5
- 238000001514 detection method Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000004904 shortening Methods 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
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
- G06F9/4843—Task transfer initiation or dispatching by program, e.g. task dispatcher, supervisor, operating system
- G06F9/4881—Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues
-
- 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
-
- 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]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
- G06F9/5038—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering the execution order of a plurality of tasks, e.g. taking priority or time dependency constraints into consideration
-
- 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
- G06F2009/45575—Starting, stopping, suspending or resuming virtual machine instances
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
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
A green cloud task scheduling method for a heterogeneous cloud data center comprises the following steps: (1) calculating the expected execution time of each task on all the virtual machines to form an expected execution time matrix; (2) calculating a corresponding minimum completion time matrix and selecting a cloud task to be executed firstly; (3) listing all cloud tasks with the minimum completion time on the marked virtual machine, and searching the minimum completion time and the second minimum completion time of the cloud tasks; (4) calculating minimum differences of the minimum completion time and the second minimum completion time with the completion time of the cloud task selected firstly; (5) determining a cloud task which is deployed firstly, and updating an MCT matrix; (6) judging whether the cloud task is distributed completely, if not, repeating the step (2), and if so, ending the whole scheduling algorithm; according to the cloud task scheduling method and device, automatic acquisition of cloud task parameters of the cloud data center, calculation of loss comparison and intelligent scheduling of cloud tasks can be achieved, and therefore energy consumption of the cloud data center is reduced to the maximum extent.
Description
Technical Field
The invention relates to the technical field of distributed computing and cloud computing, in particular to a green cloud task scheduling method for a heterogeneous cloud data center.
Background
With the continuous popularization of cloud computing, more and more applications are managed to a cloud platform to run, so that the cost of enterprises and users for maintaining hardware is basically eliminated, the operation cost is further reduced, and users can be physically and mentally put into the development of the applications without paying attention to corresponding infrastructure maintenance. However, the cloud computing system is composed of a large number of physical hosts distributed in different regions, and the performances of the physical hosts are very different, specifically, the performances are represented by the strength of computing power, the size of storage capacity and the like, so that the heterogeneity of the cloud computing system is formed.
The cloud computing is divided according to a service mode and mainly comprises three layers: IaaS (infrastructure as service), PaaS (platform as a service), and SaaS (software as a service). In the three levels of research, the scheduling algorithm is always one of the core problems, especially in the IaaS layer, since the layer only provides basic computing resources, the importance of the scheduling algorithm is further shown. In some large simulation tasks, due to the repeatability of scheduling tasks, the execution time of the tasks on the heterogeneous clusters can be obtained in an early stage, due to the special structure of the scheduling tasks, existing scheduling algorithms are not applicable any more, most of the scheduling algorithms mainly aim at shortening the completion time and improving the load balance, and the energy consumption of a cloud data center is ignored.
Disclosure of Invention
In order to overcome the problems in the prior art, the invention provides a green cloud task scheduling method for a heterogeneous cloud data center, and a cloud task detection, scheduling performance analysis and automatic scheduling method is constructed by applying a cloud computing technology, a cloud task scheduling technology and a cloud data center distributed management technology; the cloud task management method can realize automatic acquisition of cloud task parameters of the cloud data center, calculation of loss comparison and intelligent scheduling of the cloud tasks, so that the energy consumption of the cloud data center is reduced to the maximum extent.
In order to achieve the purpose, the invention adopts the technical scheme that:
a green cloud task scheduling method for a heterogeneous cloud data center comprises the following steps:
(1) calculating the expected execution time of each task on all the virtual machines to form an n-m order expected Execution Time (ETC) matrix:
when a plurality of repeatedly executed cloud tasks need to be scheduled, an n x m-order expected Execution Time (ETC) matrix is listed according to the execution time of each cloud task on each virtual machine, and the matrix provides a basis for calculating a Minimum Completion Time (MCT) matrix in the next step;
(2) calculate the corresponding minimum time to complete (MCT) matrix and select the cloud task that is executed first:
if the cloud task is executed for the first time, the ETC matrix is the same as the MCT matrix, otherwise, the MCT matrix needs to be updated, the minimum completion time of each task in the MCT matrix is selected to form a time set, the minimum completion time is selected from the time set, and the column number of the matrix is marked, namely the virtual machine number;
(3) listing all cloud tasks whose minimum completion times are on the marked virtual machine, and finding their minimum completion times and next minimum completion times:
further selecting all cloud tasks with the minimum completion time on the virtual machine according to the requirements of the loss comparison rule, and sequentially searching the minimum and second minimum completion times for loss comparison;
(4) calculating minimum differences of the minimum completion time and the second minimum completion time with the first selected cloud task:
for all cloud tasks selected in the step (2) and the step (3), subtracting the minimum completion time of the cloud tasks by using the minimum and the second minimum completion time of the cloud tasks to calculate the minimum difference of the completion time of the cloud tasks, wherein the minimum difference represents the minimum execution time required by the cloud task if the cloud task is not executed on the virtual machine with the minimum completion time;
(5) determining the cloud task deployed first, and updating an MCT matrix:
firstly, respectively subtracting the minimum difference value of the completion time of the cloud task selected in the third step from the minimum difference value of the completion time of the cloud task selected in the second step, wherein the result represents a loss comparison value; secondly, if the loss comparison value is less than or equal to 0, abandoning, continuing to deploy the cloud task selected in the second step, and if the loss comparison value is greater than 0, selecting the cloud task with the maximum loss comparison value for deployment; finally, deleting the selected deployed cloud tasks from the matrix;
(6) and (3) judging whether the cloud task is distributed and completed, if not, repeating the step (2), and if so, ending the whole scheduling algorithm:
the scheduling process is completed and the final scheduling result is generated, so that the execution efficiency of the cloud task and the utilization rate of the virtual machines are guaranteed, the number of the virtual machines can be reduced, and the energy consumption of the cloud data center is reduced.
Compared with the prior art, the invention has the beneficial effects that:
firstly, the method comprehensively analyzes a plurality of factors influencing the energy consumption of the cloud data center, and mainly comprises scheduling energy consumption, dynamic energy consumption, static energy consumption and other energy consumption. Secondly, the minimum completion time and the second minimum completion time of the cloud task are comprehensively considered, and the cloud task is selected by utilizing the difference value of the two completion times of different cloud tasks on the same virtual machine. Finally, the characteristics of the cloud task are comprehensively considered, as the real data center is mostly formed by heterogeneous machines with different configurations, each physical machine can show different performances due to different configurations of a CPU, an internal memory and the like, and the method provided by the invention gives consideration to the heterogeneity of the cluster.
Drawings
Fig. 1 is an example of cloud task scheduling.
Fig. 2 is a cloud task scheduling layer flow diagram.
Detailed Description
In order to make the features, processes and advantages of the present invention more apparent, the present invention will be described in further detail below with reference to the accompanying drawings and examples. It is to be understood that the embodiments described herein are merely illustrative of the basic idea of the invention and are not intended to limit the invention.
The invention discloses a green cloud task scheduling method for a heterogeneous cloud data center, which is applied to a cloud data center platform, wherein the platform mainly comprises three parts: the system comprises a cloud task queue layer, a virtual machine queue layer and a cloud task scheduling layer. The cloud task queue layer comprises all cloud tasks needing to be scheduled and belongs to a demand side. The virtual machine queue layer contains all virtual machines in the cloud data center and belongs to the supplier. However, the two layers need to be connected through a cloud task scheduling layer, and the layer is used for scheduling the cloud task to the virtual machine to complete the deployment of the cloud task.
The cloud task queue layer contains all cloud task information submitted by a user, and the cloud task queue layer can classify the cloud tasks according to parameters such as execution time and deadline of the cloud tasks. Due to the development of modern cloud computing technology, a user puts higher requirements on the real-time performance and the computing capacity of cloud computing, so that a cloud task can show different characteristics according to different requirements of the user, wherein the attributes mainly comprise CPU (Central processing Unit) utilization rate, memory, hard disk and the like, each attribute represents the requirement of the cloud task, and all the attributes can be met only when the cloud task is deployed.
The virtual machine queue layer is used as an expression form of cloud data center resources, and since the modern large-scale cloud data center adopts a virtualization technology to improve the utilization rate of the resources, the virtual machines are generally used as basic units to provide services for users and cloud tasks. Therefore, the generation, migration and destruction of the virtual machine can affect the energy consumption of the cloud data center. In the project of cloud task deployment, the virtual machine is timely turned on and off according to the condition of the current cloud data center.
The cloud task scheduling layer is used as the middle part of the platform and mainly comprises three methods: the data processing method comprises a data reasonability detection method, a cloud task scheduling method and a data formatting method. The data reasonableness detection method is mainly used for processing parameter information of a cloud task queue layer and a virtual machine queue layer, analyzing demand information and supply information of the cloud task queue layer and the virtual machine queue layer, and removing parts with incomplete or unreasonable information. And outputting the detected reasonable data to a cloud task scheduling method. The cloud task scheduling method analyzes received data, firstly, an ETC matrix and an MCT matrix are generated, and the two matrixes are used as basic scheduling information and must be generated firstly. Secondly, calculating a corresponding MCT matrix, selecting the cloud task which is executed firstly, listing all the cloud tasks with the minimum completion time on the marked virtual machine, and searching the minimum completion time and the second minimum completion time of the cloud tasks. And thirdly, respectively calculating the minimum difference value of the completion time between the cloud tasks and the first selected cloud task, and determining the cloud task to be deployed. And finally, updating the MCT matrix, and continuing to deploy other cloud tasks until all the cloud tasks are deployed. The data formatting method formats the cloud task information generated by the cloud task scheduling method according to the requirements of the platform, determines the final deployment information and completes the deployment of the cloud task.
Based on the introduction of the cloud data center platform, the invention provides a green cloud task scheduling method for a heterogeneous cloud data center, which comprises the following steps:
(1) calculating the expected execution time of each task on all the virtual machines to form an n-m order expected Execution Time (ETC) matrix:
when a plurality of repeatedly executed cloud tasks need to be scheduled, an n x m-order expected Execution Time (ETC) matrix is listed according to the execution time of each cloud task on each virtual machine, and the matrix provides a basis for calculating a Minimum Completion Time (MCT) matrix in the next step;
(2) calculate the corresponding minimum time to complete (MCT) matrix and select the cloud task that is executed first:
if the cloud task is executed for the first time, the ETC matrix is the same as the MCT matrix, otherwise, the MCT matrix needs to be updated, the minimum completion time of each task in the MCT matrix is selected to form a time set, the minimum completion time is selected from the time set, and the column number of the matrix is marked, namely the virtual machine number;
(3) listing all cloud tasks whose minimum completion times are on the marked virtual machine, and finding their minimum completion times and next minimum completion times:
further selecting all cloud tasks with the minimum completion time on the virtual machine according to the requirements of the loss comparison rule, and sequentially searching the minimum and second minimum completion times for loss comparison;
(4) calculating minimum differences of the minimum completion time and the second minimum completion time with the first selected cloud task:
for all cloud tasks selected in the step (2) and the step (3), subtracting the minimum completion time of the cloud tasks by using the minimum and the second minimum completion time of the cloud tasks to calculate the minimum difference of the completion time of the cloud tasks, wherein the minimum difference represents the minimum execution time required by the cloud task if the cloud task is not executed on the virtual machine with the minimum completion time;
(5) determining the cloud task deployed first, and updating an MCT matrix:
firstly, respectively subtracting the minimum difference value of the completion time of the cloud task selected in the third step from the minimum difference value of the completion time of the cloud task selected in the second step, wherein the result represents a loss comparison value; secondly, if the loss comparison value is less than or equal to 0, abandoning, continuing to deploy the cloud task selected in the second step, and if the loss comparison value is greater than 0, selecting the cloud task with the maximum loss comparison value for deployment; finally, deleting the selected deployed cloud tasks from the matrix;
(6) and (3) judging whether the cloud task is distributed and completed, if not, repeating the step (2), and if so, ending the whole scheduling algorithm:
the scheduling process is completed and the final scheduling result is generated, so that the execution efficiency of the cloud task and the utilization rate of the virtual machines are guaranteed, the number of the virtual machines can be reduced, and the energy consumption of the cloud data center is reduced.
As shown in fig. 1, the following illustrates a scheduling process of a cloud task.
Assume that a user requested cloud task includes T1、T2、T3、T4、T5And T6The virtual machine comprises V1、V2And V3As shown in fig. 1. Step1, listing an ETC matrix and an MCT matrix, wherein the two matrices are the same; step2, selecting the minimum completion time of each task to form a sequence [100,150,110,190,110,200 ]]Selecting the smallest 100 of the sequence to execute, which should be T0Is deployed at V0Above, but found T2And T5At V0The completion time of (c) is also minimal; step3, calculating T2And T0By a loss ratio of 40, using T2Instead of T0Then, calculate T again5And T2Is 120, and T is used5Instead of T2Finally, will T5Is deployed at V0The above step (1); step4, delete T5I.e. the 5 th row in the matrix, and then updating the MCT matrix, i.e. the second matrix; step4, repeat the above method to select 110 for deployment, i.e. T4Deployed at T2And sequentially executing the tasks until all the tasks are deployed, wherein the execution time of each virtual machine is as follows: v0=120+110=230,V1=150+220=370,V 2110+120 230. The specific processing flow is shown in fig. 2.
According to the invention, according to different characteristics of the cloud task and the virtual machine, the deployment result of the cloud task is changed through loss comparison, the average execution time of the cloud data center is reduced, and the energy consumption of the cloud data center is further reduced. Compared with other cloud task scheduling algorithms, the performance of the algorithm is improved by about 22% on average.
Claims (1)
1. A green cloud task scheduling method for a heterogeneous cloud data center is characterized by comprising the following steps:
(1) calculating the expected execution time of each task on all the virtual machines to form an n-m order expected execution time ETC matrix:
when a plurality of repeatedly executed cloud tasks need to be scheduled, an n x m-order estimated execution time ETC matrix is listed according to the execution time of each cloud task on each virtual machine, and the matrix provides a basis for calculating a minimum completion time MCT matrix in the next step;
(2) calculating a corresponding minimum completion time MCT matrix and selecting a cloud task to be executed firstly:
if the cloud task is executed for the first time, the ETC matrix is the same as the MCT matrix, otherwise, the MCT matrix needs to be updated, the minimum completion time of each task in the MCT matrix is selected to form a time set, the minimum completion time is selected from the time set, and the column number of the matrix is marked, namely the virtual machine number;
(3) listing all cloud tasks whose minimum completion times are on the marked virtual machine, and finding their minimum completion times and next minimum completion times:
further selecting all cloud tasks with the minimum completion time on the virtual machine according to the requirements of the loss comparison rule, and sequentially searching the minimum and second minimum completion times for loss comparison;
(4) calculating minimum differences of the minimum completion time and the second minimum completion time with the first selected cloud task:
for all cloud tasks selected in the step (2) and the step (3), subtracting the minimum completion time of the cloud tasks by using the minimum and the second minimum completion time of the cloud tasks to calculate the minimum difference of the completion time of the cloud tasks, wherein the minimum difference represents the minimum execution time required by the cloud task if the cloud task is not executed on the virtual machine with the minimum completion time;
(5) determining the cloud task deployed first, and updating an MCT matrix:
firstly, respectively subtracting the minimum difference value of the completion time of the cloud task selected in the third step from the minimum difference value of the completion time of the cloud task selected in the second step, wherein the result represents a loss comparison value; secondly, if the loss comparison value is less than or equal to 0, abandoning, continuing to deploy the cloud task selected in the second step, and if the loss comparison value is greater than 0, selecting the cloud task with the maximum loss comparison value for deployment; finally, deleting the selected deployed cloud tasks from the matrix;
(6) and (3) judging whether the cloud task is distributed and completed, if not, repeating the step (2), and if so, ending the whole scheduling algorithm:
the scheduling process is completed and the final scheduling result is generated, so that the execution efficiency of the cloud task and the utilization rate of the virtual machines are guaranteed, the number of the virtual machines can be reduced, and the energy consumption of the cloud data center is reduced.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911117023.XA CN110851263A (en) | 2019-11-15 | 2019-11-15 | Green cloud task scheduling method for heterogeneous cloud data center |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911117023.XA CN110851263A (en) | 2019-11-15 | 2019-11-15 | Green cloud task scheduling method for heterogeneous cloud data center |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110851263A true CN110851263A (en) | 2020-02-28 |
Family
ID=69600740
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911117023.XA Pending CN110851263A (en) | 2019-11-15 | 2019-11-15 | Green cloud task scheduling method for heterogeneous cloud data center |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110851263A (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040093601A1 (en) * | 2002-11-07 | 2004-05-13 | Quicksilver Technology, Inc. | Method, system and program for developing and scheduling adaptive integrated circuity and corresponding control or configuration information |
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 |
CN106201701A (en) * | 2016-07-14 | 2016-12-07 | 扬州大学 | A kind of workflow schedule algorithm of band task duplication |
CN107479949A (en) * | 2017-08-21 | 2017-12-15 | 重庆电子工程职业学院 | Low energy consumption cloud computing method for scheduling task |
-
2019
- 2019-11-15 CN CN201911117023.XA patent/CN110851263A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040093601A1 (en) * | 2002-11-07 | 2004-05-13 | Quicksilver Technology, Inc. | Method, system and program for developing and scheduling adaptive integrated circuity and corresponding control or configuration information |
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 |
CN106201701A (en) * | 2016-07-14 | 2016-12-07 | 扬州大学 | A kind of workflow schedule algorithm of band task duplication |
CN107479949A (en) * | 2017-08-21 | 2017-12-15 | 重庆电子工程职业学院 | Low energy consumption cloud computing method for scheduling task |
Non-Patent Citations (5)
Title |
---|
BO LI等: ""Minimum Completion Time Offloading Algorithm for Mobile Edge Computing"", 《2018 IEEE 4TH INTERNATIONAL CONFERENCE ON COMPUTER AND COMMUNICATIONS (ICCC)》 * |
何翔等: ""基于系统损失和任务复杂度的网格调度算法"", 《计算机科学》 * |
刘永等: "节能及信任驱动的虚拟机资源调度", 《计算机应用研究》 * |
董仕: ""网格仿真器及任务调度算法研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
袁佳欣: ""云计算平台任务调度算法研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US20190332422A1 (en) | Dynamic accelerator scheduling and grouping for deep learning jobs in a computing cluster | |
CN102984137A (en) | Multi-target server scheduling method based on multi-target genetic algorithm | |
CN114399227A (en) | Production scheduling method and device based on digital twins and computer equipment | |
US20190377606A1 (en) | Smart accelerator allocation and reclamation for deep learning jobs in a computing cluster | |
CN114820279B (en) | Distributed deep learning method and device based on multiple GPUs and electronic equipment | |
CN106339475A (en) | Distributed storage system for mass data | |
Zhou et al. | Strategy optimization of resource scheduling based on cluster rendering | |
CN110134646B (en) | Knowledge platform service data storage and integration method and system | |
Fu et al. | Network traffic based virtual machine migration in cloud computing environment | |
CN115759979B (en) | Intelligent process processing method and system based on RPA and process mining | |
CN111966447A (en) | Container placing method based on double-row genetic algorithm | |
CN110851263A (en) | Green cloud task scheduling method for heterogeneous cloud data center | |
RU2609076C2 (en) | Method and system for smart control over distribution of resources in cloud computing environments | |
Lin et al. | A workload-driven approach to dynamic data balancing in MongoDB | |
CN116431281A (en) | Virtual machine migration method based on whale optimization algorithm | |
CN110888713A (en) | Trusted virtual machine migration algorithm for heterogeneous cloud data center | |
CN113778627B (en) | Scheduling method for creating cloud resources | |
Chen et al. | A cloud computing resource scheduling scheme based on estimation of distribution algorithm | |
CN113504966B (en) | GPU cluster scheduling strategy simulation method and GPU cluster simulator | |
Liu et al. | Design of a service robot cloud service platform | |
CN114417577A (en) | Cross-platform resource scheduling and optimization control method | |
Mao et al. | A fine-grained and dynamic MapReduce task scheduling scheme for the heterogeneous cloud environment | |
CN114237884A (en) | Cloud computing data center energy consumption optimization method based on gray Markov chain | |
CN110647381B (en) | Virtual machine resource balancing and deployment optimizing method | |
CN110908800B (en) | Scheduling optimization method based on scheduling entropy in edge cloud collaborative environment |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200228 |
|
RJ01 | Rejection of invention patent application after publication |