CN110399206A - One kind is based on IDC virtualization scheduling energy conserving system under cloud computing environment - Google Patents
One kind is based on IDC virtualization scheduling energy conserving system under cloud computing environment Download PDFInfo
- Publication number
- CN110399206A CN110399206A CN201910529656.5A CN201910529656A CN110399206A CN 110399206 A CN110399206 A CN 110399206A CN 201910529656 A CN201910529656 A CN 201910529656A CN 110399206 A CN110399206 A CN 110399206A
- Authority
- CN
- China
- Prior art keywords
- node
- centermanager
- job
- resource
- cloud computing
- 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
- G06F9/46—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
- G06F9/46—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
- G06F9/505—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 load
-
- 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
Abstract
The invention discloses one kind based on IDC virtualization scheduling energy conserving system under cloud computing environment, including tetra- modules of CenterManager, Executor, Zoo-Keeper and ClientAPI, wherein CenterManager is divided to for two submodules of ResourceManager and JobScheduler, user is submitted to CenterManager, inquires and is controlled operation, and the resource status of inquiry application cluster or individual node by order line, WebPortal and ClientAPI;This programme, not only need rational management system resource, the higher utilization rate of guarantee system, and it needs sufficiently to meet the different customization demand of operation, establish shifting to new management mechanisms for the soft or hard constraint of operation, type effectively, which is resided, for different types of batch processing job and internet is served by unified management, not only it is added significantly to the scale of Application share cluster, and make full use of the feature of different types of application performance bottleneck and the complementation of high offpeak period, the resource utilization for improving application cluster significantly reduces the total cost of ownership of data center.
Description
Technical field
The invention belongs to dispatch systems technology field, more specifically, more particularly to it is a kind of based on IDC under cloud computing environment
Virtualization scheduling energy conserving system.Meanwhile the invention further relates to one kind based on IDC virtualization scheduling energy conserving system under cloud computing environment
Method.
Background technique
Internet service has more type, performance bottleneck multiplicity, deployment process are complicated, the fluctuation of load is big to deposit between application
The dependence the features such as, therefore the demand to resource proposes different constraint condition.A kind of constraint condition is to must satisfy
's.Due to developing the limitation with translation and compiling environment, some applications may require that the version of node operating system, GLIBC or GCC are necessary
Higher than some specific version;In order to guarantee performance, the application such as graphics process and biological DNA sequence dna matching may require that node is matched
High performance GPU is set instead of common CPU;The memory of matrix operation application requirement node CPU with high performance and large capacity,
Etc., the above-mentioned constraint condition that must satisfy is referred to as " hard constraint " by us.
In addition, application programming model can propose some operation constraint conditions to optimize transaction capabilities also.Such as Hadoop
It runs it is required that operation is issued on the node of storing data to save network bandwidth;Network
Intensive applications, it is desirable that operation is issued on identical or neighboring switch node and is run, it is identical to make full use of
Or the node of neighboring switch has relatively high network bandwidth.These constraint conditions usually require preferentially to meet, we claim
Be " soft-constraint ".
In order to support mixed type operation, job scheduling strategy must simultaneously meet above-mentioned a variety of different types of operation constraints
Condition.Existing operation constrained dispatch all towards batch application scene, is unable to satisfy batch processing and service type mixing applied field
The demand of scape can not meet the diversified performance optimization demand of different type operation, lead to serious waste of resources.
Summary of the invention
The purpose of the present invention is to solve problems in the prior art, and the one kind proposed is based on IDC under cloud computing environment
Virtualization scheduling energy conserving system.
To achieve the above object, the invention provides the following technical scheme: a kind of adjusted based on IDC virtualization under cloud computing environment
Energy conserving system, including tetra- modules of CenterManager, Executor, Zoo-Keeper and ClientAPI are spent, wherein
CenterManager is divided to for two submodules of ResourceManager and JobScheduler, user by order line,
WebPortal and ClientAPI submits to CenterManager, inquires and controls operation, and inquiry application cluster or list
The resource status of a node;JobScheduler is responsible for application packet management and job scheduling, is matched according to the resource of application packet
Volume and with resource quantity selection priority it is high, there is the application packet for waiting operation to carry out job scheduling, according to application packet
The job scheduling algorithm of selection carries out job scheduling, and the job information being selected (JobClassAd) is sent to
ResourceManager carries out resource matched.
Preferably, the ResourceManager is responsible for condition monitoring, management and the operation money of all nodes of application cluster
Source matching, the Executor are responsible for the starting, execution and condition monitoring of operation, and regularly converge to Res0urceManager
Report node resource state information.
Preferably, the zoo-Keeper is configuration center, it saves all modules and application cluster section, meanwhile, Zoo-
The configuration information of Keeper or the core point of high availability mechanism;Also act as the effect of name Service.
Preferably, the CenterMan.ager delay machine or service it is unavailable when, backup CenterManager passes through
The new main CenterManager of ZooKeeper selection.
Preferably, all job run status informations, user is written to ZooKeeper in real time in the JobScheduler
The job informations such as operating status, Node distribution and the network port of the operation can be obtained by job title with application.
Preferably, the CenterManager uses distributed document, such as Google file system (GFS), Hadoop points
Cloth file system (HDFS) and Network File System (NFS) etc. solve submission and deployment, work data and the log letter of operation
Data sharing problem between the preservation and application of breath, in order to improve application performance, system uses resource container
(ResourceContainer), such as LinuxContainer etc., rather than traditional virtual machine, realize application performance and
Security isolation is possessing high performance simultaneously as resource container itself does not have independently operated operating system, but and place
Host sharing operation system kernel and runtime environment, this requires systems must provide additional support scheme, such as unified
The runtime environments such as operating system version and Glibc, Gcc, Java avoid the Internet application by identical port from being deployed in same
Node etc..
Preferably, the CenterManager further includes dynamic port: the operation for not requiring fixed port, such as
Business Logic and MySQL database etc., user setting job property Dynamic.Port attribute are True, indicate to need to make
Industry distributes a dynamic port;ResourceManager is dynamically its point according to node port occupancy situation when resource matched
With an available port number;When Executor initiating task can by the end No. El write people to ZooKeeper name Service, facilitate it
Its operation inquiry.
Preferably, the CenterManager further includes dependent form operation module: service type job load fluctuates big spy
Sign is so that operation in the deployment scenario of data center's node is also dynamic change;Operation can only specify them to rely on work when submitting
The title of industry;When operation is resource matched, ResourceManager is relied on operation from name Service inquiry according to job title and transports
Capable node listing generates job dependence constraint condition.
Preferably, the CenterManager further includes node failure processing module: being compiled with MapReduce, Hadoop etc.
Journey frame is similar, and scheduling strategy needs to formulate the treatment mechanism that a set of reply node ceases to be in force automatically;However, dependent form operation from
Dynamic crash handling requires to keep the dependence between operation, that is, is relied on after operation is rescheduled and issues execution, relies on
Operation is issued on identical node automatically and runs;This allows for rescheduling when being relied on operation, and ResourceManager must
The constraint condition for being relied on operation and relying on operation must be combined, be dynamically generated one while meeting the combination of the two service requirement
Constraint condition.
A kind of method based on IDC virtualization scheduling energy conserving system under cloud computing environment provided by the invention, including it is as follows
Step:
It can operations dynamic be believed by the node rack of operation, IP address, port and disk be exclusive etc. when S1, Executor initiating task
Breath is written in ZooKeeper name Service in real time;
When S2, operation are resource matched, ResourceManager inquires according to job title from ZooKeeper name Service above-mentioned
Operation multidate information, and replace operation dynamic constrained attribute with it and generate operation dynamic constrained condition;
S3, ResourceManager are resource matched according to the progress operation of operation dynamic constrained condition, and node listing information is returned
Back to JobScheduler;
Operation is issued to Executor and executed by S4, JobScheduler.
Technical effect and advantage of the invention:
This programme not only needs rational management system resource, guarantees the higher utilization rate of system, and needs sufficiently to meet operation
Different customization demands, establishes shifting to new management mechanisms for the soft or hard constraint of operation, effectively for different types of batch processing job and mutually
Resident type of networking is served by unified management, is not only added significantly to the scale of Application share cluster, but also make full use of
The feature of different types of application performance bottleneck and the complementation of high offpeak period, improves the utilization of resources of application cluster significantly
Rate reduces the total cost of ownership of data center.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, below in conjunction with specific embodiment, to this
Invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, not
For limiting the present invention.Based on the embodiments of the present invention, those of ordinary skill in the art are not before making creative work
Every other embodiment obtained is put, shall fall within the protection scope of the present invention.
Embodiment 1
One kind is based on IDC virtualization scheduling energy conserving system under cloud computing environment, including CenterManager, Executor, Zoo-
Tetra- modules of Keeper and ClientAPI, wherein CenterManager points are ResourceManager and JobScheduler
Two submodules, user are submitted to CenterManager, inquire and are controlled by order line, WebPortal and ClientAPI
Operation, and the resource status of inquiry application cluster or individual node;JobScheduler is responsible for application packet management and work
Industry scheduling selects the application point that priority is high, has waiting operation according to the resource quota of application packet and with resource quantity
Group carries out job scheduling, carries out job scheduling according to the job scheduling algorithm of application packet selection, and the operation being selected is believed
It is resource matched that breath (JobClassAd) is sent to ResourceManager progress.
Further, the ResourceManager is responsible for condition monitoring, management and the operation of all nodes of application cluster
Resource matched, the Executor is responsible for the starting, execution and condition monitoring of operation, and regularly to Res0urceManager
Report node resource state information.
Further, the zoo-Keeper is configuration center, it saves all modules and application cluster section, meanwhile,
The configuration information of Zoo-Keeper or the core point of high availability mechanism;Also act as the effect of name Service.
Further, the CenterMan.ager delay machine or service it is unavailable when, backup CenterManager passes through
The new main CenterManager of ZooKeeper selection.
Further, all job run status informations are written to ZooKeeper in real time in the JobScheduler, use
Family and application can obtain the job informations such as operating status, Node distribution and the network port of the operation by job title.
Further, the CenterManager uses distributed document, such as Google file system (GFS), Hadoop
Distributed file system (HDFS) and Network File System (NFS) etc. solve submission and deployment, the work data and log of operation
Data sharing problem between the preservation and application of information, in order to improve application performance, system uses resource container
(ResourceContainer), such as LinuxContainer etc., rather than traditional virtual machine, realize application performance and
Security isolation is possessing high performance simultaneously as resource container itself does not have independently operated operating system, but and place
Host sharing operation system kernel and runtime environment, this requires systems must provide additional support scheme, such as unified
The runtime environments such as operating system version and Glibc, Gcc, Java avoid the Internet application by identical port from being deployed in same
Node etc..
Further, the CenterManager further includes dynamic port: the operation for not requiring fixed port, example
Such as Business Logic and MySQL database, user setting job property Dynamic.Port attribute are True, indicate that needs are
Operation distributes a dynamic port;ResourceManager is dynamically it according to node port occupancy situation when resource matched
Distribute an available port number;The end El can be write to people to ZooKeeper name Service when Executor initiating task, it is convenient
Other operation inquiries.
Further, the CenterManager further includes dependent form operation module: service type job load fluctuates big
Feature makes operation in the deployment scenario of data center's node be also dynamic change;Operation can only specify them to rely on when submitting
The title of operation;When operation is resource matched, ResourceManager is relied on operation from name Service inquiry according to job title
The node listing of operation generates job dependence constraint condition.
Further, the CenterManager further includes node failure processing module: with MapReduce, Hadoop etc.
Programming framework is similar, and scheduling strategy needs to formulate the treatment mechanism that a set of reply node ceases to be in force automatically;However, dependent form operation
The processing requirement that ceases to be in force automatically keeps the dependence between operation, that is, is relied on after operation is rescheduled and issues execution, according to
Bad operation is issued on identical node automatically and runs;This allows for rescheduling when being relied on operation, ResourceManager
It must be dynamically generated one in conjunction with the constraint condition for being relied on operation and dependence operation while meet the group of the two service requirement
Close constraint condition.
A kind of method based on IDC virtualization scheduling energy conserving system under cloud computing environment provided by the invention, including it is as follows
Step:
It can operations dynamic be believed by the node rack of operation, IP address, port and disk be exclusive etc. when S1, Executor initiating task
Breath is written in ZooKeeper name Service in real time;
When S2, operation are resource matched, ResourceManager inquires according to job title from ZooKeeper name Service above-mentioned
Operation multidate information, and replace operation dynamic constrained attribute with it and generate operation dynamic constrained condition;
S3, ResourceManager are resource matched according to the progress operation of operation dynamic constrained condition, and node listing information is returned
Back to JobScheduler;
Operation is issued to Executor and executed by S4, JobScheduler.
Therefore this programme not only needs rational management system resource, guarantees the higher utilization rate of system, and needs abundant
Meet the different customization demand of operation, establish shifting to new management mechanisms for the soft or hard constraint of operation, is effectively different types of batch processing
Operation and internet reside type and are served by unified management, are not only added significantly to the scale of Application share cluster, but also
The feature for making full use of different types of application performance bottleneck and the complementation of high offpeak period, improves the money of application cluster significantly
Source utilization rate reduces the total cost of ownership of data center.
Finally, it should be noted that the foregoing is only a preferred embodiment of the present invention, it is not intended to restrict the invention,
Although the present invention is described in detail referring to the foregoing embodiments, for those skilled in the art, still may be used
To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features,
All within the spirits and principles of the present invention, any modification, equivalent replacement, improvement and so on should be included in of the invention
Within protection scope.
Claims (10)
1. one kind is based on IDC virtualization scheduling energy conserving system under cloud computing environment, it is characterised in that: including CenterManager,
Tetra- modules of Executor, Zoo-Keeper and ClientAPI, wherein CenterManager point for ResourceManager and
Two submodules of JobScheduler, user are mentioned by order line, WebPortal and ClientAPI to CenterManager
Hand over, inquire and control operation, and the resource status of inquiry application cluster or individual node;JobScheduler is responsible for application
Grouping management and job scheduling select priority height according to the resource quota of application packet and with resource quantity, have waiting
The application packet of operation carries out job scheduling, carries out job scheduling according to the job scheduling algorithm of application packet selection, and will be by
It is resource matched that the job information (JobClassAd) of selection is sent to ResourceManager progress.
2. a kind of virtualized based on IDC under cloud computing environment according to claim 1 dispatches energy conserving system, feature exists
In: condition monitoring, management and the operation that the ResourceManager is responsible for all nodes of application cluster are resource matched, described
Executor is responsible for the starting, execution and condition monitoring of operation, and regularly reports node resource shape to Res0urceManager
State information.
3. a kind of virtualized based on IDC under cloud computing environment according to claim 2 dispatches energy conserving system, feature exists
In: the zoo-Keeper is configuration center, it saves all modules and application cluster section, meanwhile, Zoo-Keeper or height
The configuration information of the core point of available mechanism;Also act as the effect of name Service.
4. a kind of virtualized based on IDC under cloud computing environment according to claim 3 dispatches energy conserving system, feature exists
In: when the CenterMan.ager delay machine or unavailable service, backup CenterManager is selected by ZooKeeper
New main CenterManager.
5. a kind of virtualized based on IDC under cloud computing environment according to claim 4 dispatches energy conserving system, feature exists
In: all job run status informations are written to ZooKeeper in real time in the JobScheduler, and user and application can lead to
It crosses job title and obtains the job informations such as operating status, Node distribution and the network port of the operation.
6. a kind of virtualized based on IDC under cloud computing environment according to claim 5 dispatches energy conserving system, feature exists
In: the CenterManager uses distributed document, such as Google file system (GFS), Hadoop distributed file system
(HDFS) and Network File System (NFS) etc., solve the submission and deployment of operation, the preservation of work data and log information and
Data sharing problem between, in order to improve application performance, system uses resource container (ResourceContainer),
Such as LinuxContainer etc., rather than traditional virtual machine realize the performance and security isolation of application, are possessing high-performance
While, since resource container itself does not have independently operated operating system, but with host sharing operation system kernel
And runtime environment, this requires system must provide additional support scheme, such as uniform operating system version and Glibc,
The runtime environments such as Gcc, Java avoid the Internet application by identical port from being deployed in same node etc..
7. a kind of virtualized based on IDC under cloud computing environment according to claim 6 dispatches energy conserving system, feature exists
In: the CenterManager further includes dynamic port: the operation for not requiring fixed port, such as Business Logic and
MySQL database etc., user setting job property Dynamic.Port attribute are True, and expression, which needs to distribute one for operation, to be moved
State port;ResourceManager dynamically distributes an available end according to node port occupancy situation for it when resource matched
Slogan;The end El can be write to people to ZooKeeper name Service when Executor initiating task, other operations is facilitated to inquire.
8. a kind of virtualized based on IDC under cloud computing environment according to claim 6 dispatches energy conserving system, feature exists
In: the CenterManager further includes dependent form operation module: service type job load fluctuates big feature and operation is existed
The deployment scenario of data center's node is also dynamic change;Operation can only specify them to rely on the title of operation when submitting;Make
When industry is resource matched, ResourceManager is arranged according to job title from the node that name Service inquiry is relied on job run
Table generates job dependence constraint condition.
9. a kind of virtualized based on IDC under cloud computing environment according to claim 6 dispatches energy conserving system, feature exists
In: the CenterManager further includes node failure processing module: it is similar with the programming frameworks such as MapReduce, Hadoop,
Scheduling strategy needs to formulate the treatment mechanism that a set of reply node ceases to be in force automatically;However, the processing that ceases to be in force automatically of dependent form operation
It is required that keep the dependence between operation, that is, be relied on after operation is rescheduled and issues execution, rely on operation it is automatic under
It is sent on identical node and runs;This allows for rescheduling when being relied on operation, ResourceManager must in conjunction with by according to
Rely operation and rely on the constraint condition of operation, be dynamically generated one while meeting the combined constraint conditions of the two service requirement.
10. a kind of method described in claim 1 based on IDC virtualization scheduling energy conserving system under cloud computing environment, feature
It is: includes the following steps:
It can operations dynamic be believed by the node rack of operation, IP address, port and disk be exclusive etc. when S1, Executor initiating task
Breath is written in ZooKeeper name Service in real time;
When S2, operation are resource matched, ResourceManager inquires according to job title from ZooKeeper name Service above-mentioned
Operation multidate information, and replace operation dynamic constrained attribute with it and generate operation dynamic constrained condition;
S3, ResourceManager are resource matched according to the progress operation of operation dynamic constrained condition, and node listing information is returned
Back to JobScheduler;
Operation is issued to Executor and executed by S4, JobScheduler.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910529656.5A CN110399206B (en) | 2019-06-19 | 2019-06-19 | IDC virtualization scheduling energy-saving system based on cloud computing environment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910529656.5A CN110399206B (en) | 2019-06-19 | 2019-06-19 | IDC virtualization scheduling energy-saving system based on cloud computing environment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110399206A true CN110399206A (en) | 2019-11-01 |
CN110399206B CN110399206B (en) | 2022-04-05 |
Family
ID=68323267
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910529656.5A Active CN110399206B (en) | 2019-06-19 | 2019-06-19 | IDC virtualization scheduling energy-saving system based on cloud computing environment |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110399206B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113641495A (en) * | 2021-08-12 | 2021-11-12 | 成都中科大旗软件股份有限公司 | Distributed scheduling method and system based on big data calculation |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103399787A (en) * | 2013-08-06 | 2013-11-20 | 北京华胜天成科技股份有限公司 | Map Reduce task streaming scheduling method and scheduling system based on Hadoop cloud computing platform |
US20140006620A1 (en) * | 2012-06-27 | 2014-01-02 | International Business Machines Corporation | System, method and program product for local client device context-aware shared resource and service management |
US20150256481A1 (en) * | 2014-03-06 | 2015-09-10 | Jisto Inc. | Elastic Compute Cloud Based On Underutilized Server Resources Using A Distributed Container System |
CN107341051A (en) * | 2016-05-03 | 2017-11-10 | 北京京东尚科信息技术有限公司 | Cluster task coordination approach, system and device |
-
2019
- 2019-06-19 CN CN201910529656.5A patent/CN110399206B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140006620A1 (en) * | 2012-06-27 | 2014-01-02 | International Business Machines Corporation | System, method and program product for local client device context-aware shared resource and service management |
CN103399787A (en) * | 2013-08-06 | 2013-11-20 | 北京华胜天成科技股份有限公司 | Map Reduce task streaming scheduling method and scheduling system based on Hadoop cloud computing platform |
US20150256481A1 (en) * | 2014-03-06 | 2015-09-10 | Jisto Inc. | Elastic Compute Cloud Based On Underutilized Server Resources Using A Distributed Container System |
CN107341051A (en) * | 2016-05-03 | 2017-11-10 | 北京京东尚科信息技术有限公司 | Cluster task coordination approach, system and device |
Non-Patent Citations (2)
Title |
---|
ALBERT REUTHER: "Scheduler Technologies in Support of High Perfomance Data Analysis", 《2016 IEEE HIGH PERFORMANCE EXTREME COMPUTING CONFERENCE》 * |
大鱼-瓶邪: "hadoop版本YARN架构理解", 《HTTPS://BLOG.CSDN.NET/QQ_25948717/ARTICLE/DETAILS/80554809》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113641495A (en) * | 2021-08-12 | 2021-11-12 | 成都中科大旗软件股份有限公司 | Distributed scheduling method and system based on big data calculation |
Also Published As
Publication number | Publication date |
---|---|
CN110399206B (en) | 2022-04-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US9542223B2 (en) | Scheduling jobs in a cluster by constructing multiple subclusters based on entry and exit rules | |
US20190377604A1 (en) | Scalable function as a service platform | |
US20200326993A1 (en) | Reducing overlay network overhead across container hosts | |
US11301290B1 (en) | System and method for controlled sharing of consumable resources in a computer cluster | |
US10733019B2 (en) | Apparatus and method for data processing | |
CN103069390B (en) | Method and system for re-scheduling workload in a hybrid computing environment | |
CN108701059A (en) | Multi-tenant resource allocation methods and system | |
EP2930618A2 (en) | System and method for load balancing compute resources | |
JP2015537307A (en) | Component-oriented hybrid cloud operating system architecture and communication method thereof | |
JPH02127757A (en) | Execution of dispersion application program for data processing network | |
CN107992359A (en) | The task scheduling algorithm that cost perceives under a kind of cloud environment | |
JPH02148224A (en) | Scheduling of time starting task | |
CN105786603B (en) | Distributed high-concurrency service processing system and method | |
CN103927225A (en) | Multi-core framework Internet information processing and optimizing method | |
Amalarethinam et al. | An Overview of the scheduling policies and algorithms in Grid Computing | |
WO2024016596A1 (en) | Container cluster scheduling method and apparatus, device, and storage medium | |
Yeh et al. | Realizing dynamic resource orchestration on cloud systems in the cloud-to-edge continuum | |
CN112764909B (en) | Sharing method and system based on cloud architecture workstation | |
CN110399206A (en) | One kind is based on IDC virtualization scheduling energy conserving system under cloud computing environment | |
CN107528871A (en) | Data analysis in storage system | |
EP2930620A2 (en) | Dynamic provisioning of processing resources in a virtualised computational architecture | |
US11630834B2 (en) | Label-based data representation I/O process and system | |
CN109450913A (en) | A kind of multinode registration dispatching method based on strategy | |
Li et al. | Cress: Dynamic scheduling for resource constrained jobs | |
Awad et al. | Tasks Scheduling Techniques in Cloud Computing |
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 |