CN101604264B - Task scheduling method and system for supercomputer - Google Patents
Task scheduling method and system for supercomputer Download PDFInfo
- Publication number
- CN101604264B CN101604264B CN2009101087124A CN200910108712A CN101604264B CN 101604264 B CN101604264 B CN 101604264B CN 2009101087124 A CN2009101087124 A CN 2009101087124A CN 200910108712 A CN200910108712 A CN 200910108712A CN 101604264 B CN101604264 B CN 101604264B
- Authority
- CN
- China
- Prior art keywords
- task
- computing node
- parallel section
- energy consumption
- application load
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 30
- 238000005265 energy consumption Methods 0.000 claims abstract description 57
- 238000010586 diagram Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Images
Classifications
-
- 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
- Multi Processors (AREA)
Abstract
The invention provides a task scheduling method and a task scheduling system for a supercomputer. The method comprises that: A, a computing node looks up an application load sheet to judge whether the computing node can complete a parallel part of a task to be operated, and if so, the parallel part of the task is requested to be operated; B, a scheduling center receives an operation request of the computing node to acquire energy consumption increased when the computing node completes the parallel part of the task; and C, the scheduling center selects a computing node with the minimally increased energy consumption, and schedules the parallel part of the task to the computing node. The task scheduling method and the task scheduling system for the supercomputer can effectively reduce the energy consumption of the supercomputer.
Description
Technical field
The present invention relates to field of computer technology, more particularly, relate to a kind of method for scheduling task and system of supercomputer.
Background technology
Supercomputer is meant the parallel computing machine that carries out large-scale calculations or data processing of ability that a plurality of computing nodes combine, and is also referred to as parallel computer.Supercomputer can consume a large amount of electric energy when operation, usually, the supercomputing chance is through the state of dispatching center's all computing nodes of statistics, according to the node state scheduling task of computing node.And supercomputer has thousands of computing nodes; Only add up, screen and dispatch, be difficult to travel through all computing nodes, and cause congested easily by the dispatching center; Thereby consume lot of energy when making the supercomputer operation, also reduced the operational efficiency of supercomputer simultaneously.
Summary of the invention
Based on this, be necessary to provide a kind of method for scheduling task that can reduce the supercomputer of energy consumption.
In addition, also be necessary to provide a kind of task scheduling system that can reduce the supercomputer of energy consumption.
The method for scheduling task of said supercomputer comprises: the A. computing node is consulted the application load table, judges whether computing node can accomplish the parallel section of the task of waiting for operation, if then request moves the parallel section of said task; B. the dispatching center receives the operation request of computing node, obtains the energy consumption that parallel section increased that computing node is accomplished said task; C. the minimum computing node of the energy consumption of said increase is selected by said dispatching center, and the parallel section of said task is dispatched to this computing node.
Preferably, the content of said application load table record comprises: the computing power that the operation address of the running status of task, task, the parallel section of operation task are required and the required energy consumption of parallel section of operation task.
Judge in the said steps A computing node can accomplish the task of waiting for operation parallel section step specifically: computing node is consulted the application load table; Obtain the required computing power of parallel section of accomplishing the task of waiting for operation; When the residue computing power of said computing node self during, judge that then said computing node can accomplish the parallel section of the task of said wait operation greater than the required computing power of said parallel section.
Also comprise after the said step C: D. moves the running status of the parallel section of said task by waiting for that operation is updated to, and the operation address of upgrading the parallel section of said task.
In addition, said method also comprises: the E. computing node is consulted the application load table, judges whether computing node can accomplish the parallel section of moving of task, if then request moves the parallel section of said moving of task; F. the dispatching center receives the operation request of computing node, obtains the energy consumption that parallel section increased of the parallel section of said moving of task being moved to computing node and this task of this computing node operation; G. the application load table is consulted by said dispatching center; Obtain the required energy consumption of parallel section of the said task of operation; The required ability of the parallel section of the said task of operation that writes down in less than said application load table when the energy consumption of said increase is consuming time, then the parallel section of said moving of task is moved to said computing node.
Preferably, also comprise after the said step G: operation address and the required energy consumption of parallel section of upgrading the said task of operation of upgrading the parallel section of said task.
The task scheduling system of said supercomputer comprises: at least one computing node; Said computing node is used to consult the application load table; Judge whether computing node can accomplish the parallel section of the task of waiting for operation, if then request moves the parallel section of said task; The dispatching center; Carry out alternately with said computing node; Be used to receive the operation request of computing node; Obtain computing node and accomplish the energy consumption that parallel section increased of said task, and the minimum computing node of energy consumption of selecting said increase, and the parallel section of the task of said wait operation is dispatched to this computing node.
Preferably, said computing node also is used to consult the application load table, judges whether computing node can accomplish the parallel section of moving of task, if then request moves the parallel section of said task.
Further preferably; Said dispatching center also is used to receive the operation request of computing node; The parallel section that obtains will moving of task is moved to the energy consumption that parallel section increased that computing node and this computing node move said task; And consult the application load table; Obtain the required energy consumption of parallel section of the said task of operation, the required ability of the parallel section of the said task of operation that writes down in less than said application load table when the energy consumption of said increase is consuming time, then the parallel section of said task is moved to said computing node.
In the method for scheduling task and system of above-mentioned supercomputer; Consult the application load table through computing node; When computing node self can be accomplished the parallel section of the task of waiting for operation, then active request was moved this parallel section, and the dispatching center obtains each computing node completion energy consumption that this parallel section increased; Select the minimum computing node of energy consumption increased, and with task scheduling to this computing node.Because computing node is according to self-ability active request operation task, and the minimum computing node operation task of energy consumption is selected by the dispatching center, like this, can effectively reduce the energy consumption of supercomputer.
Description of drawings
Fig. 1 is the process flow diagram of the method for scheduling task of supercomputer among the embodiment;
Fig. 2 is the process flow diagram of the method for scheduling task of supercomputer among another embodiment;
Fig. 3 is the synoptic diagram of the task scheduling system of supercomputer among the embodiment.
Embodiment
Fig. 1 shows the method for scheduling task flow process of supercomputer among the embodiment, and detailed process is following:
In step S101, computing node is consulted the application load table, judges when computing node can be accomplished the parallel section of the task of waiting for operation, then the parallel section of the said task of request operation.
In step S102, the dispatching center receives the operation request of computing node, obtains the energy consumption that parallel section increased that computing node is accomplished said task.
In step S103, the minimum computing node of energy consumption of said increase is selected by the dispatching center, and the parallel section of said task is dispatched to this computing node.
In one embodiment, the application load table can be stored on the computing node, supplies dispatching center and each computing node to consult.The recordable content of application load table comprises: energy consumption that the computing power that the operation address of the running status of task, task, each parallel section of operation task are required and each parallel section of operation task are required etc.Wherein, the running status of task comprises: wait for and move, moving and moving completion; The computing node address of each parallel section of this task is promptly moved in the operation address of task.
In one embodiment; Judge computing node can accomplish the task of waiting for operation parallel section process specifically: computing node is consulted the application load table; Obtain the required computing power of parallel section of accomplishing the task of waiting for operation; When the residue computing power of computing node self during greater than the required computing power of said parallel section, this computing node of then judging can be accomplished the parallel section of the task that said wait moves.The computing power here is meant central processing power (specifically can be flops), memory power (specifically can be memory size) and the network capabilities (specifically can be the network bandwidth) etc. of computing node.
In one embodiment, task scheduling after increase the minimum computing node of energy consumption, is upgraded the application load table, specifically: the running status of the parallel section of said task is moved by waiting for that operation is updated to; Upgrade the operation address (promptly moving the address of the computing node of this parallel section) of the parallel section of said task.Because all computing nodes can both initiatively be participated in the process of dispatching; When computing node was judged a certain parallel section that self can finish the work, then this parallel section of active request operation calculated it simultaneously and accomplishes the energy consumption that this parallel section increased and offer the dispatching center; The minimum computing node of energy consumption is then selected by the dispatching center; Make task scheduling more reasonable, fair and quick like this, can effectively reduce the energy consumption of supercomputer, also can improve the operational efficiency of supercomputer.
Fig. 2 shows the method for scheduling task flow process of supercomputer among another embodiment, and this flow process has been described the process of the parallel section of the task that computing node request operation moving, and detailed process is following:
In step S201, computing node is consulted the application load table, judges when computing node can be accomplished the parallel section of moving of task, then the parallel section of the said task of request operation.
In step S202, the dispatching center receives the operation request of computing node, obtains the energy consumption that parallel section increased of the parallel section of this moving of task being moved to computing node and this computing node operation task.
In step S203, the application load table is consulted by the dispatching center, obtains the required energy consumption of parallel section of the said task of operation.This energy consumption is promptly just at the energy that computing node consumed of the parallel section of operation task.
In step S204, whether the energy consumption of judging said increase is less than the required energy consumption of parallel section of the said task of operation that writes down in the application load table, if then get into step S205, otherwise finish.
In step S205, the parallel section of said moving of task is moved to said computing node.
In step S206, upgrade operation address and the required energy consumption of parallel section of upgrading the said task of operation of the parallel section of said task.In one embodiment; Accomplish the energy consumption that a certain parallel section increased of moving of task and this parallel section is moved to the summation of the energy consumption that this computing node increases consuming time less than the ability of the computing node that moves this parallel section when computing node, then computing node can be moved to this parallel section of task in the dispatching center.And the operation address of the parallel section of this task that writes down in the application load table also will be updated to the address of new computing node by the address of former computing node, and the required energy consumption of parallel section of this task of operation of record also will be changed to the energy consumption of new this parallel section of computing node operation by the energy consumption of former computing node.
Fig. 3 shows the task scheduling system of supercomputer among the embodiment, this system call center 100 and carry out at least one mutual computing node (computing node 200,300......N) with it.Wherein:
Each computing node (computing node 200,300......N) is used to consult the application load table, judges whether computing node accomplishes the parallel section of the task of waiting for operation, if then request moves the parallel section of said task.
In one embodiment, the application load table can be stored on the computing node, supplies dispatching center and each computing node to consult.The recordable content of application load table comprises: energy consumption that the computing power that the operation address of the running status of task, task, each parallel section of operation task are required and each parallel section of operation task are required etc.Wherein, the running status of task comprises: wait for and move, moving and moving completion; The address at computing node place of the parallel section of this task is promptly moved in the operation address of task.
In one embodiment, each computing node (computing node 200,300......N) also can be used for consulting the application load table, judges whether it can accomplish the parallel section of moving of task, if then request moves the parallel section of this moving of task.The 100 operation requests that receive computing node in dispatching center; The parallel section that obtains will moving of task is moved to the energy consumption that parallel section increased that computing node and this computing node move said task; Simultaneously; The application load table is consulted by dispatching center 100; Obtain the required energy consumption of parallel section of this task of operation that writes down in the application load table, the required ability of the parallel section of the operation task that writes down in less than the application load table when the energy consumption of above-mentioned increase is consuming time, then the parallel section of this moving of task is moved to this computing node.
The above is merely preferred embodiment of the present invention, not in order to restriction the present invention, all any modifications of within spirit of the present invention and principle, being done, is equal to and replaces and improvement etc., all should be included within protection scope of the present invention.
Claims (5)
1. the method for scheduling task of a supercomputer is characterized in that, said method comprises:
A. computing node is consulted the application load table, judges whether computing node can accomplish the parallel section of the task of waiting for operation, if then request moves the parallel section of said task;
B. the dispatching center receives the operation request of computing node, obtains the energy consumption that parallel section increased that computing node is accomplished said task;
C. the minimum computing node of the energy consumption of said increase is selected by said dispatching center, and the parallel section of said task is dispatched to this computing node;
Also comprise after the said step C:
D. the running status of the parallel section of said task is moved by waiting for that operation is updated to, and the operation address of upgrading the parallel section of said task;
Said method also comprises:
E. computing node is consulted the application load table, judges whether computing node can accomplish the parallel section of moving of task, if then request moves the parallel section of said moving of task;
F. the dispatching center receives the operation request of computing node, obtains the energy consumption that parallel section increased of the parallel section of said moving of task being moved to computing node and this task of this computing node operation;
G. the application load table is consulted by said dispatching center; Obtain the required energy consumption of parallel section of the said task of operation; The required ability of the parallel section of the said task of operation that writes down in less than said application load table when the energy consumption of said increase is consuming time, then the parallel section of said moving of task is moved to said computing node.
2. the method for scheduling task of supercomputer according to claim 1; It is characterized in that the content of said application load table record comprises: the computing power that the operation address of the running status of task, task, the parallel section of operation task are required and the required energy consumption of parallel section of operation task.
3. the method for scheduling task of supercomputer according to claim 2 is characterized in that, judge in the said steps A computing node can accomplish the task of waiting for operation parallel section step specifically:
Computing node is consulted the application load table; Obtain the required computing power of parallel section of accomplishing the task of waiting for operation; When the residue computing power of said computing node self during, judge that then said computing node can accomplish the parallel section of the task of said wait operation greater than the required computing power of said parallel section.
4. the method for scheduling task of supercomputer according to claim 1 is characterized in that, also comprises after the said step G:
Upgrade operation address and the required energy consumption of parallel section of upgrading the said task of operation of the parallel section of said task.
5. the task scheduling system of a supercomputer is characterized in that, said system comprises:
At least one computing node, said computing node is used to consult the application load table, judges whether computing node can accomplish the parallel section of the task of waiting for operation, if then request moves the parallel section of said task;
The dispatching center; Carry out alternately with said computing node; Be used to receive the operation request of computing node; Obtain computing node and accomplish the energy consumption that parallel section increased of said task, and the minimum computing node of energy consumption of selecting said increase, and the parallel section of the task of said wait operation is dispatched to this computing node;
Said computing node also is used to consult the application load table, judges whether computing node can accomplish the parallel section of moving of task, if then request moves the parallel section of said task;
Said dispatching center also is used to receive the operation request of computing node; The parallel section that obtains will moving of task is moved to the energy consumption that parallel section increased that computing node and this computing node move said task; And consult the application load table; Obtain the required energy consumption of parallel section of the said task of operation; The required ability of the parallel section of the said task of operation that writes down in less than said application load table when the energy consumption of said increase is consuming time, then the parallel section of said task is moved to said computing node.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009101087124A CN101604264B (en) | 2009-07-08 | 2009-07-08 | Task scheduling method and system for supercomputer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009101087124A CN101604264B (en) | 2009-07-08 | 2009-07-08 | Task scheduling method and system for supercomputer |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101604264A CN101604264A (en) | 2009-12-16 |
| CN101604264B true CN101604264B (en) | 2012-07-25 |
Family
ID=41470005
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2009101087124A Expired - Fee Related CN101604264B (en) | 2009-07-08 | 2009-07-08 | Task scheduling method and system for supercomputer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101604264B (en) |
Families Citing this family (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102567106B (en) * | 2010-12-30 | 2014-10-01 | 中国移动通信集团云南有限公司 | Task scheduling method, system and device |
| CN102289277B (en) * | 2011-07-06 | 2013-12-11 | 中国科学院深圳先进技术研究院 | Dispatching method for data center application services |
| CN102508714A (en) * | 2011-11-03 | 2012-06-20 | 南京邮电大学 | Green-computer-based virtual machine scheduling method for cloud computing |
| CN102420863B (en) * | 2011-11-30 | 2015-04-15 | 北京蓝汛通信技术有限责任公司 | Rapid file distribution system, method thereof and apparatus thereof |
| CN102567076B (en) * | 2011-12-12 | 2014-12-17 | 华中科技大学 | Selection method of server integration scheme based on energy consumption |
| GB2505412A (en) * | 2012-08-28 | 2014-03-05 | Ibm | Collaborative modified consensus workload distribution between computing nodes |
| CN104572272B (en) * | 2013-10-12 | 2018-02-09 | 杭州华为数字技术有限公司 | A kind of method for scheduling task, apparatus and system |
| CN104331326A (en) * | 2014-11-25 | 2015-02-04 | 华南师范大学 | Scheduling method and system for cloud computing |
| CN106203757A (en) * | 2015-05-06 | 2016-12-07 | 阿里巴巴集团控股有限公司 | Managing computing resources method and apparatus |
| CN105162844B (en) * | 2015-08-05 | 2018-07-13 | 中国联合网络通信集团有限公司 | A kind of method and device of task distribution |
| CN107153573A (en) * | 2016-03-02 | 2017-09-12 | 阿里巴巴集团控股有限公司 | Distributed task scheduling treating method and apparatus |
| CN106095572B (en) * | 2016-06-08 | 2019-12-06 | 东方网力科技股份有限公司 | distributed scheduling system and method for big data processing |
| CN106126323B (en) * | 2016-06-17 | 2019-11-22 | 广州商品清算中心股份有限公司 | Real-time task scheduling method based on cloud platform |
| CN106095586A (en) * | 2016-06-23 | 2016-11-09 | 东软集团股份有限公司 | A kind of method for allocating tasks, Apparatus and system |
| CN110543148B (en) * | 2018-05-28 | 2021-04-09 | 华为技术有限公司 | Task scheduling method and device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1596392A (en) * | 2001-01-05 | 2005-03-16 | 国际商业机器公司 | Computer system having low energy consumption |
| CN1617126A (en) * | 2003-11-11 | 2005-05-18 | 联想(北京)有限公司 | Computer group system and its operation managing method |
-
2009
- 2009-07-08 CN CN2009101087124A patent/CN101604264B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1596392A (en) * | 2001-01-05 | 2005-03-16 | 国际商业机器公司 | Computer system having low energy consumption |
| CN1617126A (en) * | 2003-11-11 | 2005-05-18 | 联想(北京)有限公司 | Computer group system and its operation managing method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101604264A (en) | 2009-12-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101604264B (en) | Task scheduling method and system for supercomputer | |
| CN103365726A (en) | Resource management method and system facing GPU (Graphic Processing Unit) cluster | |
| CN105487930A (en) | Task optimization scheduling method based on Hadoop | |
| CN104050137B (en) | The method and device that kernel is run in a kind of OS | |
| CN100517221C (en) | Efficient multiprocessor system and methods thereof | |
| CN101414270A (en) | Method for implementing assist nuclear task dynamic PRI scheduling with hardware assistant | |
| US20130167152A1 (en) | Multi-core-based computing apparatus having hierarchical scheduler and hierarchical scheduling method | |
| CN102508718A (en) | Method and device for balancing load of virtual machine | |
| CN105786600A (en) | Task scheduling method and device | |
| CN102779075A (en) | Method, device and system for scheduling in multiprocessor nuclear system | |
| CN102707995A (en) | Service scheduling method and device based on cloud computing environments | |
| CN102262567A (en) | Virtual machine scheduling decision system, platform and method | |
| CN105580407A (en) | Network resource deployment method and device | |
| CN102968344A (en) | Method for migration scheduling of multiple virtual machines | |
| CN101103336A (en) | Data processing system and method of task scheduling | |
| CN102999377A (en) | Service concurrent access control method and device | |
| CN102662754A (en) | Multi-field supportable virtual machine dispatching device and multi-field supportable virtual machine dispatching method | |
| CN103019854A (en) | Distributed dynamic process management method and system | |
| JP6081212B2 (en) | Computer system, resource management apparatus, and resource management method | |
| CN113132456A (en) | Edge cloud cooperative task scheduling method and system based on deadline perception | |
| CN114138488A (en) | Cloud-native implementation method and system based on elastic high-performance computing | |
| CN102185726A (en) | Method for improving business resource management capability in information grid system | |
| CN102455932B (en) | Serial execution method, device and system for task instances | |
| CN103744735A (en) | Method and device for scheduling multi-core resource | |
| CN117435324B (en) | Task scheduling method based on containerization |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20170523 Address after: Room office building No. 1068 Shenzhen Institute of advanced technology A-301 518000 in Guangdong city of Shenzhen province Nanshan District Shenzhen University city academy Avenue Patentee after: Shenzhen shen-tech advanced Cci Capital Ltd. Address before: 1068 No. 518055 Guangdong city in Shenzhen Province, Nanshan District City Xili Road School of Shenzhen University Patentee before: SHENZHEN INSTITUTES OF ADVANCED TECHNOLOGY |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20170607 Address after: 4, 8288 building, 66 Shenzhen software Town, Longgang Road, Longgang District, Shenzhen, Guangdong 518000, China Patentee after: SHENZHEN YAOYUAN TECHNOLOGY Co.,Ltd. Address before: Room office building No. 1068 Shenzhen Institute of advanced technology A-301 518000 in Guangdong city of Shenzhen province Nanshan District Shenzhen University city academy Avenue Patentee before: Shenzhen shen-tech advanced Cci Capital Ltd. |
|
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20120725 |