CN107885549A - Remove the method and system that process is remained in TORQUE computing cluster calculate nodes - Google Patents
Remove the method and system that process is remained in TORQUE computing cluster calculate nodes Download PDFInfo
- Publication number
- CN107885549A CN107885549A CN201711137327.3A CN201711137327A CN107885549A CN 107885549 A CN107885549 A CN 107885549A CN 201711137327 A CN201711137327 A CN 201711137327A CN 107885549 A CN107885549 A CN 107885549A
- Authority
- CN
- China
- Prior art keywords
- calculate
- user name
- file
- torque computing
- calculate node
- 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
- G06F9/44—Arrangements for executing specific programs
- G06F9/445—Program loading or initiating
- G06F9/44594—Unloading
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Debugging And Monitoring (AREA)
Abstract
The invention belongs to field of computer technology, and in particular to the management of high-performance computer cluster, more particularly to remove the method and system that process is remained in TORQUE computing cluster calculate nodes.The method that process is remained in TORQUE computing cluster calculate nodes is removed, including:All user names are obtained from user name listing file, are stored in user list;The user name of operation task in current all calculate nodes is obtained, is stored in relation array;A file with the entitled filename of calculate node is established for each calculate node, be stored in file does not currently have the user name of tasks carrying in the calculate node;Residual process is deleted for each calculate node;Automatically update user name listing file.The system that process is remained in TORQUE computing cluster calculate nodes is removed, including:First acquisition module;Second acquisition module;File establishes module;Remain process-kill module;Update module.The present invention can fast and accurately remove the residual process in TORQUE computing cluster calculate nodes.
Description
Technical field
The invention belongs to field of computer technology, and in particular to the management of high-performance computer cluster, more particularly to remove
The method and system of process are remained in TORQUE computing cluster calculate nodes.
Background technology
High-performance calculation can be used for numerical analysis and simulated experiment, is a kind of generally acknowledged important research method, is science
The important means of innovation.Currently, high-performance calculation ability is the embodiment of a national overall national strength, is had emphatically to national strategy
The influence wanted.
Lifting computing capability has two ways:When the computing capability of lifting unit, such as lifting CPU dominant frequency, use
Many-core technology uses GPU calculating etc.;Another method is to cross express network together(Such as InfiniBand)Carry out parallel more meters
Calculation machine.At present, building for HPCC often considers above-mentioned two factor, selects a cost performance highest side
Case, software for calculation, result of calculation are all stored in document storage system, to ensure the synchronism of each node data.Job run
When, management node is assigned to task in calculate node and calculated, corresponding in calculate node to produce a calculation procedure.Work as meter
After the completion of calculation, calculation procedure terminates in calculate node, and management node again can not find out the state of task.
In cluster actual motion, it is often found that indivedual calculate nodes have calculation procedure running, but looked into management node
Less than its running status.Such as s-th of calculate node has calculation procedure running, but management node thinks that s-th of node is in
Idle condition.If new task, management node may proceed to submit task to s-th of node, cause the multiple meters of s-th of node
Add journey while run, substantially reduce computational efficiency.Here, the calculation procedure that we can not find out management node, is referred to as remained
Process.The reason for causing residual process has a lot, and most importantly user program is write lack of standardization, next to that user not over
Correct method submits calculating task, it is also possible to is system bug.In a word, residual calculation procedure has a strong impact on normal calculating
Tasks carrying, to dispose.It is badly in need of a kind of method and system for the residual process fast and accurately removed in calculate node.
The content of the invention
Remained it is an object of the present invention to overcome the above mentioned deficiencies, providing and removing in TORQUE computing cluster calculate nodes
The method and system of process, to solve influence of the residual process to calculate node and TORQUE computing clusters, ensure the fast of task
Speed performs.
To achieve these goals, the present invention uses following technical scheme:
A kind of method that process is remained in removing TORQUE computing cluster calculate nodes, comprises the following steps:
Step 1:All user names are obtained from user name listing file, are stored in user list;
Step 2:The user name of operation task in current all calculate nodes is obtained, is stored in relation array;
Step 3:A file with the entitled filename of calculate node is established for each calculate node, is stored in currently at this in file
Calculate node does not have the user name of tasks carrying;
Step 4:Residual process is deleted for each calculate node;
Step 5:Automatically update user name listing file.
Preferably, also include before the step 1:
User name listing file is created for TORQUE computing clusters.
Preferably, also include after the step 4:
Record the situation that each calculate node deletes residual process.
The system that process is remained in TORQUE computing cluster calculate nodes is removed, including:
First acquisition module, for obtaining all user names from user name listing file, it is stored in user list;
Second acquisition module, for obtaining the user name of operation task in current all calculate nodes, it is stored in relation array;
File establishes module, for establishing a file with the entitled filename of calculate node for each calculate node, in file
Deposit does not currently have the user name of tasks carrying in the calculate node;
Process-kill module is remained, for deleting residual process for each calculate node;
Update module, for automatically updating user name listing file.
Preferably, in addition to:
Creation module, for creating user name listing file for TORQUE computing clusters.
Preferably, in addition to:
Logging modle, the situation of residual process is deleted for recording each calculate node.
Compared with prior art, the device have the advantages that:
Then the present invention obtains current institute by obtaining all user names from user name listing file and being stored in user list
There is the user name of operation task in calculate node and be stored in relation array, then establish one for each calculate node to calculate
The file of the entitled filename of node, be stored in file does not currently have the user name of tasks carrying in the calculate node, finally to be every
Individual calculate node deletes residual process, automatically updates the mode of user name listing file, deletes each TORQUE computing clusters and respectively counts
The residual process of operator node, influence of the residual process to calculating task is eliminated with this so that calculating task is within the predetermined time
Complete, the specification use of cluster.
Brief description of the drawings
Fig. 1 is the basic procedure schematic diagram that the present invention removes the method that process is remained in TORQUE computing cluster calculate nodes
One of.
Fig. 2 is the basic procedure schematic diagram that the present invention removes the method that process is remained in TORQUE computing cluster calculate nodes
Two.
Fig. 3 be the present invention remove TORQUE computing cluster calculate nodes in remain process system structural representation it
One.
Fig. 4 be the present invention remove TORQUE computing cluster calculate nodes in remain process system structural representation it
Two.
Embodiment
In order to make it easy to understand, explanation explained below is made to the part noun occurred in the embodiment of the present invention:
TORQUE computing clusters:A kind of common HPCC job management system.
Management node:For managing and distributing the node of computing resource, general cluster only has one.
Calculate node:It is mainly used in the node calculated in the cluster, general cluster there are many.
Residual process:The calculation procedure that can not find out in management node.
Below in conjunction with the accompanying drawings with specific embodiment the present invention will be further explained explanation:
Embodiment one:
As shown in figure 1, the method for process is remained in a kind of removing TORQUE computing cluster calculate nodes of the present invention, including it is following
Step:
Step S101:All user names are obtained from user name listing file, are stored in user list;
Step S102:The user name of operation task in current all calculate nodes is obtained, is stored in relation array;
Step S103:A file with the entitled filename of calculate node is established for each calculate node, is stored in file current
There is no the user name of tasks carrying in the calculate node;
Step S104:Residual process is deleted for each calculate node;
Step S105:Automatically update user name listing file.
Embodiment two:
As shown in Fig. 2 another method for removing residual process in TORQUE computing cluster calculate nodes of the present invention, including with
Lower step:
Step S201:User name listing file user.list is created for TORQUE computing clusters.
Step S202:All user names, deposit user list uli are obtained from user name listing file user.list;
For example, it is assumed that there are 10 users in system:
Uli=[user1, user2, user3, user4, user5, user6, user7, user8, user9, user10];
Step S203:The user name situation of operation task in current all calculate nodes is obtained, is stored in relation array idic, form
For idic [node name]=[user list];
For example, it is assumed that TORQUE computing clusters there are 16 calculate nodes, node 1, node 2 ... node 16 are respectively designated as, with
Exemplified by node 1, node 2, node 16:
Idic [node1]=[user1, user3, user8];
Idic [node2]=[user1, user4];
Idic [node16]=[user5, user6].
Step S204:A file with the entitled filename of calculate node is established for each calculate node, is stored in file
Currently there is no the user name of tasks carrying in the calculate node;
Such as:
It is stored in node1 files:User2, user4, user5, user6, user7, user9, user10;
It is stored in node2 files:User2, user3, user5, user6, user7, user8, user9, user10;
It is stored in node16 file:User1, user2, user3, user4, user7, user8, user9, user10.
Step S205:Currently there is no the user name of tasks carrying in the calculate node according in each calculate node file,
Residual process is deleted for each calculate node;
Such as:
User2, user4, user5, user6, user7, user9, user10 all processes are deleted under node1;
User2, user3, user5, user6, user7, user8, user9, user10 all processes are deleted under node2;
User1, user2, user3, user4, user7, user8, user9, all processes of user10 are deleted under node16.
Step S206:Record the situation that each calculate node deletes residual process, each calculate node is deleted remain into
The situation write-in now.list of journey, until deleting all residual processes.
Step S207:If keeper newly with the addition of user, and the user is in and performs task status, then automatically updates
User name listing file user.list.
Embodiment three:
As shown in figure 3, the system of process is remained in a kind of removing TORQUE computing cluster calculate nodes of the present invention, including:
First acquisition module 301, for obtaining all user names from user name listing file, it is stored in user list;
Second acquisition module 302, for obtaining the user name of operation task in current all calculate nodes, it is stored in relation array;
File establishes module 303, for establishing a file with the entitled filename of calculate node, file for each calculate node
In be stored in and currently there is no the user name of tasks carrying in the calculate node;
Process-kill module 304 is remained, for deleting residual process for each calculate node;
Update module 305, for automatically updating user name listing file.
Example IV:
As shown in figure 4, another system for removing residual process in TORQUE computing cluster calculate nodes of the present invention, including:
Creation module 401, for creating user name listing file for TORQUE computing clusters.
First acquisition module 402, for obtaining all user names from user name listing file, it is stored in user list;
Second acquisition module 403, for obtaining the user name of operation task in current all calculate nodes, it is stored in relation array;
File establishes module 404, for establishing a file with the entitled filename of calculate node, file for each calculate node
In be stored in and currently there is no the user name of tasks carrying in the calculate node;
Process-kill module 405 is remained, for deleting residual process for each calculate node;
Logging modle 406, the situation of residual process is deleted for recording each calculate node.
Update module 407, for automatically updating user name listing file.
Illustrated above is only the preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, under the premise without departing from the principles of the invention, some improvements and modifications can also be made, these improvements and modifications also should
It is considered as protection scope of the present invention.
Claims (6)
1. the method for process is remained in a kind of removing TORQUE computing cluster calculate nodes, it is characterised in that comprise the following steps:
Step 1:All user names are obtained from user name listing file, are stored in user list;
Step 2:The user name of operation task in current all calculate nodes is obtained, is stored in relation array;
Step 3:A file with the entitled filename of calculate node is established for each calculate node, is stored in currently at this in file
Calculate node does not have the user name of tasks carrying;
Step 4:Residual process is deleted for each calculate node;
Step 5:Automatically update user name listing file.
2. remaining the method for process in removing TORQUE computing cluster calculate nodes according to claim 1, its feature exists
In also including before the step 1:
User name listing file is created for TORQUE computing clusters.
3. remaining the method for process in removing TORQUE computing cluster calculate nodes according to claim 1, its feature exists
In also including after the step 4:
Record the situation that each calculate node deletes residual process.
4. based on the clear of the method that process is remained in any described removing TORQUE computing cluster calculate nodes of claim 1-3
Except the system that process is remained in TORQUE computing cluster calculate nodes, it is characterised in that including:
First acquisition module, for obtaining all user names from user name listing file, it is stored in user list;
Second acquisition module, for obtaining the user name of operation task in current all calculate nodes, it is stored in relation array;
File establishes module, for establishing a file with the entitled filename of calculate node for each calculate node, in file
Deposit does not currently have the user name of tasks carrying in the calculate node;
Process-kill module is remained, for deleting residual process for each calculate node;
Update module, for automatically updating user name listing file.
5. remaining the system of process in removing TORQUE computing cluster calculate nodes according to claim 4, its feature exists
In, in addition to:
Creation module, for creating user name listing file for TORQUE computing clusters.
6. remaining the system of process in removing TORQUE computing cluster calculate nodes according to claim 4, its feature exists
In, in addition to:
Logging modle, the situation of residual process is deleted for recording each calculate node.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711137327.3A CN107885549B (en) | 2017-11-16 | 2017-11-16 | Method and system for clearing residual process in TORQUE computing cluster computing node |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201711137327.3A CN107885549B (en) | 2017-11-16 | 2017-11-16 | Method and system for clearing residual process in TORQUE computing cluster computing node |
Publications (2)
Publication Number | Publication Date |
---|---|
CN107885549A true CN107885549A (en) | 2018-04-06 |
CN107885549B CN107885549B (en) | 2020-10-23 |
Family
ID=61777060
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201711137327.3A Active CN107885549B (en) | 2017-11-16 | 2017-11-16 | Method and system for clearing residual process in TORQUE computing cluster computing node |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107885549B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112463739A (en) * | 2019-09-09 | 2021-03-09 | 山东省计算中心(国家超级计算济南中心) | Data processing method and system based on ocean mode ROMS |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1391386A (en) * | 2001-06-12 | 2003-01-15 | 华为技术有限公司 | Method for protecting task process in multitask operating system |
US20150222506A1 (en) * | 2014-01-31 | 2015-08-06 | Apollo Education Group, Inc. | Mechanism for controlling a process on a computing node based on the participation status of the computing node |
CN105022666A (en) * | 2014-04-24 | 2015-11-04 | 中国电信股份有限公司 | Method, device and system for controlling MapReduce task scheduling |
CN106681753A (en) * | 2016-11-01 | 2017-05-17 | 腾讯科技(深圳)有限公司 | Software residual process processing method and device |
CN107066879A (en) * | 2017-03-22 | 2017-08-18 | 山东中创软件商用中间件股份有限公司 | A kind of method and system hidden for computer application program process |
-
2017
- 2017-11-16 CN CN201711137327.3A patent/CN107885549B/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1391386A (en) * | 2001-06-12 | 2003-01-15 | 华为技术有限公司 | Method for protecting task process in multitask operating system |
US20150222506A1 (en) * | 2014-01-31 | 2015-08-06 | Apollo Education Group, Inc. | Mechanism for controlling a process on a computing node based on the participation status of the computing node |
CN105022666A (en) * | 2014-04-24 | 2015-11-04 | 中国电信股份有限公司 | Method, device and system for controlling MapReduce task scheduling |
CN106681753A (en) * | 2016-11-01 | 2017-05-17 | 腾讯科技(深圳)有限公司 | Software residual process processing method and device |
CN107066879A (en) * | 2017-03-22 | 2017-08-18 | 山东中创软件商用中间件股份有限公司 | A kind of method and system hidden for computer application program process |
Non-Patent Citations (1)
Title |
---|
朱军等: "节点删除法的虚拟网络映射算法", 《安徽大学学报(自然科学版)》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112463739A (en) * | 2019-09-09 | 2021-03-09 | 山东省计算中心(国家超级计算济南中心) | Data processing method and system based on ocean mode ROMS |
Also Published As
Publication number | Publication date |
---|---|
CN107885549B (en) | 2020-10-23 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104618693B (en) | A kind of monitor video based on cloud computing handles task management method and system online | |
WO2020259717A1 (en) | Method, apparatus, and terminal device for controlling federated learning system, and storage medium | |
US9798774B1 (en) | Graph data search method and apparatus | |
Lo'ai et al. | A mobile cloud computing model using the cloudlet scheme for big data applications | |
US20160048415A1 (en) | Systems and Methods for Auto-Scaling a Big Data System | |
CN101887367B (en) | Multi-level parallel programming method | |
CN104731595A (en) | Big-data-analysis-oriented mixing computing system | |
CN102739785B (en) | Method for scheduling cloud computing tasks based on network bandwidth estimation | |
CN103902593A (en) | Data transfer method and device | |
CN106302640A (en) | Data request processing method and device | |
CN108121511A (en) | Data processing method, device and equipment in a kind of distributed edge storage system | |
WO2017148297A1 (en) | Method and device for joining tables | |
WO2015051685A1 (en) | Task scheduling method, device and system | |
CN107357873A (en) | A kind of big data storage management system | |
WO2021082928A1 (en) | Data reduction method and apparatus, computing device, and storage medium | |
CN106487601A (en) | Resource monitoring method, apparatus and system | |
CN109618308A (en) | A method of internet of things data is handled based on Spark Streaming | |
WO2015196885A1 (en) | Method and apparatus for acquiring and storing performance data of cloud computing system | |
CN107885549A (en) | Remove the method and system that process is remained in TORQUE computing cluster calculate nodes | |
CN107273527A (en) | A kind of Hadoop clusters and distributed system | |
CN106230623A (en) | A kind of VIM site selection method and device | |
CN104361000A (en) | Access method and system for relational database based on self-built database connection pool | |
CN106257447A (en) | The video storage of cloud storage server and search method, video cloud storage system | |
WO2017113865A1 (en) | Method and device for big data increment calculation | |
CN104580498B (en) | A kind of adaptive cloud management platform |
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 |