CN104461502A - Task management method and system based on Hadoop - Google Patents
Task management method and system based on Hadoop Download PDFInfo
- Publication number
- CN104461502A CN104461502A CN201410614126.8A CN201410614126A CN104461502A CN 104461502 A CN104461502 A CN 104461502A CN 201410614126 A CN201410614126 A CN 201410614126A CN 104461502 A CN104461502 A CN 104461502A
- Authority
- CN
- China
- Prior art keywords
- task
- current
- subordinate
- current task
- higher level
- 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
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a task management method and system based on Hadoop. The method comprises the steps that a task dependency relationship configured by a user is acquired, a current task is loaded, and upper-level tasks and lower-level tasks of the current task are determined according to the task dependency relationship; the execution results of the upper-level tasks are queried in a task status manager, and if all the upper-level tasks are successfully executed, the current task is triggered to be executed; the execution result of the current task is monitored, and if the current task is successfully executed, the execution result is sent to the task status manager, and the lower-level tasks are triggered to be loaded. By means of the task management method and system, the situations that the tasks are executed independently, and consequently concurrent execution of the tasks is caused are avoided, and the tasks can be uniformly managed. Once the current task is successfully executed, the lower-level tasks are loaded; under the condition that the upper-level tasks are successfully executed, the lower-level tasks can be immediately triggered to be executed. Thus, the running time needed for executing all the tasks can be shortened, and accordingly the execution efficiency of Hadoop can be improved.
Description
[technical field]
The present invention relates to field of cloud computer technology, particularly relate to a kind of task management method based on Hadoop and system.
[background technology]
In recent years, large data platform Hadoop (Distributed Calculation) technology application reaches its maturity.Hadoop provides a distributed cluster system, for large Deta bearer and computing.Hadoop is a technology of increasing income, also constantly ripe and perfect in this technology of community of increasing income, and has reached commercial requirement at present, has been applied simultaneously in various large data platform.
Hadoop is a basic framework platform, provides the API (ApplicationProgramming Interface, application programming interface) of a set of bottom for Hadoop interface interchange.User at control desk, can call Hadoop interface, to finish the work with application program and hiveql (data warehouse) script mode.Calling of Hadoop interface belongs to control desk level or supervisor level usually, and task management gets up difficulty greatly, uses inconvenience.When mutually there is dependence when between task, can only realize single task management by the mode of user program, the script file that easy generation disperses in a large number, need keeper or user to monitor the execution result of each task in real time simultaneously, and according to execution result, triggered by script and perform follow-up work.When task quantity is more, normally by setting the different execution time to each task according to execution sequence, to reach task management effect, but easily like this cause task independently to perform and cause task concurrence performance, easily occur that task result exports entanglement and must re-execute the situation of task, make task cannot unified management, finally affect the execution efficiency of Hadoop.
[summary of the invention]
Based on this, be necessary the problem cannot carrying out unified management for Hadoop in prior art to task, a kind of task management method based on Hadoop and system are provided, unified management can be carried out to the task with dependence.
Based on a task management method of Hadoop, comprise step:
Obtain user configured Task Dependent relation, load current task, and according to higher level's task of described Task Dependent relation determination current task and subordinate's task;
From task status manager, inquire about the execution result of described higher level's task, if all higher level's tasks carrying successes, then trigger and perform current task;
The execution result of monitoring current task, if current task runs succeeded, then sends to task status manager and triggers loading subordinate task by described execution result.
Correspondingly, the present invention also provides a kind of task scheduling system based on Hadoop, comprising:
Acquisition module, for obtaining user configured Task Dependent relation, loads current task, and according to higher level's task of described Task Dependent relation determination current task and subordinate's task;
Task trigger module, for inquiring about the execution result of described higher level's task from task status manager, if all higher level's tasks carrying successes, then triggering and performing current task;
Result monitoring sending module, for monitoring the execution result of current task, if current task runs succeeded, then sends to task status manager and triggers loading subordinate task by described execution result.
First the present invention obtains user configured Task Dependent relation, and according to higher level's task of described Task Dependent relation determination current task and subordinate's task, then from task status manager, inquire about the execution result of described higher level's task, if all higher level's tasks carrying successes, then trigger and perform current task; In the middle of the implementation of current task, the execution result of monitoring current task, if current task runs succeeded, then sends to task status manager and triggers loading subordinate task by described execution result.The present invention obtains higher level's task and subordinate's task of current task according to Task Dependent relation, under the successful prerequisite of higher level's tasks carrying, can trigger execution current task, the task that therefore avoids independently performs and the situation causing task concurrence performance, makes task can unified management.When current task runs succeeded, just load subordinate's task, described subordinate task just can immediately trigger execution when level task runs succeeded thereon, avoid the free time between task, can save whole tasks carrying complete needed for working time of expending, thus improve the execution efficiency of Hadoop.
[accompanying drawing explanation]
Fig. 1 is the process flow diagram of a kind of embodiment of task management method that the present invention is based on Hadoop;
Fig. 2 is the dependence schematic diagram of a kind of embodiment of task management method that the present invention is based on Hadoop;
Fig. 3 is the node relationships net schematic diagram of a kind of embodiment of task management method that the present invention is based on Hadoop;
Fig. 4 is the structured flowchart of a kind of embodiment of task management system that the present invention is based on Hadoop.
[embodiment]
In order to make the object, technical solutions and advantages of the present invention clearly, below in conjunction with accompanying drawing, the present invention is described in further detail.
Refer to Fig. 1, it is the process flow diagram of a kind of embodiment of task management method that the present invention is based on Hadoop.
Based on a task management method of Hadoop, comprise step:
S101: obtain user configured Task Dependent relation, loads current task, and according to higher level's task of described Task Dependent relation determination current task and subordinate's task;
User is by the Task Dependent relation of each task of control desk interface configurations, described dependence comprises starting point task and the executive condition of other each tasks except starting point task, as shown in Figure 2, it is the dependence schematic diagram of a kind of embodiment of task management method that the present invention is based on Hadoop.The dependence task A of task C and task B runs succeeded, and just can execute the task under the prerequisite that is run succeeded at task A and task B C, as long as task A and task B wherein has an execution unsuccessful, then task C does not perform.
Obtain user configured Task Dependent relation, starting point task is determined from described dependence, and subordinate's task of starting point task, start starting point tasks carrying, if there is multiple starting point task, then by starting point task concurrence performance, after starting point tasks carrying, be current task by subordinate's task definition of starting point task, and load this current task, then according to higher level's task and subordinate's task of described Task Dependent relation determination current task.In example above, for task C, task A and task B is exactly higher level's task of task C, is exactly subordinate's task for task A or task B, task C.
S102: the execution result inquiring about described higher level's task from task status manager, if all higher level's tasks carrying successes, then triggers and performs current task;
The task status manager of access for storing each execution status of task, therefrom inquire about and obtain the execution result of higher level's task of current task, under the prerequisite that all higher level's tasks of current task all run succeeded, illustrate that current task meets the executive condition of this task in Task Dependent relation, then trigger performing current task.
S103: the execution result of monitoring current task, if current task runs succeeded, then sends to task status manager and triggers loading subordinate task by described execution result.
In the middle of the implementation of current task, the execution result of Real-Time Monitoring current task, if current task is performing do not obtain result, then in task manager, demarcate execution status of task before deserving is being " performing ".If current task runs succeeded, then this execution result is sent to task status manager, and demarcate this current task executing state for " running succeeded " by task status manager, then subordinate's task of current task is loaded, as long as all higher level's task of described subordinate task all runs succeeded just can trigger perform subordinate's task, the like, until all tasks carryings are complete, or till task termination execution.
First the present invention obtains user configured Task Dependent relation, and according to higher level's task of described Task Dependent relation determination current task and subordinate's task, then from task status manager, inquire about the execution result of described higher level's task, if all higher level's tasks carrying successes, then trigger and perform current task; In the middle of the implementation of current task, the execution result of monitoring current task, if current task runs succeeded, then sends to task status manager and triggers loading subordinate task by described execution result.The present invention obtains higher level's task and subordinate's task of current task according to Task Dependent relation, under the successful prerequisite of higher level's tasks carrying, can trigger execution current task, the task that therefore avoids independently performs and the situation causing task concurrence performance, makes task can unified management.When current task runs succeeded, just load subordinate's task, described subordinate task just can immediately trigger execution when level task runs succeeded thereon, avoid the free time between task, can save whole tasks carrying complete needed for working time of expending, thus improve the execution efficiency of Hadoop.
In one embodiment, in above-mentioned steps S101, according to described higher level's task of Task Dependent relation determination current task and the step of subordinate's task, specifically can comprise the following steps.
S201: generate node relationships net according to described Task Dependent relation, and determine the present node at current task place in described node relationships net, wherein, node mapping task in described node relationships net;
Starting point using starting point task as node relationships net, then obtains the task of the starting point task of higher level task, then using this task as the downward extension of starting point in node relationships net, the like, to generate node relationships net.Such as, a Task Dependent relation describes as follows, task 1, background task 2 and background task 3 based on starting point task, processing tasks 1 relies on background task 1 and runs succeeded, processing tasks 2 relies on background task 2 and background task 3 runs succeeded, business task 1 relies on processing tasks 1 and runs succeeded, and business task 2 relies on processing tasks 1 and processing tasks 2 runs succeeded.The node relationships net generated according to this Task Dependent relation as shown in Figure 3.
After generating node relationships net, determine the position of current task in described node relationships net, and the node of this position is demarcated position present node.
S202: travel through the superior node and downstream site that are connected with described present node respectively up and down, the task that described superior node maps is judged to be higher level's task of current task, the task that described downstream site maps is judged to be subordinate's task of current task.
In node relationships net, take present node as benchmark, upwards travel through the superior node be connected with present node, then the task that described superior node maps is judged to be higher level's task of current task, then travel through the downstream site be connected with present node downwards, then the task that described downstream site maps is judged to be subordinate's task of current task.In figure 3, for processing tasks 2, with this node for the node that benchmark upwards travels through has background task 2 and background task 3, the node of downward traversal has business task 2, therefore background task 2 and background task 3 are judged to be higher level's task of processing tasks 2, business task 2 are judged to be subordinate's task of processing tasks 2.
Node relationships net is generated according to described Task Dependent relation, then higher level's task and subordinate's task of current task is obtained by the mode of traverse node, can be quick, accurately also intactly inquire about higher level's task and subordinate's task of current task, can ensure that Task Dependent relation is strictly followed in the execution of each task, avoid the situation causing task concurrence performance, and make to be more prone to be realized by computer program according to described higher level's task of Task Dependent relation determination current task and the process of subordinate's task, what alleviate this step realizes difficulty, the real-time of enhancing system.
In one embodiment, in above-mentioned steps S102, inquire about the step of the execution result of described higher level's task from task status manager after, can also comprise the following steps further.
S301: if a certain higher level's tasks carrying failure of current task, then stop performing current task, and the failed execution result of current task execution is sent to task status manager.
Inquire about the execution result of described higher level's task from task status manager after, if a certain higher level's tasks carrying failure of current task, then illustrate that current task does not possess the condition that can perform, now stop performing current task, then current task is performed failed execution result and send to task status manager, by task status manager current task is labeled as and performs failure.
When a certain higher level's tasks carrying failure of current task, stop performing current task, current task mistake can be avoided to perform, decrease the possibility that Hadoop mistake is executed the task, thus hardware and thread resources can be distributed running other and meeting executive condition of task by Hadoop, improve the operational efficiency of Hadoop.
In one embodiment, in above-mentioned steps S103, after the step of the execution result of monitoring current task, further comprise step:
S401: if current task performs failure, then stop performing subordinate's task, and execution result current task being performed failed execution result and subordinate's tasks carrying failure sends to task status manager.
When monitoring the execution result of current task, if current task performs failure, then illustrate that subordinate's task of current task does not possess the condition that can perform, now stop performing subordinate's task, then execution result current task being performed failed execution result and subordinate's tasks carrying failure sends to task status manager, is perform failure by task status manager by subordinate's task flagging of current task and current task.
When current task performs unsuccessfully, stop performing subordinate's task, subordinate's task mistake of current task can be avoided to perform, further reduce the possibility that Hadoop mistake is executed the task, improve the operational efficiency of Hadoop further.
Refer to Fig. 4, it is the structured flowchart of a kind of embodiment of task management system that the present invention is based on Hadoop.
Based on a task scheduling system of Hadoop, comprising:
Acquisition module 301, for obtaining user configured Task Dependent relation, loads current task, and according to higher level's task of described Task Dependent relation determination current task and subordinate's task;
User is by the Task Dependent relation of each task of control desk interface configurations, described dependence comprises starting point task and the executive condition of other each tasks except starting point task, as shown in Figure 2, the dependence task A of task C and task B runs succeeded, that is just can execute the task under the prerequisite run succeeded at task A and task B C, as long as task A and task B wherein has an execution unsuccessful, then task C does not perform.
Acquisition module 301 obtains user configured Task Dependent relation, starting point task is determined from described dependence, and subordinate's task of starting point task, start starting point tasks carrying, if there is multiple starting point task, then by starting point task concurrence performance, after starting point tasks carrying, then subordinate's task definition of starting point task is current task by acquisition module 301, and loads this current task.Acquisition module 301 is according to higher level's task of described Task Dependent relation determination current task and subordinate's task afterwards.In example above, for task C, task A and task B is exactly higher level's task of task C, is exactly subordinate's task for task A or task B, task C.
Task trigger module 302, for inquiring about the execution result of described higher level's task from task status manager, if all higher level's tasks carrying successes, then triggering and performing current task;
Task trigger module 302 accesses the task status manager for storing each execution status of task, therefrom inquires about and obtains the execution result of higher level's task of current task.Under the prerequisite that all higher level's tasks of current task all run succeeded, illustrate that current task meets the executive condition of this task in Task Dependent relation, then triggered performing current task by task trigger module 302.
Result monitoring sending module 303, for monitoring the execution result of current task, if current task runs succeeded, then sends to task status manager and triggers loading subordinate task by described execution result.
In the middle of the implementation of current task, the execution result of result monitoring sending module 303 Real-Time Monitoring current task, if current task is performing do not obtain result, then task manager is demarcating this current task executing state for " performing ".If current task runs succeeded, then this execution result is sent to task status manager by result monitoring sending module 303, and demarcate this current task executing state for " running succeeded " by task status manager, then result monitoring sending module 303 loads subordinate's task of current task, as long as all higher level's task of described subordinate task all runs succeeded just can trigger perform subordinate's task, the like, until all tasks carryings are complete, or till task termination execution.
The present invention first acquisition module 301 obtains user configured Task Dependent relation, and according to higher level's task of described Task Dependent relation determination current task and subordinate's task, then task trigger module 302 inquires about the execution result of described higher level's task from task status manager, if all higher level's tasks carrying successes, then trigger and perform current task; In the middle of the implementation of current task, result monitoring sending module 303 monitors the execution result of current task, if current task runs succeeded, then described execution result is sent to task status manager and triggers loading subordinate task.The present invention obtains higher level's task and subordinate's task of current task according to Task Dependent relation by acquisition module 301, under the successful prerequisite of higher level's tasks carrying, execution current task can be triggered, therefore the task that avoids independently performs and the situation causing task concurrence performance, makes task can unified management.When current task runs succeeded, just load subordinate's task, described subordinate task just can immediately trigger execution when level task runs succeeded thereon, avoid the free time between task, can save whole tasks carrying complete needed for working time of expending, thus improve the execution efficiency of Hadoop.
In one embodiment, above-mentioned acquisition module 301 can comprise following submodule.
Generation module, for generating node relationships net according to described Task Dependent relation, and determines the present node at current task place in described node relationships net, wherein, and node mapping task in described node relationships net;
The starting point of generation module using starting point task as node relationships net, then generation module obtains the task of the starting point task of higher level task, then using this task as the downward extension of starting point in node relationships net, the like, to generate node relationships net.Such as, a Task Dependent relation describes as follows, task 1, background task 2 and background task 3 based on starting point task, processing tasks 1 relies on background task 1 and runs succeeded, processing tasks 2 relies on background task 2 and background task 3 runs succeeded, business task 1 relies on processing tasks 1 and runs succeeded, and business task 2 relies on processing tasks 1 and processing tasks 2 runs succeeded.The node relationships net generated according to this Task Dependent relation as shown in Figure 3.
Generation module determines the position of current task in described node relationships net after generating node relationships net, and the node of this position is demarcated position present node.
Traversal determination module, for traveling through the superior node and downstream site that are connected with described present node respectively up and down, the task that described superior node maps is judged to be higher level's task of current task, the task that described downstream site maps is judged to be subordinate's task of current task.
In node relationships net, traversal determination module take present node as benchmark, upwards travel through the superior node be connected with present node, then the task that described superior node maps is judged to be higher level's task of current task, then travel through determination module and travel through the downstream site be connected with present node downwards, afterwards the task that described downstream site maps is judged to be subordinate's task of current task.In figure 3, for processing tasks 2, with this node for the node that benchmark upwards travels through has background task 2 and background task 3, the node of downward traversal has business task 2, therefore travel through higher level's task that background task 2 and background task 3 are judged to be processing tasks 2 by determination module, business task 2 is judged to be subordinate's task of processing tasks 2.
Generation module generates node relationships net according to described Task Dependent relation, then determination module obtains current task higher level's task and subordinate's task by the mode of traverse node is traveled through, can be quick, accurately also intactly inquire about higher level's task and subordinate's task of current task, can ensure that Task Dependent relation is strictly followed in the execution of each task, avoid the situation causing task concurrence performance, and make to be more prone to be realized by computer program according to described higher level's task of Task Dependent relation determination current task and the process of subordinate's task, what alleviate this step realizes difficulty, the real-time of enhancing system.
In one embodiment, can also comprise with lower module further.
First stops module, for when a certain higher level's tasks carrying is failed, stops performing current task, and the failed execution result of current task execution is sent to task status manager.
Inquire about the execution result of described higher level's task from task status manager after, if a certain higher level's tasks carrying failure of current task, then illustrate that current task does not possess the condition that can perform, now stop end-of-module by first and perform current task, then first stop module and current task is performed failed execution result send to task status manager, by task status manager current task is labeled as and performs failure.
When a certain higher level's tasks carrying failure of current task, first stops end-of-module performs current task, current task mistake can be avoided to perform, decrease the possibility that Hadoop mistake is executed the task, thus hardware and thread resources can be distributed running other and meeting executive condition of task by Hadoop, improve the operational efficiency of Hadoop.
In one embodiment, can also comprise with lower module further.
Second stops module, for when current task performs unsuccessfully, stop performing subordinate's task, and execution result current task being performed failed execution result and subordinate's tasks carrying failure sends to task status manager.
When monitoring the execution result of current task, if current task performs failure, then illustrate that subordinate's task of current task does not possess the condition that can perform, now stop end-of-module by second and perform subordinate's task, then second stopping module the execution result that current task performs failed execution result and subordinate's tasks carrying failure is sent to task status manager, is perform failure by task status manager by subordinate's task flagging of current task and current task.
When current task performs unsuccessfully, second stops end-of-module performs subordinate's task, subordinate's task mistake of current task can be avoided to perform, further reduce the possibility that Hadoop mistake is executed the task, improve the operational efficiency of Hadoop further.
The above embodiment only have expressed several embodiment of the present invention, and it describes comparatively concrete and detailed, but therefore can not be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be as the criterion with claims.
Claims (8)
1. based on a task management method of Hadoop, it is characterized in that, comprise step:
Obtain user configured Task Dependent relation, load current task, and according to higher level's task of described Task Dependent relation determination current task and subordinate's task;
From task status manager, inquire about the execution result of described higher level's task, if all higher level's tasks carrying successes, then trigger and perform current task;
The execution result of monitoring current task, if current task runs succeeded, then sends to task status manager and triggers loading subordinate task by described execution result.
2. the task management method based on Hadoop according to claim 1, is characterized in that, according to described higher level's task of Task Dependent relation determination current task and the step of subordinate's task, specifically comprises:
Generate node relationships net according to described Task Dependent relation, and determine the present node at current task place in described node relationships net, wherein, node mapping task in described node relationships net;
Travel through the superior node and downstream site that are connected with described present node respectively up and down, the task that described superior node maps is judged to be higher level's task of current task, the task that described downstream site maps is judged to be subordinate's task of current task.
3. the task management method based on Hadoop according to claim 1, is characterized in that, after inquiring about the step of the execution result of described higher level's task, further comprises step from task status manager:
If a certain higher level's tasks carrying failure of current task, then stop performing current task, and the failed execution result of current task execution is sent to task status manager.
4. the task management method based on Hadoop according to claim 1, is characterized in that, after the step of the execution result of monitoring current task, further comprises step:
If current task performs failure, then stop performing subordinate's task, and execution result current task being performed failed execution result and subordinate's tasks carrying failure sends to task status manager.
5. based on a task scheduling system of Hadoop, it is characterized in that, comprising:
Acquisition module, for obtaining user configured Task Dependent relation, loads current task, and according to higher level's task of described Task Dependent relation determination current task and subordinate's task;
Task trigger module, for inquiring about the execution result of described higher level's task from task status manager, if all higher level's tasks carrying successes, then triggering and performing current task;
Result monitoring sending module, for monitoring the execution result of current task, if current task runs succeeded, then sends to task status manager and triggers loading subordinate task by described execution result.
6. the task management system based on Hadoop according to claim 5, is characterized in that, described acquisition module comprises:
Generation module, for generating node relationships net according to described Task Dependent relation, and determines the present node at current task place in described node relationships net, wherein, and node mapping task in described node relationships net;
Traversal determination module, for traveling through the superior node and downstream site that are connected with described present node respectively up and down, the task that described superior node maps is judged to be higher level's task of current task, the task that described downstream site maps is judged to be subordinate's task of current task.
7. the task management system based on Hadoop according to claim 5, is characterized in that, also comprise:
First stops module, for when a certain higher level's tasks carrying is failed, stops performing current task, and the failed execution result of current task execution is sent to task status manager.
8. the task management system based on Hadoop according to claim 5, is characterized in that, also comprise:
Second stops module, for when current task performs unsuccessfully, stop performing subordinate's task, and execution result current task being performed failed execution result and subordinate's tasks carrying failure sends to task status manager.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410614126.8A CN104461502A (en) | 2014-11-03 | 2014-11-03 | Task management method and system based on Hadoop |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410614126.8A CN104461502A (en) | 2014-11-03 | 2014-11-03 | Task management method and system based on Hadoop |
Publications (1)
Publication Number | Publication Date |
---|---|
CN104461502A true CN104461502A (en) | 2015-03-25 |
Family
ID=52907614
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410614126.8A Pending CN104461502A (en) | 2014-11-03 | 2014-11-03 | Task management method and system based on Hadoop |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104461502A (en) |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104915260A (en) * | 2015-06-19 | 2015-09-16 | 北京搜狐新媒体信息技术有限公司 | Hadoop cluster management task distributing method and system |
CN105045832A (en) * | 2015-06-30 | 2015-11-11 | 北京奇艺世纪科技有限公司 | Data acquisition method and apparatus |
CN105159783A (en) * | 2015-10-09 | 2015-12-16 | 上海瀚之友信息技术服务有限公司 | System task distribution method |
CN106293920A (en) * | 2016-08-15 | 2017-01-04 | 北京票之家科技有限公司 | Method for scheduling task and device |
CN106293971A (en) * | 2016-08-15 | 2017-01-04 | 张家林 | A kind of method and apparatus of distributed task dispatching |
CN106970834A (en) * | 2017-01-20 | 2017-07-21 | 阿里巴巴集团控股有限公司 | The method and device of task processing |
CN107016479A (en) * | 2016-01-28 | 2017-08-04 | 五八同城信息技术有限公司 | Task scheduling and managing method, apparatus and system |
CN107870948A (en) * | 2016-09-28 | 2018-04-03 | 平安科技(深圳)有限公司 | Method for scheduling task and device |
CN108804206A (en) * | 2017-04-26 | 2018-11-13 | 武汉斗鱼网络科技有限公司 | The processing method and system of synchronous task |
CN110390493A (en) * | 2019-07-31 | 2019-10-29 | 北京字节跳动网络技术有限公司 | Task management method, device, storage medium and electronic equipment |
CN112905321A (en) * | 2021-02-07 | 2021-06-04 | 北京红山信息科技研究院有限公司 | Event response type task triggering method and device, electronic equipment and storage medium |
CN113485820A (en) * | 2021-08-03 | 2021-10-08 | 北京百度网讯科技有限公司 | Task scheduling system and implementation method, device and medium thereof |
CN115878272A (en) * | 2021-09-22 | 2023-03-31 | 中科寒武纪科技股份有限公司 | Graph task scheduling method, execution end device, storage medium and program product |
CN115904850A (en) * | 2023-01-09 | 2023-04-04 | 深流微智能科技(深圳)有限公司 | Power-on detection method of multi-core processor, readable storage medium and GPU |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060149786A1 (en) * | 2003-01-21 | 2006-07-06 | Shuhei Nishiyama | Distributed database system |
CN102387208A (en) * | 2011-10-21 | 2012-03-21 | 百度在线网络技术(北京)有限公司 | Distribution type task scheduling method and system |
CN102508716A (en) * | 2011-09-29 | 2012-06-20 | 用友软件股份有限公司 | Task control device and task control method |
CN102508639A (en) * | 2011-10-10 | 2012-06-20 | 北京邮电大学 | Distributed parallel processing method based on satellite remote sensing data characteristics |
CN103530740A (en) * | 2013-10-25 | 2014-01-22 | 金蝶软件(中国)有限公司 | Method and device for carrying out node control on task flow |
-
2014
- 2014-11-03 CN CN201410614126.8A patent/CN104461502A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060149786A1 (en) * | 2003-01-21 | 2006-07-06 | Shuhei Nishiyama | Distributed database system |
CN102508716A (en) * | 2011-09-29 | 2012-06-20 | 用友软件股份有限公司 | Task control device and task control method |
CN102508639A (en) * | 2011-10-10 | 2012-06-20 | 北京邮电大学 | Distributed parallel processing method based on satellite remote sensing data characteristics |
CN102387208A (en) * | 2011-10-21 | 2012-03-21 | 百度在线网络技术(北京)有限公司 | Distribution type task scheduling method and system |
CN103530740A (en) * | 2013-10-25 | 2014-01-22 | 金蝶软件(中国)有限公司 | Method and device for carrying out node control on task flow |
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104915260A (en) * | 2015-06-19 | 2015-09-16 | 北京搜狐新媒体信息技术有限公司 | Hadoop cluster management task distributing method and system |
CN104915260B (en) * | 2015-06-19 | 2018-05-25 | 北京搜狐新媒体信息技术有限公司 | A kind of distribution method and system of Hadoop cluster managements task |
CN105045832A (en) * | 2015-06-30 | 2015-11-11 | 北京奇艺世纪科技有限公司 | Data acquisition method and apparatus |
CN105045832B (en) * | 2015-06-30 | 2018-08-07 | 北京奇艺世纪科技有限公司 | A kind of collecting method and device |
CN105159783A (en) * | 2015-10-09 | 2015-12-16 | 上海瀚之友信息技术服务有限公司 | System task distribution method |
CN107016479A (en) * | 2016-01-28 | 2017-08-04 | 五八同城信息技术有限公司 | Task scheduling and managing method, apparatus and system |
CN106293971A (en) * | 2016-08-15 | 2017-01-04 | 张家林 | A kind of method and apparatus of distributed task dispatching |
CN106293920A (en) * | 2016-08-15 | 2017-01-04 | 北京票之家科技有限公司 | Method for scheduling task and device |
CN107870948A (en) * | 2016-09-28 | 2018-04-03 | 平安科技(深圳)有限公司 | Method for scheduling task and device |
CN106970834A (en) * | 2017-01-20 | 2017-07-21 | 阿里巴巴集团控股有限公司 | The method and device of task processing |
CN108804206A (en) * | 2017-04-26 | 2018-11-13 | 武汉斗鱼网络科技有限公司 | The processing method and system of synchronous task |
CN110390493A (en) * | 2019-07-31 | 2019-10-29 | 北京字节跳动网络技术有限公司 | Task management method, device, storage medium and electronic equipment |
CN110390493B (en) * | 2019-07-31 | 2021-10-22 | 北京字节跳动网络技术有限公司 | Task management method and device, storage medium and electronic equipment |
CN112905321A (en) * | 2021-02-07 | 2021-06-04 | 北京红山信息科技研究院有限公司 | Event response type task triggering method and device, electronic equipment and storage medium |
CN113485820A (en) * | 2021-08-03 | 2021-10-08 | 北京百度网讯科技有限公司 | Task scheduling system and implementation method, device and medium thereof |
CN115878272A (en) * | 2021-09-22 | 2023-03-31 | 中科寒武纪科技股份有限公司 | Graph task scheduling method, execution end device, storage medium and program product |
CN115904850A (en) * | 2023-01-09 | 2023-04-04 | 深流微智能科技(深圳)有限公司 | Power-on detection method of multi-core processor, readable storage medium and GPU |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104461502A (en) | Task management method and system based on Hadoop | |
US10866840B2 (en) | Dependent system optimization for serverless frameworks | |
US11012503B2 (en) | Application deployment engine | |
US10394663B2 (en) | Low impact snapshot database protection in a micro-service environment | |
US9934105B2 (en) | Fault tolerance for complex distributed computing operations | |
US9338061B2 (en) | Open station as a stream analysis operator container | |
US20150081885A1 (en) | Automated event management | |
US20210297504A1 (en) | Software-specific auto scaling | |
CN104428760A (en) | Managing a multitenant cloud service | |
CN105554102A (en) | Elastic expansion method based on container cluster and application system thereof | |
US20120117242A1 (en) | Service linkage system and information processing system | |
CN103605613A (en) | Method and system for dynamically adjusting virtual-machine memory in cloud computing environment | |
US10374914B2 (en) | Rerouting data of a streaming application | |
CN103605574A (en) | Virtual machine resource scheduling method and system for server clusters | |
US9342369B2 (en) | System and method for increasing throughput of a PaaS system | |
US9785507B2 (en) | Restoration of consistent regions within a streaming environment | |
Fan et al. | Agent-based service migration framework in hybrid cloud | |
US20170302588A1 (en) | Method and apparatus for provisioning of resources to support applications and their varying demands | |
CN109960579B (en) | Method and device for adjusting service container | |
US11283697B1 (en) | Scalable real time metrics management | |
CN104410511A (en) | Server management method and system | |
CN110324837B (en) | Network management method and device | |
US20170010948A1 (en) | Monitoring a computing environment | |
Tang et al. | Application centric lifecycle framework in cloud | |
US8645957B2 (en) | Optimized datacenter management by centralized task execution through dependency inversion |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20190215 |
|
AD01 | Patent right deemed abandoned |