CN114237858A - Task scheduling method and system based on multi-cluster network - Google Patents
Task scheduling method and system based on multi-cluster network Download PDFInfo
- Publication number
- CN114237858A CN114237858A CN202210162360.6A CN202210162360A CN114237858A CN 114237858 A CN114237858 A CN 114237858A CN 202210162360 A CN202210162360 A CN 202210162360A CN 114237858 A CN114237858 A CN 114237858A
- Authority
- CN
- China
- Prior art keywords
- task
- cluster
- information
- scheduler
- network
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 32
- 230000001419 dependent effect Effects 0.000 claims abstract description 18
- 238000013500 data storage Methods 0.000 claims description 15
- 230000005540 biological transmission Effects 0.000 claims description 4
- 238000004891 communication Methods 0.000 abstract description 7
- 230000008569 process Effects 0.000 description 8
- 230000008878 coupling Effects 0.000 description 4
- 238000010168 coupling process Methods 0.000 description 4
- 238000005859 coupling reaction Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/48—Program initiating; Program switching, e.g. by interrupt
- G06F9/4806—Task transfer initiation or dispatching
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5061—Partitioning or combining of resources
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer And Data Communications (AREA)
Abstract
The utility model provides a task scheduling method and system based on multi-cluster network, the workflow server sends task information of tasks to the global scheduler in turn according to the dependency relationship among the tasks, and ensures that each task can be executed in turn, the global scheduler distributes each task to the local scheduler of the cluster which can meet the respective resource demand according to the task information of each task and the resource condition of each cluster in the multi-cluster network for scheduling execution, the resource on each cluster can be utilized efficiently, the task distributed on each cluster can be executed as far as possible, and the utilization rate of communication resources is improved; in addition, for each task, when the task is determined to have no dependent task or the dependent task of the task is executed completely, the task information of the task is sent to the global scheduler, so that the task sent to the global scheduler can be effectively executed, the waiting time for executing each task is reduced, and the task execution efficiency is improved.
Description
Technical Field
The present application relates to the field of communications technologies, and in particular, to a task scheduling method and system based on a multi-cluster network.
Background
In the conventional task scheduling system, most of the tasks do not provide the task dependent function, or even if the task dependent function is provided, there is no reasonable solution for how to efficiently utilize the resources on the plurality of clusters when each task is implemented.
Disclosure of Invention
An object of the embodiments of the present application is to provide a task scheduling method and system based on a multi-cluster network, so as to solve the problem that resources on multiple clusters cannot be efficiently utilized in a task scheduling system of an existing multi-cluster network.
The embodiment of the application provides a task scheduling method based on a multi-cluster network, which comprises the following steps:
the workflow server receives a workflow scheduling request;
the workflow server determines a dependency relationship between tasks to be executed and task information of each task according to the workflow scheduling request, and sequentially sends the task information of each task to a global scheduler according to the dependency relationship; the task information comprises resource demand information of resources required by executing the task;
the global scheduler distributes each task to a local scheduler of a cluster which can meet the respective resource requirement according to the task information of each task and the resource condition of each cluster in the multi-cluster network;
and each local scheduler distributes each received task to a corresponding node to execute.
In the implementation process, the workflow server sequentially sends the task information of each task to the global scheduler according to the dependency relationship among the tasks to ensure that each task can be sequentially executed, and the global scheduler distributes each task to the local scheduler of the cluster which can meet the resource requirement of each task according to the task information of each task and the resource condition of each cluster in the multi-cluster network to perform scheduling, so that the resources on each cluster can be efficiently utilized, the tasks distributed on each cluster can be ensured to be executed as much as possible, and the utilization rate of communication resources is improved.
Further, the sequentially sending task information of each task to a global scheduler according to the dependency relationship includes:
for each task, when the task is determined to meet one of preset task dependence conditions, sending task information of the task to the global scheduler; the preset task dependence conditions comprise:
the first condition is as follows: the task has no dependent task;
and a second condition: the dependent task of the task is completed.
In the implementation process, for each task, when it is determined that the task does not have a dependent task or the dependent task of the task is completely executed, the task information of the task is sent to the global scheduler, so that the task sent to the global scheduler can be effectively executed, the waiting time for executing each task is reduced, and the task execution efficiency is improved.
Further, the multi-cluster network is a multi-cluster computational power network, and the task is a computational task.
In the implementation process, the workflow scheduling in the multi-cluster computational power network is supported, so that complex large-scale cross-cluster computing tasks can be conveniently completed.
Further, the resource requirement information of each task further includes: and data storage address information of data required for executing the task.
Further, the task information of each task further includes: storing address information of an execution result of the task execution result; and each node in each local scheduler is used for storing the task execution result in a corresponding position according to the execution result storage address information.
Further, the distributing, by the global scheduler, each of the tasks to the local scheduler of the cluster that can meet the resource requirement of each of the tasks according to the task information of each of the tasks and the resource condition of each of the clusters in the multi-cluster network includes:
the global scheduler determines a target cluster where data required by execution of each task is located according to the received data storage address information corresponding to each task;
and when the resource condition of the target cluster corresponding to a certain task is judged to meet the resource requirement of the task, distributing the task to the local scheduler of the target cluster.
Further, the data storage address information is address index information, and the global scheduler determines, according to the received data storage address information corresponding to each of the tasks, a target cluster in which data required for executing each of the tasks is located, including:
the global scheduler inquires a target cluster corresponding to each address index information from a distributed database according to the received address index information corresponding to each task; the distributed database stores the corresponding relation between the address index information of each data and the cluster storing the data.
In the implementation process, the storage and use requirements of the multi-cluster network on the data are met by establishing the corresponding relation between the data address index and the cluster where the data is located in the distributed database.
Further, the method further comprises:
when the resource condition of a target cluster corresponding to a certain task is judged not to meet the resource requirement of the task, distributing the task to a local scheduler of an optimal candidate cluster of the task; the optimal candidate cluster of the task is the cluster which has the fastest data transmission rate with the target cluster of the task and the resource condition of which meets the resource requirement of the task in the multi-cluster network.
In the implementation process, the tasks are distributed to the optimal candidate clusters to optimize the cost of data access.
Further, the method further comprises:
and each node pulls a mirror image from a public mirror image warehouse, executes the received task according to the mirror image, and each cluster shares a task execution algorithm through the mirror image in the public mirror image warehouse.
The embodiment of the present application further provides a task scheduling system based on a multi-cluster network, including:
the workflow server is used for receiving a workflow scheduling request, determining the dependency relationship among tasks to be executed and the task information of each task according to the workflow scheduling request, and sequentially sending the task information of each task to the global scheduler according to the dependency relationship; the task information comprises resource demand information of resources required by executing the task;
the global scheduler is used for distributing each task to the local scheduler of the cluster which can meet the respective resource requirement according to the task information of each task and the resource condition of each cluster in the multi-cluster network;
and the local scheduler is used for distributing each received task to the corresponding node for execution.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are required to be used in the embodiments of the present application will be briefly described below, it should be understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered as limiting the scope, and that those skilled in the art can also obtain other related drawings based on the drawings without inventive efforts.
Fig. 1 is a schematic flowchart of a task scheduling method based on a multi-cluster network according to an embodiment of the present application;
FIG. 2 is a schematic diagram of dependencies between tasks according to an embodiment of the present disclosure;
fig. 3 is a schematic diagram of an information structure of task information corresponding to a task flow according to an embodiment of the present application;
fig. 4 is a schematic structural diagram of a task scheduling system based on a multi-cluster network according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be described below with reference to the drawings in the embodiments of the present application.
An embodiment of the present application provides a task scheduling method based on a multi-cluster network, please refer to fig. 1, which may include the following steps:
s101: the workflow server receives a workflow scheduling request.
The workflow in this embodiment is composed of a plurality of tasks having dependency relationships, and each task constituting the workflow needs to be executed according to the dependency relationships.
S102: and the workflow server determines the dependency relationship among the tasks to be executed and the task information of each task according to the workflow scheduling request, and sequentially sends the task information of each task to the global scheduler according to the dependency relationship.
It is understood that the workflow server may receive a workflow scheduling request submitted by a user through a terminal, and then analyze the workflow scheduling request to determine the dependency relationship between tasks that need to be executed.
It should be noted that, in this embodiment, task information of each task may be sent to the global scheduler in a jobb form, the jobb in this embodiment provides an extensible mechanism to describe the task, information included in the mechanism corresponds to specific content of the task information in this embodiment, and in this embodiment, the task information of each task may include, but is not limited to, at least one of the following content:
(1) resource requirement information of resources required to execute the task.
(2) A task execution method.
The resource requirement information of the resource required for executing the task may include, but is not limited to, hardware resources, such as CPU, GPU, memory, disk space, and the like, and may also include algorithm resources, data resources, and the like. The data resources may directly include specific contents of the data, but in general, the specific contents of the data may be too huge, the terminal may not normally directly send the data contents to the workflow server, and the workflow server may analyze the received workflow scheduling request to extract data storage address information of the data required to execute each task. Of course, the data storage address information may directly indicate cluster information of a cluster where the data is located, or may be address index information, which will be described below.
For each task to be executed, when it is determined that the task meets one of preset task dependency conditions, sending task information of the task to a global scheduler, where the preset task dependency conditions may include:
the first condition is as follows: the task has no dependent task;
and a second condition: the dependent task of the task is completed.
It should be noted that, when it is determined that a certain task does not satisfy any of the preset task dependency conditions, the task information of the task may not be sent to the global scheduler, and even if the task information of the task is sent to the global scheduler at this time, because the task dependent on the task is not executed completely, the task cannot be executed smoothly, the task information of the task may not be sent to the global scheduler for the task that does not satisfy the preset task dependency conditions at present, so as to achieve the purpose of saving communication resources. Preferably, the task information of the task without dependency on the task is sent to the global scheduler, and after the task is executed, the next nested task of the task is sent to the global scheduler.
For ease of understanding, a specific example is described herein.
It is assumed that the workflow in this embodiment is composed of 7 tasks, and the dependency relationship between the tasks is shown in fig. 2, where inputs and outputs of the 7 tasks are connected to form a directed acyclic graph, and no data interaction occurs between two unconnected tasks, and the two unconnected tasks can run independently from each other, in fig. 2, j1 is a dependent task of j2 and j3, j2 is a dependent task of j4 and j5, and so on, and details are not described here.
The execution process of the workflow in fig. 2 is briefly described here, because j1 has no dependent item, the task information of j1 is sent to the global scheduler first, j1 is executed completely, the dependent tasks representing j2 and j3 are executed completely, the task information of j2 and j3 is sent to the global scheduler, after j2 is executed completely, the task information of j4 is sent to the global scheduler, after j2 and j3 are executed completely, the task information of j5 is sent to the global scheduler, if j5 fails to execute, the task information of j6 and j7 may not be sent to the global scheduler temporarily, and when it is detected that j5 is executed successfully, the task information of j6 and j7 is sent to the global scheduler again.
S103: and the global scheduler distributes each task to the local scheduler of the cluster which can meet the resource requirement of each task according to the task information of each task and the resource condition of each cluster in the multi-cluster network.
In this embodiment, the meeting of the resource requirement of a certain task by the resource condition of the cluster may include: the computing resources of the cluster meet the computing resource requirements of the task, and the data stored on the cluster meet the data requirements of the task.
The global scheduler in this embodiment allocates clusters to each task according to the resource requirements of each task, and each task may enter the same cluster or different clusters.
Specifically, the global scheduler may determine, according to the received data storage address information corresponding to each task, a target cluster where data required for executing each task is located; when the resource condition of a target cluster corresponding to a certain task is judged to meet the resource requirement of the task, the task is distributed to a local scheduler of the target cluster.
The data storage address information in this embodiment may be address index information, and the global scheduler may query a target cluster corresponding to each address index information from the distributed database according to the received address index information corresponding to each task; the distributed database stores the corresponding relation between the address index information of each data and the cluster storing the data.
The distributed database in this embodiment is used to provide a global data storage service, and the storage and use requirements of the multi-cluster network on data are met by establishing a data address index and a corresponding relationship between clusters where the data are located in the distributed database.
It should be noted that when it is determined that the resource condition of the target cluster corresponding to a certain task does not meet the resource requirement of the task, the task may be distributed to the local scheduler of the optimal candidate cluster of the task; the optimal candidate cluster of the task is the cluster which has the fastest data transmission rate with the target cluster of the task and the resource condition of which meets the resource requirement of the task in the multi-cluster network.
In this embodiment, when determining the cluster that needs to be entered for each task, each corresponding cluster may download the data required for executing the task to the memory in advance, so that the local scheduler of each cluster can conveniently retrieve the data, and the task execution time is reduced.
S104: and each local scheduler distributes each received task to a corresponding node to execute.
It should be noted that, in this embodiment, the local scheduler in each cluster may distribute each task to a specific node in the cluster according to a preset local scheduling policy for execution. In this embodiment, the task information of each task may further include: the execution result of the task execution result stores address information.
In step S104, each node in each local scheduler may store the task execution result at a corresponding location according to the execution result storage address information. Optionally, the task in this embodiment may be a model training calculation task, the execution result storage address information is used to indicate a storage address of a training model, and after each node performs model training to obtain a training model, each training model may be stored in a corresponding position in the training model database according to the execution result storage address information, so as to facilitate subsequent calling of each training model.
In this embodiment, each node may pull a mirror from the common mirror repository, execute the received task according to the mirror, and each cluster may share a task execution algorithm through the mirror in the common mirror repository.
To facilitate understanding of the solution provided by the present embodiment, the following description is continued with a specific example. Referring to fig. 3, fig. 3 is an example of task information of each task constituting a workflow, where the workflow includes 2 jobs, that is, tasks, named data-transform and model-training respectively, where an output of the data-transform is an input of the model-training, and thus the task data-transform is a dependent task of the task model-training, the workflow server first sends the task information of the task data-transform to the global scheduler, and after the task data-transform is executed, sends the task information of the task model-training to the global scheduler, and the specific content of the task information of the jobs is described by taking the model-training task as an example:
the input field is the data source of the job, and corresponds to the data storage address information described above.
The output field corresponds to the data storage address information described above.
"TensorFlow" in the type field indicates that the task is a task using the TensorFlow framework.
the tasks field defines the execution mode of the task:
two nodes, i.e., worker (executor) and PS (parameter server), are required, of which there are 4 workers and 1 PS.
image and command represent the mirror used and the commands executed by the job, respectively.
(3) resources represents the job's demand for resources.
After receiving task information of a model-tracing task, a global scheduler can query information of a target cluster where data is located from a distributed database according to the information in an input, if resources of the target cluster meet requirements of the model-tracing task, the model-tracing task is allocated to the target cluster, otherwise, the model-tracing task is allocated to a candidate cluster which meets the requirements of the resources and has the highest data transmission speed with the target cluster. After receiving the task information of the task, the cluster that allocates the model-tracing task distributes the task to each node by the local scheduler, it should be noted that after being allocated with the task, the cluster can download corresponding data to the memory, so as to improve the efficiency of executing subsequent tasks. And each node in the local scheduler pulls a mirror image from the container mirror image warehouse and starts to execute a computing task, and after the execution of the model-tracing task is finished, the trained model is uploaded to a corresponding position according to the output field of the joba.
Referring to fig. 4, an embodiment of the present application further provides a task scheduling system based on a multi-cluster network, including:
the workflow server is used for receiving a workflow scheduling request, determining the dependency relationship among tasks to be executed and the task information of each task according to the workflow scheduling request, and sequentially sending the task information of each task to the global scheduler according to the dependency relationship; the task information comprises resource demand information of resources required by executing the task;
the global scheduler is used for distributing each task to the local scheduler of the cluster which can meet the respective resource requirement according to the task information of each task and the resource condition of each cluster in the multi-cluster network;
and the local scheduler is used for distributing each received task to the corresponding node for execution.
It should be noted that, in this embodiment, the definition manner of each cluster for the task may be the same, that is, specific contents of the task information that each cluster can process the task may include the aforementioned input field, output field, type field, and tasks field, and for other clusters outside the system, if the task information of the task that can be processed by the local scheduler of another cluster is the same as the definition format of the task information of the job in this embodiment, that is, if the task information of the task that can be processed by the local scheduler of another cluster includes the aforementioned input field, output field, type field, and tasks field, the other cluster may be added to the system. Therefore, for the task scheduling system based on the multi-cluster network provided in this embodiment, when it is detected that the task information of the task in the local scheduler in some other cluster except the local scheduling system is the same as the task information format of the task in the local scheduling system, the other cluster may be added to the local scheduling system.
By the multi-cluster-network-based task scheduling method and system provided by the embodiment, the calculation tasks in the computational power network can be abstracted into a jobobject, jobdefinition is unified, and a concise expression form is provided to express task information, for example, a YAML file with the size of dozens to hundreds of lines can be provided, so that the task information can be conveniently, quickly and inexpensively transmitted in the network; in addition, a multilayer scheduling structure is adopted, the coupling between the local scheduler and the global scheduler is low, and the original cluster scheduler can be directly reused in the process of adding the computing node into the computational power network; the task scheduling can be combined with various global services, and the computing tasks are supported to run and migrate in the computing network efficiently; and the workflow in the computational network is supported, so that the complex large-scale, cross-cluster and cross-geographic-region computing tasks can be conveniently completed.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units is only one logical division, and there may be other divisions when actually implemented, and for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection of devices or units through some communication interfaces, and may be in an electrical, mechanical or other form.
In addition, units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In this document, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions.
The above description is only an example of the present application and is not intended to limit the scope of the present application, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present application shall be included in the protection scope of the present application.
Claims (10)
1. A task scheduling method based on a multi-cluster network is characterized by comprising the following steps:
the workflow server receives a workflow scheduling request;
the workflow server determines a dependency relationship between tasks to be executed and task information of each task according to the workflow scheduling request, and sequentially sends the task information of each task to a global scheduler according to the dependency relationship; the task information comprises resource demand information of resources required by executing the task;
the global scheduler distributes each task to a local scheduler of a cluster which can meet the respective resource requirement according to the task information of each task and the resource condition of each cluster in the multi-cluster network;
and each local scheduler distributes each received task to a corresponding node to execute.
2. The method for task scheduling based on multi-cluster network as claimed in claim 1, wherein said sending task information of each of said tasks to a global scheduler in sequence according to said dependency relationship comprises:
for each task, when the task is determined to meet one of preset task dependence conditions, sending task information of the task to the global scheduler; the preset task dependence conditions comprise:
the first condition is as follows: the task has no dependent task;
and a second condition: the dependent task of the task is completed.
3. The multi-cluster-network-based task scheduling method of claim 1, wherein the multi-cluster network is a multi-cluster computational power network and the task is a computational task.
4. The method for task scheduling based on multi-cluster network as claimed in claim 1, wherein the resource requirement information of each task comprises: and data storage address information of data required for executing the task.
5. The multi-cluster-network-based task scheduling method of claim 4, wherein the task information of each task further comprises: storing address information of an execution result of the task execution result; and each node in each local scheduler is used for storing the task execution result in a corresponding position according to the execution result storage address information.
6. The method as claimed in claim 4, wherein the step of the global scheduler distributing each task to the local scheduler of the cluster that can meet the respective resource requirement according to the task information of each task and the resource condition of each cluster in the multi-cluster network comprises:
the global scheduler determines a target cluster where data required by execution of each task is located according to the received data storage address information corresponding to each task;
and when the resource condition of the target cluster corresponding to a certain task is judged to meet the resource requirement of the task, distributing the task to the local scheduler of the target cluster.
7. The method as claimed in claim 6, wherein the data storage address information is address index information, and the determining, by the global scheduler, a target cluster where data required for executing each of the tasks is located according to the received data storage address information corresponding to each of the tasks includes:
the global scheduler inquires a target cluster corresponding to each address index information from a distributed database according to the received address index information corresponding to each task; the distributed database stores the corresponding relation between the address index information of each data and the cluster storing the data.
8. The multi-cluster-network-based task scheduling method of claim 6, wherein the method further comprises:
when the resource condition of a target cluster corresponding to a certain task is judged not to meet the resource requirement of the task, distributing the task to a local scheduler of an optimal candidate cluster of the task; the optimal candidate cluster of the task is the cluster which has the fastest data transmission rate with the target cluster of the task and the resource condition of which meets the resource requirement of the task in the multi-cluster network.
9. The multi-cluster-network-based task scheduling method of claim 1, wherein the method further comprises:
and each node pulls a mirror image from a public mirror image warehouse, executes the received task according to the mirror image, and each cluster shares a task execution algorithm through the mirror image in the public mirror image warehouse.
10. A task scheduling system based on a multi-cluster network, comprising:
the workflow server is used for receiving a workflow scheduling request, determining the dependency relationship among tasks to be executed and the task information of each task according to the workflow scheduling request, and sequentially sending the task information of each task to the global scheduler according to the dependency relationship; the task information comprises resource demand information of resources required by executing the task;
the global scheduler is used for distributing each task to the local scheduler of the cluster which can meet the respective resource requirement according to the task information of each task and the resource condition of each cluster in the multi-cluster network;
and the local scheduler is used for distributing each received task to the corresponding node for execution.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210162360.6A CN114237858A (en) | 2022-02-22 | 2022-02-22 | Task scheduling method and system based on multi-cluster network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210162360.6A CN114237858A (en) | 2022-02-22 | 2022-02-22 | Task scheduling method and system based on multi-cluster network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114237858A true CN114237858A (en) | 2022-03-25 |
Family
ID=80747734
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210162360.6A Pending CN114237858A (en) | 2022-02-22 | 2022-02-22 | Task scheduling method and system based on multi-cluster network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114237858A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116302457A (en) * | 2023-05-25 | 2023-06-23 | 之江实验室 | Cloud primary workflow engine implementation method, system, medium and electronic equipment |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130346994A1 (en) * | 2012-06-20 | 2013-12-26 | Platform Computing Corporation | Job distribution within a grid environment |
| CN111782370A (en) * | 2020-06-30 | 2020-10-16 | 平安科技（深圳）有限公司 | Kubernetes multi-cluster management system, method, terminal and storage medium |
| CN113032126A (en) * | 2021-04-07 | 2021-06-25 | 北京理工大学 | High-concurrency cloud workflow scheduling engine cross-cluster communication system and method |
| CN113918270A (en) * | 2020-07-08 | 2022-01-11 | 电科云(北京)科技有限公司 | Cloud resource scheduling method and system based on Kubernetes |
-
2022
- 2022-02-22 CN CN202210162360.6A patent/CN114237858A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130346994A1 (en) * | 2012-06-20 | 2013-12-26 | Platform Computing Corporation | Job distribution within a grid environment |
| CN111782370A (en) * | 2020-06-30 | 2020-10-16 | 平安科技（深圳）有限公司 | Kubernetes multi-cluster management system, method, terminal and storage medium |
| CN113918270A (en) * | 2020-07-08 | 2022-01-11 | 电科云(北京)科技有限公司 | Cloud resource scheduling method and system based on Kubernetes |
| CN113032126A (en) * | 2021-04-07 | 2021-06-25 | 北京理工大学 | High-concurrency cloud workflow scheduling engine cross-cluster communication system and method |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116302457A (en) * | 2023-05-25 | 2023-06-23 | 之江实验室 | Cloud primary workflow engine implementation method, system, medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US7185046B2 (en) | Submitting jobs in a distributed computing environment | |
| CA3000422C (en) | Workflow service using state transfer | |
| CN102033777B (en) | Distributed type job scheduling engine based on ICE (internet communication engine) | |
| US8584136B2 (en) | Context-aware request dispatching in clustered environments | |
| CN101146127B (en) | A client buffer update method and device in distributed system | |
| CN103092698A (en) | System and method of cloud computing application automatic deployment | |
| CN105808334A (en) | MapReduce short job optimization system and method based on resource reuse | |
| CN102521044A (en) | Distributed task scheduling method and system based on messaging middleware | |
| CN104735095A (en) | Method and device for job scheduling of cloud computing platform | |
| CN112445598B (en) | Task scheduling method and device based on quartz, electronic equipment and medium | |
| CN105187327A (en) | Distributed message queue middleware | |
| JP7161560B2 (en) | Artificial intelligence development platform management method, device, medium | |
| CN112787999B (en) | Cross-chain calling method, device, system and computer readable storage medium | |
| CN112199353A (en) | Data processing method and electric power customer service platform | |
| CN114116149A (en) | Task scheduling method, device, equipment and medium | |
| US8812578B2 (en) | Establishing future start times for jobs to be executed in a multi-cluster environment | |
| CN114237858A (en) | Task scheduling method and system based on multi-cluster network | |
| CN115982273A (en) | Data synchronization method, system, electronic equipment and storage medium | |
| CN115617480A (en) | Task scheduling method, device and system and storage medium | |
| CN115829266A (en) | Method and device for job cooperation of cross-scheduling system, electronic equipment and storage medium | |
| CN115658248A (en) | Task scheduling method and device, electronic equipment and storage medium | |
| CN116643854A (en) | Service arrangement method and device, flow controller and task processor | |
| CN113098960A (en) | Service operation method, device, server and storage medium | |
| CN116991823A (en) | Data migration method, device, terminal equipment and storage medium | |
| US20230418681A1 (en) | Intelligent layer derived deployment of containers |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20220325 |
|
| RJ01 | Rejection of invention patent application after publication |