CN-121996381-A - Task allocation method, device, equipment, medium and program product

CN121996381ACN 121996381 ACN121996381 ACN 121996381ACN-121996381-A

Abstract

The application provides a task allocation method, a device, equipment, a medium and a program product, which relate to the field of cloud computing and are characterized in that the method comprises the steps of obtaining matching degree of each node of a task processing system and requirements of an actuator, constructing a matching degree matrix of the actuator and the node based on the matching degree, arranging the actuator to a corresponding node based on the matching degree matrix, obtaining a maximum parallel node value of a target task, splitting the target task to obtain a subtask set, obtaining a load-balanced task allocation plan based on the maximum parallel node value, and allocating the subtasks in the subtask set to the actuators of the corresponding nodes according to the allocation plan. The application can effectively improve the resource utilization rate and the task execution efficiency.

Inventors

WANG XIAOSU
ZHAO YU
CHEN GUO
ZHANG CHUN
MA FUXIAO
SHI XUAN
HU DONGNI
Yan Chenrui
BAO XIANGJIN
SONG XIAOXUAN

Assignees

中移动信息技术有限公司
中国移动通信集团有限公司

Dates

Publication Date: 20260508
Application Date: 20260127

Claims (10)

1. A method of task allocation, the method comprising: obtaining the matching degree of each node of the task processing system and the requirements of an actuator; Constructing a matching degree matrix of the actuator and the nodes based on the matching degree, and arranging the actuator to the corresponding nodes based on the matching degree matrix; obtaining the maximum parallel node value of a target task; Splitting the target task to obtain a subtask set, and acquiring a task allocation plan of load balancing based on the maximum parallel node value; and distributing the subtasks in the subtask set to the executors of the corresponding nodes according to the distribution plan.
2. The method of claim 1, wherein the obtaining the maximum parallel node value for the target task comprises: acquiring task attribute parameters affecting task parallelism, wherein the task attribute parameters comprise at least one of subtask dependency relationship, data interaction complexity, calculation logic relevance and data processing scale; Quantizing each task attribute parameter, and acquiring a comprehensive quantized value based on the quantized task attribute parameters and corresponding weights; normalizing the comprehensive quantized values to obtain task parallelism; Constructing a parallelism-node number mapping function based on historical task data, and mapping the task parallelism into a maximum parallel node number candidate value through the parallelism-node number mapping function; Based on the maximum parallel node number candidate value, simulating a plurality of task allocation schemes through a load balancing algorithm, obtaining the load unbalance degree of each scheme, and determining the node number corresponding to the scheme with the minimum load unbalance degree as the maximum parallel node number value.
3. The method of claim 2, wherein constructing the parallelism-node number mapping function based on historical task data comprises: Collecting a historical task data set, wherein the historical task data comprises task parallelism of each historical task, the number of optimal parallel nodes actually operated by each historical task, resource capacity parameters and resource occupancy rate data during execution of each historical task; Carrying out standardization processing on the historical task data set to obtain a standardized training data set; taking task parallelism of historical tasks as input characteristics and the number of corresponding optimal parallel nodes as output labels, and constructing a mapping function model; Training the mapping function model based on a model training algorithm and the standardized training data set, and determining that the trained mapping function model is a parallelism-node number mapping function, wherein the model training algorithm comprises at least one of a linear regression algorithm, a decision tree algorithm and a neural network algorithm.
4. A method according to any one of claims 1 to 3, wherein said obtaining a load balanced mission plan based on said maximum parallel node values comprises: Acquiring a node range capable of participating in task allocation and available executor resource information of each node based on the maximum parallel node value; Acquiring the total amount of the subtasks; based on the total amount of the subtasks, the range of the nodes which can participate in task allocation and the resource information of the executor, adopting an allocation optimization algorithm, and under the constraint of not exceeding the maximum parallel node value, simulating and generating a plurality of groups of different subtask allocation schemes, wherein each group of subtask allocation schemes comprises the subtask allocation quantity of each node; Aiming at each group of subtask allocation schemes, according to the subtask allocation quantity of each node, combining the node resource bearing capacity, and acquiring the load unbalance degree corresponding to each subtask allocation scheme through a preset load balance evaluation index; And acquiring a subtask allocation scheme with minimum load unbalance.
5. The method of any one of claims 1 to 3, further comprising monitoring node load and task execution status in real time after the sub-tasks in the sub-task set are distributed to the corresponding node executors according to the distribution plan, and adjusting a task distribution scheme based on the monitoring result.
6. The method of claim 5, wherein the real-time monitoring of the node load and the task execution status, and adjusting the task allocation scheme based on the monitoring result, comprises: monitoring node load and task execution state; Determining that a node with the change of the node load exceeding a preset threshold or the delay time of task execution exceeding a preset time is an overload node, and migrating the task of the overload node to a node meeting a first condition; The first condition includes that the matching degree of the resource capacity parameter of the node and the actuator demand corresponding to the migration task is within a preset matching range, and the load of the node is lower than the average load of the system.
7. A task assigning apparatus, comprising: The first acquisition module is used for acquiring the matching degree of each node of the task processing system and the requirement of the executor; The arrangement module is used for constructing a matching degree matrix of the executor and the nodes based on the matching degree, and arranging the executor to the corresponding nodes based on the matching degree matrix; the second acquisition module is used for acquiring the maximum parallel node value of the target task; the third acquisition module is used for splitting the target task to obtain a subtask set, and acquiring a task allocation plan with balanced load based on the maximum parallel node value; and the distribution module is used for distributing the subtasks in the subtask set to the executors of the corresponding nodes according to the distribution plan.
8. An electronic device comprising a processor, a memory and a program stored on the memory and executable on the processor, the program when executed by the processor implementing the steps of the task allocation method according to any one of claims 1 to 6.
9. A computer-readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, implements the steps of the task allocation method according to any one of claims 1 to 6.
10. A computer program product comprising computer instructions which, when executed by a processor, implement the steps of the task allocation method according to any one of claims 1 to 6.

Description

Task allocation method, device, equipment, medium and program product Technical Field The present application relates to the field of cloud computing, and in particular, to a task allocation method, apparatus, device, medium, and program product. Background The intelligent center of the telecom operator has extremely high requirements on the execution efficiency and the resource utilization rate of tasks. In the related art, a task is split into a plurality of subtasks, and then the subtasks are sequentially or randomly distributed to different nodes for execution, and the simple distribution strategy cannot fully utilize the resources of the nodes, so that the utilization rate of the resources is low. Disclosure of Invention The embodiment of the application provides a task allocation method and a task allocation device, which can solve the problems that resources of nodes cannot be fully utilized during task allocation and the utilization rate of the resources is low. In order to solve the technical problems, the application is realized as follows: in a first aspect, an embodiment of the present application provides a task allocation method, where the method includes: obtaining the matching degree of each node of the task processing system and the requirements of an actuator; Constructing a matching degree matrix of the actuator and the nodes based on the matching degree, and arranging the actuator to the corresponding nodes based on the matching degree matrix; obtaining the maximum parallel node value of a target task; Splitting the target task to obtain a subtask set, and acquiring a task allocation plan of load balancing based on the maximum parallel node value; and distributing the subtasks in the subtask set to the executors of the corresponding nodes according to the distribution plan. In a second aspect, an embodiment of the present application provides a task allocation device, including: The first acquisition module is used for acquiring the matching degree of each node of the task processing system and the requirement of the executor; The arrangement module is used for constructing a matching degree matrix of the executor and the nodes based on the matching degree, and arranging the executor to the corresponding nodes based on the matching degree matrix; the second acquisition module is used for acquiring the maximum parallel node value of the target task; the third acquisition module is used for splitting the target task to obtain a subtask set, and acquiring a task allocation plan with balanced load based on the maximum parallel node value; and the distribution module is used for distributing the subtasks in the subtask set to the executors of the corresponding nodes according to the distribution plan. In a third aspect, an embodiment of the present application provides an electronic device, including a processor and a memory, where the memory stores a program or instructions executable on the processor, the program or instructions implementing the steps of the task allocation method according to the first aspect when executed by the processor. In a fourth aspect, embodiments of the present application provide a readable storage medium having stored thereon a program or instructions which when executed by a processor implement the steps of the task allocation method according to the first aspect. In the embodiment of the application, the matching degree of each node of a task processing system and the demand of an actuator is obtained, a matching degree matrix of the actuator and the node is constructed based on the matching degree, the actuator is arranged to the corresponding node based on the matching degree matrix, the maximum parallel node value of a target task is obtained, the target task is split to obtain a subtask set, a task allocation plan with balanced load is obtained based on the maximum parallel node value, and the subtasks in the subtask set are allocated to the actuators of the corresponding nodes according to the allocation plan. In this way, the matching degree of each node of the task processing system and the demand of the executor is obtained, the matching degree matrix of the executor and the node is constructed based on the matching degree, the executor is arranged to the corresponding node, meanwhile, the maximum parallel node value of the target task is obtained, the target task is split to obtain the subtask set, then the task allocation plan with balanced load is obtained based on the maximum parallel node value, and finally, the subtasks are allocated to the executors of the corresponding nodes according to the allocation plan, so that the accurate adaptation of the executor and the nodes and the balanced distribution of the system load are realized, the resource utilization rate and the task execution efficiency are effectively improved, and the task processing demand with high efficiency and intelligence is better satisfied. Drawings FIG. 1 is a schematic flow c