A task scheduling method and related device

By having the first computing node actively request task allocation and manage identification information, the problem of low task execution efficiency in existing task scheduling methods is solved. This achieves load balancing and fault migration among computing nodes, thereby improving task execution efficiency and system stability.

CN120144243BActive Publication Date: 2026-05-29SHENZHEN HUAWEI CLOUD COMPUTING TECHNOLOGIES CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SHENZHEN HUAWEI CLOUD COMPUTING TECHNOLOGIES CO LTD
Filing Date
2025-01-21
Publication Date
2026-05-29

AI Technical Summary

Technical Problem

Existing task scheduling methods may fail to schedule tasks in a timely manner when there are a large number of computing nodes, resulting in low task execution efficiency. Furthermore, computing nodes cannot perceive the task execution status, leading to uneven task distribution and affecting execution efficiency.

Method used

The first computing node actively sends request information to the second computing node, determines the time to send the request based on its own load, prioritizes the allocation of tasks to computing nodes with lower loads, and manages the task status through identification information to avoid duplicate instructions and fault migration.

Benefits of technology

It improves task execution efficiency, avoids task switching problems caused by uneven task allocation and computing node failure, and enhances system stability and task processing capabilities.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120144243B_ABST
    Figure CN120144243B_ABST
Patent Text Reader

Abstract

The application provides a task scheduling method and related equipment. The method is executed by infrastructure managed by a cloud management platform. The method comprises: receiving information of a first task at a first time; sending first request information for requesting execution of the task at a second time, the second time being at least separated from the first time by a first time length, the first time length being determined according to a first load value, the first load value being determined according to a number of tasks currently executed by a computing node and / or resource utilization; receiving first response information, the first response information being used for indicating execution of a second task or being used for indicating that there is no unexecuted task; when the first response information is used for indicating execution of the second task, executing the second task according to the first response information. The method can avoid over-reliance on scheduling of a scheduling node to cause the task to be unable to be timely distributed, and make a computing node with lower load preferentially send the request information, thereby improving execution efficiency of the task.
Need to check novelty before this filing date? Find Prior Art