Task allocation method and device, electronic equipment and storage medium

By generating task queues and node priority sequences, and dynamically adjusting task allocation based on task volume and load value, the problem of load imbalance in distributed computing systems is solved, and resource utilization efficiency is improved.

CN122132144APending Publication Date: 2026-06-02SHENZHEN LEXIN SOFTWARE TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
SHENZHEN LEXIN SOFTWARE TECH CO LTD
Filing Date
2026-03-20
Publication Date
2026-06-02

AI Technical Summary

Technical Problem

In distributed computing systems, traditional task allocation methods fail to effectively consider differences in task volume and heterogeneity of processing nodes, resulting in unbalanced loads, overloaded or idle resources on some nodes, and low overall resource utilization efficiency.

Method used

By calculating the load value of task processing nodes, a task queue and node priority sequence are generated. Tasks are allocated in order of task volume from high to low and load value from low to high. After each round of allocation, the node load value is updated and the priority sequence is adjusted to ensure that high-load tasks are allocated to lightly loaded nodes first, thus avoiding node saturation.

Benefits of technology

It achieves dynamic matching between the actual load of each processing node and the target processing capacity, improving global load balancing and resource utilization, and avoiding node overload and resource idleness.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122132144A_ABST
    Figure CN122132144A_ABST
Patent Text Reader

Abstract

This invention relates to the field of data request processing, disclosing a task allocation method, apparatus, device, and computer-readable storage medium. The method includes: calculating the ratio of the number of tasks to be processed to the maximum task processing capacity of each task processing node to obtain the load value corresponding to each task processing node; generating a task queue from all tasks in the task set according to the task volume in descending order; generating a node priority sequence for the current round from all task processing nodes according to the load value in ascending order; combining the task queue and the node priority sequence; sequentially retrieving tasks from the head of the queue and allocating them in order; repeating the steps from retrieving tasks from the head of the task queue to generating the node priority sequence for the next round until all tasks in the task queue have been allocated. This invention improves the global load balancing and resource utilization of task allocation.
Need to check novelty before this filing date? Find Prior Art