Multi-tenant-oriented task scheduling method and related device

By using dynamic priority labels and sliding window rate limiting mechanisms, combined with concurrent consumption tokens and time-slice scheduling algorithms, the problem of low resource utilization and efficiency in task scheduling in multi-tenant SaaS environments is solved. This enables priority processing of critical tasks and intelligent isolation of resources, thereby improving the stability and efficiency of the system.

CN122293748APending Publication Date: 2026-06-26FAN RUAN SOFTWARE CO LTD
0 Cites 0 Cited by

Patent Information

Application Number
CN202610393642.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-03-27
Publication Date
2026-06-26

AI Technical Summary

Technical Problem

In multi-tenant SaaS environments, existing technologies cannot intelligently adjust strategies to prioritize critical business operations, resulting in low resource utilization and scheduling efficiency. They also suffer from uneven resource allocation, poor resilience to sudden traffic surges, rigid scheduling strategies, and rudimentary task failure handling mechanisms.

Method used

Employing dynamic priority tags and a sliding window rate limiting mechanism, the system determines dynamic priorities by receiving task metadata, allocates tasks to the corresponding target priority queues, and uses a concurrent consumption token mechanism. Combined with a time-slice scheduling algorithm and an adaptive scheduler, it achieves intelligent task adjustment and resource isolation.

Benefits of technology

It improves resource utilization and scheduling efficiency, ensures that critical tasks are prioritized, prevents sudden traffic surges, optimizes resource allocation and task failure handling, and enhances system stability and efficiency.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a multi-tenant-oriented task scheduling method and a related device, and relates to the technical field of cloud computing, a to-be-scheduled task and task meta-information thereof are received, dynamic priority information is determined, under the condition that tenants corresponding to the to-be-scheduled task have valid quotas, the to-be-scheduled task is sent to a target priority queue, and the to-be-scheduled task is scheduled according to the target priority queue. And when the tenant has the concurrent consumption token in the target priority queue, obtaining the concurrent consumption token, executing the to-be-scheduled task, and releasing the concurrent consumption token. Different dynamic priorities are set for the to-be-scheduled tasks, it is guaranteed that the tasks are distributed to the resource-isolated independent queues, it is guaranteed that all the tasks are processed preferentially, and the resource utilization rate is increased; according to the method, sliding window flow limiting is implemented before tasks are issued, concurrent consumption tokens of tenant granularity are distributed during task consumption, the tasks among tenants are isolated, intelligent adjustment of the tasks is achieved in a complex multi-tenant and high-concurrency production scene, and the resource utilization rate and the scheduling efficiency are greatly improved.
Need to check novelty before this filing date? Find Prior Art