Multi-queue-based energy-efficient scheduling strategy for tasks with deadline constraints in cloud data center

The trade-off between energy efficiency and resource utilization incurred by the real-time tasks scheduling strategy in a cloud computing environment is still a challenging issue because of the large scale of heterogeneous tasks and physical machines. In this paper, we propose an energy-efficient sc...

Full description

Saved in:
Bibliographic Details
Published in:The Journal of supercomputing 2025, Vol.81 (1), Article 4
Main Authors: Duan, Lintao, Wang, Haiying
Format: Article
Language:English
Subjects:
Algorithms
Business metrics
Cloud computing
Compilers
Computer Science
Energy efficiency
Interpreters
Performance enhancement
Performance evaluation
Performance measurement
Processor Architectures
Programming Languages
Queues
Real time
Resource scheduling
Resource utilization
Scheduling
Task scheduling
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The trade-off between energy efficiency and resource utilization incurred by the real-time tasks scheduling strategy in a cloud computing environment is still a challenging issue because of the large scale of heterogeneous tasks and physical machines. In this paper, we propose an energy-efficient scheduling framework for cloud–edge–end architecture, which includes two-level internal schedulers, a task dispatcher and a group migrator, based on a multi-queue model. The aim of the proposed scheduling framework is to achieve improvement in energy efficiency and resource utilization without reducing task schedulability obviously during the execution of a large number of real-time tasks. To achieve this, the framework keeps tasks in a pending queue and divides these waiting tasks into independent groups with the same execution speed, and then dispatches and migrates them to the suitable machines that have the optimal energy efficiency and resource utilization. We conduct extensive simulate experiments under various workload conditions and evaluate our scheduling strategy using eight performance metrics. The experimental results show that our scheduling strategy can improve the performance metrics when compared with the competing algorithms.
ISSN:0920-8542
1573-0484
DOI:10.1007/s11227-024-06580-1