Paper Info
Title
An adaptive and deadline-constrained workflow scheduling algorithm in infrastructure as a service clouds
Abstract
Cloud computing technology, which is a model of service provisioning in distributed systems, has been introduced as a way to execute workflow applications. One of the most challenging problems for executing workflow applications in cloud environment is developing workflow scheduling algorithms that consider different QoS requirements such as time and cost. We have previously presented a workflow scheduling algorithm called the CTDC (Constrained Time and Decreased Cost), which decreases the execution cost of the workflow while meeting user-defined deadline. However, the CTDC algorithm is a static scheduling algorithm and does not take into account the virtual machine performance variability. In CTDC algorithm, if the performance of virtual machines is reduced during the execution of the workflow due to the virtualization of resources and shared nature of infrastructure, the virtual machines cannot complete the tasks within the estimated time, and thus, the defined deadline will be missed. Therefore, in this paper, we complete the CTDC algorithm and propose a scheduling algorithm named AWS-CTDC (Adaptive Workflow Scheduling-CTDC) which considers the performance variation of virtual machines in the cloud environment and tries to complete the execution of the workflow within the user-defined deadline. The AWS-CTDC algorithm involves initial static scheduling, monitoring the finish time of completed tasks and rescheduling unexecuted tasks, if needed. The proposed algorithm is evaluated by a simulation process using WorkflowSim which is based on CloudSim. To evaluate the proposed algorithm, its results are compared with the results of CTDC, JIT-C (Just-In-Time), RTC (Robustness-Time-Cost), RCT (Robustness-Cost-Time), and CEGA (Cost-Effective Genetic Algorithm) algorithms. The results show that the proposed algorithm performs better than the above-mentioned algorithms.
Year
DOI
Venue
2022
10.1007/s42044-021-00082-6
Iran Journal of Computer Science
Keywords
DocType
Volume
Workflow application, DAG, Adaptive scheduling, Deadline constraint, IaaS cloud
Journal
5
Issue
ISSN
Citations 
1
2520-8438
0
PageRank 
References 
Authors
0.34
0
2
Name
Order
Citations
PageRank
Ghafouri Robabeh100.34
A. Movaghar219732.28