A New Scheduling Method for Workflows on Cloud Computing
Year: 2015
Type: article
Abstract: Cloud computing has recently become a very popular topic. Nowadays, with the increasing demand for process automation in thecloud, the investigation on cloud workflow scheduling strategies is becoming a significant issue. Majority of existing workflow schedulingalgorithms consider only compute resources that usually cannot be provisioned on demand size of workflows or not released to the environmentuntil the workflow execution completes. That is why the performance of these algorithms is being decreased and time and cost of them is beingincreased. In this paper, we present a new workflow scheduling method based on Ant Colony Optimization algorithm in order to reduce thisscheduling overhead with considering the above problems in our dynamic environment. Furthermore, we do consider these problems and varioustype VMs during the execution dynamically based on Amazon EC2. Also in comparison with state of the art in large-scale scheduling method,our datasets are based on real workflow applications with maximum 100 nodes. The results show that performance of our algorithm issignificantly better than Greedy Randomized Adaptive Search Procedure (GRASP) and scalable for increasing nodes of workflow. Keyword: workflow scheduling, cloud computing, VM allocation, Dynamic VM, user constraint, ACO (Ant Colony Optimization).
Authors: Seyed Ebrahim Dashti, Amir Masoud Rahmani
Language: English
Subfield: Information Systems
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W2807435568