Paper Info
Title
A Budget-Aware Algorithm For Scheduling Scientific Workflows In Cloud
Abstract
Commercial clouds are quickly becoming the goto platform for hosting on-demand and dynamically scalable services for scientific analyses and computation. Dynamic provisioning of resources is a critical element when utilising the cloud for executing large-scale and complex scientific analyses. In particular, hosting and managing data intensive applications on the cloud raises new challenges in terms of workflow scheduling. In this paper, we introduce a new Budget Aware Trickling (BAT) algorithm that addresses eScience workflow scheduling in the cloud. Our main focus in this paper is on data intensive applications that appear in scientific domains dealing with a large amount of data. The BAT algorithm builds upon the concept of Constrained Critical Paths (CCP) to execute a set of tasks on the same instance to lower the cost of communication and data movement between instances. Our BAT algorithm distributes budget based on the dependency structure inherent in workflows and we show that it yields 30% reduction in makespan while maintaining consistent success rate.
Year
DOI
Venue
2016
10.1109/HPCC-SmartCity-DSS.2016.30
PROCEEDINGS OF 2016 IEEE 18TH INTERNATIONAL CONFERENCE ON HIGH PERFORMANCE COMPUTING AND COMMUNICATIONS; IEEE 14TH INTERNATIONAL CONFERENCE ON SMART CITY; IEEE 2ND INTERNATIONAL CONFERENCE ON DATA SCIENCE AND SYSTEMS (HPCC/SMARTCITY/DSS)
Field
DocType
Citations 
Job shop scheduling,Bat algorithm,Fair-share scheduling,Computer science,Scheduling (computing),Algorithm,Provisioning,Real-time computing,Dynamic priority scheduling,Workflow,Cloud computing,Distributed computing
Conference
0
PageRank 
References 
Authors
0.34
0
3
Name
Order
Citations
PageRank
Vahid Arabnejad1303.13
Kris Bubendorfer234129.28
Bryan Ng310020.84