Open Access Repository

Deployment of batch mode scientific workflow on a Computation-as-a-Service private cloud

Xu, Y, Liu, P ORCID: 0000-0003-2158-5442, Penesis, I ORCID: 0000-0003-4570-6034 and He, G 2018 , 'Deployment of batch mode scientific workflow on a Computation-as-a-Service private cloud', in Proceedings - 5th International Conference on Soft Computing and Machine Intelligence (ISCMI 2018) , IEEE, Nairobi, Kenya, pp. 123-128 .

Full text not available from this repository.

Abstract

Cloud computing is usually to address businessproblems of costly computing infrastructures but nowadays itis considered as a possible alternative to scientific workflowdeployment. Therefore, there are only limited cases forscientific and engineering computing in which there are taskparallelism closely coupled with high concurrent I/Orequirements. To address this issue, this paper developed anew resource management methodology to maximize overallmachine utilization levels while minimizing application runtime. The key strategy and algorithm in this methodologyconsist of: (i) a bottom-up architecture that utilizes resourcesfor both servers and clients. (ii) a maximum utilizationresource coloration algorithm based on node ability. Aprototype system was implemented by incorporating thepolicies and algorithms mentioned above in Cloud Computingand Distributed Systems (CLOUDS) Laboratory. Initial resultswere obtained by two different cases, by Rotorysics (formerlyPropella), a special marine hydrodynamics code for propellersand turbines and by DF_OSFBEM, a panel method code forunsteady 3D multiple-foil hydrodynamics. Results showed thatnew solution has speeded up total run time up to 50% at the2nd level --- the higher service ability level. By using thedeveloped methodology and exploration of Computation-as-a-Service (CaaS), the objective was achieved to acceleratescientific workflow efficiency in private cloud computingplatform.

Item Type: Conference Publication
Authors/Creators:Xu, Y and Liu, P and Penesis, I and He, G
Keywords: scientific workflows, resource management, private cloud computing, node ability, Computation-as-a-Service
Journal or Publication Title: Proceedings - 5th International Conference on Soft Computing and Machine Intelligence (ISCMI 2018)
Publisher: IEEE
Copyright Information:

Copyright 2018 IEEE

Related URLs:
Item Statistics: View statistics for this item

Actions (login required)

Item Control Page Item Control Page
TOP