A Service-Oriented Framework for Scientific Cloud Computing

A Service-Oriented Framework for Scientific Cloud Computing

Abstract

Research in numerous areas has shifted towards considerable utilization of IT resources. Thus, the major objective of e-Science is to provide IT infrastructures that simplify and support research in these disciplines. In order to achieve this goal, e-Science has increasingly investigated the concept of Cloud Computing within the last years. Cloud Computing promises on-demand access to virtually infinite computing and storage resources by means of virtualization and Web service technologies. Cloud computing adopts the concept of Everything as a Service (XaaS) and is usually defined as a layered model containing Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS). Firstly, IaaS refers to the provisioning of virtually infinite computing and storage resources. Secondly, PaaS focuses on support of different execution environments such as HPC programming models, data-intensive computation, data mediation, or the execution of workflows. Finally, SaaS facilitates the access to applications via Web interfaces. The primary aim of this PhD thesis is to contribute to the design and development of a scientific Cloud environment on the basis of the XaaS concept. Furthermore, the scientific Cloud environment takes the IaaS, PaaS, and the SaaS layer into account and focuses on user-centric challenges therein. The five main contributions of this thesis are: (1) design and implementation of a versatile Cloud Environment for e-Science taking user- and infrastructure-centric challenges into account, (2) support Cloud based access and mediation of data sources, (3) the provisioning and orchestration of scientific applications within the Cloud, (4) the adaptive configuration of Cloud resources, and (5) the evaluation and usage of the versatile Cloud environment. The scientific Cloud environment delineated herein tackles these objectives on the basis of virtual appliances, Web service technologies, and autonomic computing concepts. The contributions of this thesis have been carried out and evaluated within multiple collaborations and European projects (@neurIST, VPH-Share).

Grafik Top
Authors
  • Köhler, Martin
Grafik Top
Shortfacts
Category
Thesis (PhD)
Divisions
Scientific Computing
Subjects
Software Engineering
Parallele Datenverarbeitung
Date
11 June 2012
Export
Grafik Top