Cloudware support for scientific applications

Cloud computing paradigm was introduced to serve the requirements of web applications and social networks with fluctuating number of users and inputs. The scientific community has found interest mainly in the deployment model of infrastructure-as-a-service that allows the highly desired control of the execution environments not sufficiently covered by other distributed computing paradigm. We try to prove in this paper, that using the proper tools, the scientific applications can also be well served by the other deployment model, the platform-as-aservice, by releasing the scientist from the burden of setting the execution environments and allowing him or her to focus on the application performances.