This paper describes our work on STORM CLOUDS[1], a project with the main objective of migrating smart-city services, that Public Authorities (PAs) currently provided using traditional Information Technology, to a cloud-based environment. Our organization was in charge of finding the technical solutions, so we designed and implemented a cloud computing solution called Storm Clouds Platform (SCP), for that purpose. In principle, the applications we ported could run on a public cloud service, like Amazon Web ServicesTM[2] or Microsoft® Azure[3], that provide computational resources on a pay-per-use paradigm. However, these solutions have disadvantages due to their proprietary nature: vendor lock-in is one of the issues but other serious problems are related to the lack of full control on how data and applications are processed in the cloud. As an example, when using a public cloud, the users of the cloud services have very little control on the location where applications run and data are stored, if there is any. This is identified as one of the most important obstacles in cloud computing adoption, particularly in applications manage personal data and the application provider has legal obligation of preserving end user privacy[4]. This paper explains how we faced the problem and the solutions we found. We designed a cloud computing platform — completely based on open software components — that can be used for either implementing private clouds or for porting applications to public clouds.
IaaS Public Cloud Computing Platform Scheduling Model and Optimization Analysis
