Extendable Architecture and Service Infrastructure for Cloud-Aware Software

Competence centre: Agent Core Technologies
Prof. Dr.-Ing. Sahin Albayrak, Jakob Tonn
Partners: Atos It Solutions and Services GmbHCharité - Universitätsmedizin BerlinFraunhofer FOKUSOrga Systems GmbHUniversität MainzProSyst Software GmbHMATERNA GmbH


So far, development of cloud computing applications is characterized by either using the advantages of specialized cloud services such as file storage or distributed editing (PaaS) on the API level, or by using highly-scalable, pre-installed web services (SaaS). This level of specialization makes interoperability very hard to reach. In the case of isolated private clouds, which are currently replacing traditional internal data centres, there are also constraints regarding scalability and high availability. Thus, one main goal in EASI-CLOUDS is to strengthen interoperability between all levels of the cloud computing stack (IaaS, PaaS, SaaS) through the development of open standards, in order to support federation of cloud services.

A second goal is the creation of a service toolkit, which contains of services optimized for clouds, and forms a modular, dynamically scaling platform. This extends the concept of creating software by using separate services in a modular fashion, which is already being propagated by the idea of service-oriented architectures (SOA), to the domain of cloud computing. The new properties also imply new services, to be able to represent the adapted business models and additional requirements in addition to technical extensions. The german consortium focusses on the development of services in the fields of SLA management, billing and payment services, and workflow technologies.

Furthermore, the evolution of the OSGI platform towards a cloud compatible service platform for applications is a very important work package in the project. As a standardized service platform in the J2EE area, OSGI is optimally feasible for extension through new services which are designed for the cloud (cloud-sensitive services), and allow access to scalability and adaptive behavior (elasticity) already on the service level. This goal is intended to be achieved in close European cooperation.
Within the German cooperation project, these conceptional and technical works are validated in a concrete medical application scenario. This application will realize a complex, distributed diagnosis scenario, in which examination data is processed and visualized according to the requirements of the diagnosis steps. Using the processed data, medical experts then create the diagnosis for the examined patient. The large amount of data and the complex and intensive calculations in this use case are typical requirements, which can only be fulfilled efficiently by a cloud solution.

The relevance of the EASI-CLOUDS results is, in addition to the technical implementation, also validated from an economical perspective through analysis of business models, stakeholders, and market, which shows the economical value of the cloud solution and provides a foundation for exploitation.