Abstract:
Cloud computing has recently emerged as a new technology paradigm for hosting and delivering Information Technology (IT) services to users over the Internet. Cloud Computing provide optimal and efficient computing through collaboration, agility, availability and scalability.
Cloud platforms provide essential benefits for organizations such as greater elasticity. These platforms eliminate the requirement for users to plan ahead for resource provisioning, faster delivery times. They provide faster delivery times to through the use of cloud-based data centers and services. However, cloud computing brings challenges alongside its benefits. First, developers often produce cloud applications for a specific cloud platform, and does not support portability to Multiclouds environments. The evolving this application to cloud services end up with major complexity issues.
To contribute to solving these problems, we proposed a model driven design approach, relying on development for portability-based design of cloud application architecture. This architecture can be easily deployed on Multicloud platforms, avoiding related portability or cloud service provider vendor lock-in problems.
In this dissertation, we present a framework for cloud application architecture development which encompasses strategic planning, architecture design, deployment, and evaluation phases. This thesis, within focuses on the architecture design phase and propose new scenarios based iterative process for architecture design. We introduce the concept of a scenario container at the architecture design level and develop the architecture iteratively as an integrated set of containers. In each iteration, we define the container of application components for one application scenario.
The proposed development process guides the architect in designing containerized portable cloud application architectures. We finally propose new container design patterns that integrate with selected cloud application patterns to support the development of scenario-based containerized architecture.
We validate the design approach by applying it to two different case studies: a Student Academic Result Record system, and Hajj and Umrah mobile healthcare system. In the second case study, we compare our new architecture development approach with the (TOGAF) enterprise architecture framework.
We verify the deployment of portable architectures on multicloud platforms using the CloudMIG simulation tool, for migrating e-commerce online pet store case study to cloud environments. We present simulation results for cloud deployment options (CDO) based on a medium response time (MRT) of the case study, considering multiple cloud platform deployment selection.
