Modelling techniques for cloud- based solutions

Abstract

Cloud computing is the thing that everyone talks about. Various solutions were developed or rewritten and ported to the cloud. Even though, it creates certain benefits for its users, in certain instances it might lead to unexpected issues. Most of the software that was developed uses modeling technique which allow to foresee the possible bottlenecks and problems as well as tools that would be used for development. In addition, it would be able to breakdown the application into more detailed parts. Many of the techniques have been around for quite some time and in order to keep up with modern technologies, researchers build different extensions for them. At the present, cloud computing is not fully covered by the idea of software modeling techniques. This brings an idea of allowing developers and business people to have the Cloud implementation in mind., allowing to model and approximate where and how to provision. The purpose of this project is to help developers who are planning to either develop for or port to the cloud in order to reduce expenditures that may be incurred. This is to be done by introducing the extensions to the most popular modelling language nowadays, Unified Modelling Language (UML). This thesis will be tackling the three aspects of the cloud deployment service models, which are Software as a Service, Platform as a Service, and Infrastructure as a Service. The reason for that is that deployment of the application should be tight with its modeling as to take care of it in the later stages of the development. Each was looked at differently, but within same solution. Even though, these aspects differ from each other, the project proposes the solution which will fit to all of them. In a nutshell, the solution is the extension that features the classes for each of the previously mentioned models as well as it provides new relationship to connect the usual classes to the introduced ones. The main aim of the project is to reduce the amount of money that can be spent on cloud resources. The credibility of the extension is then proven by comparing the estimated costs incurred for two scenarios: using the extension for given scenario against choosing the default powers and capacities offered by the provider.