This thesis discusses the new way of delivering IT services over the Internet widely known as cloud computing. In its opening part, cloud computing is put into a historical context of the evolution of enterprise computing, and the dominant issues the IT department faces today are mentioned. Further, the paper deals with several components that make up the architecture of cloud computing and reviews the benefits and drawbacks an enterprise can have while it adopts this new model. One of the primary aims of this thesis is to identify the impact of the technology trends on cloud computing. The thesis brings together four major computing trends, namely virtualization, multi-tenant architecture, service-oriented architecture and grid computing. Another aim is to focus on two trends related to IT infrastructure that will lead to fundamental changes in IT industry. The first of them is the emergence of extremely large-scale data centers at low cost locations, which can serve tremendous amount of customers and achieve considerable economies of scale. The second trend this paper points out is the shift from multi-purpose all-in-one computers into a wide range of mobile devices dedicated to a specific user's needs. The last aim of this thesis is to clarify the economic impact of cloud computing in terms of costs and changes in business models. The thesis concludes by evaluating the current adoption and predicting the future trend of cloud computing.

Made available in DSpace on 2016-06-02T19:06:22Z (GMT). No. of bitstreams: 1 6616.pdf: 3224245 bytes, checksum: 3bea5c064e483543e0961a78c51613d2 (MD5) Previous issue date: 2014-10-20 / Software as a Service (SaaS) represent a form of software distribution on demand and accessible via the Internet. The development of SaaS enables service users (tenants) to benefit from the low cost of deployment and maintenance. The Multi-Tenant (MT) architecture is often used in the development of SaaS, it allows the sharing of software and hardware resources, providing cost savings to the costumers and the service provider. A challenge that the SaaS developer has is to keep sharing the instance of the software in scenarios with tenants who have differing needs. In this project, to address the variability of tenants in SaaS MT were set some patterns, based on Software Product Lines techniques for customizing user interfaces, data, processes and permissions. A framework called FoSaaS (Framework of Software as a Service) has also been developed to support the use of patterns, providing support to developers in SaaS MT coding. In addition to allowing the customization of SaaS MT, experiments in this project indicated that patterns improve reuse, quality and productivity in developed SaaS MT. The applicability of the patterns was evidenced through real projects in developing commercial applications. / Softwares como um Serviço (SaaS) representam uma forma de distribuição de software sob demanda e acessíveis via Internet. O desenvolvimento de SaaS possibilita aos consumidores do serviço (tenants) se beneficiarem do baixo custo de implantação e manutenção. A arquitetura Multi-Tenant (MT) é frequentemente utilizada no desenvolvimento de SaaS, pois permite o compartilhamento de recursos de software e hardware, proporcionando a redução de custos ao consumidor e ao fornecedor do serviço. Um desafio que o desenvolvedor de SaaS tem é o de manter o compartilhamento da instância do software em cenários com tenants que possuem necessidades divergentes. Neste projeto, para tratar a variabilidade dos tenants em SaaS MT foram criados alguns padrões, com base em técnicas de linhas de produtos de software, para customização de interfaces de usuários, dados, processos e permissões. Um framework denominado FoSaaS (Framework of Software as a Service) também foi desenvolvido para apoiar o uso dos padrões criados fornecendo apoio aos desenvolvedores na codificação de SaaS MT. Além de permitir a customização de SaaS MT, experimentos realizados neste projeto indicaram que os padrões melhoram o reúso, a qualidade e a produtividade dos SaaS MT desenvolvidos. A aplicabilidade dos padrões foi evidenciada por meio da realização de projetos reais no desenvolvimento de aplicações comerciais.

Queueing theory is widely used in practical queuing applications. It can be applied for specific models of queuing systems, especially the ones that follow the Markovian property. Its purpose is to predict system behaviour in order to be used for performance optimization. In this case study, it was used to evaluate an extended queuing model with agents serving multiple queues. The purpose was to try to capture more variability and input factors into the theoretical model and test its applicability on more extended models. The main objective was to use relevant queuing theory models to estimate the wait time using real contact center data. Different from the theoretical model, the service rates of the system model depended on how many queues an agent served concurrently, which increased the complexity of the model. The obtained results demonstrated some limitations that made the models too restrictive to be applied to a model with multi-skilled agents that were not equally available. Moreover, it was shown that heuristical approaches might be more suitable for more complex queuing systems that are not covered in queueing theory models. / Köteori används i stor utsträckning i praktiska kö-applikationer. Den kan tillämpas för specifika modeller av kö-system, speciellt de som följer Markovegenskapen. Dess syfte är att förutse systembeteende för att kunna användas för prestandaoptimering. I denna fallstudie användes den för att utvärdera en utökad kömodell med agenter som betjänade flera köer. Syftet var att försöka fånga mer variabilitet och inputfaktorer i den teoretiska modellen och testa dess tillämplighet för mer utökade modeller. Huvudmålet var att använda relevanta kö-teorimodeller för att estimera väntetiden med användning av riktiga contact center data. Till skillnad från den teoretiska modellen, betjäningsintensiteten för systemmodellen berodde på hur många köer en agent betjänade samtidigt, vilket ökade komplexiteten av modellen. De erhållna resultaten visade begränsningar som gjorde modellen för restriktiv för att appliceras på en modell med fler-kvalificerade agenter som inte var lika tillgängliga. Utöver detta så visade det sig att heuristiska metoder kan vara mer lämpliga för mer komplicerade system som inte täcks av kö-teori modeller.