Cloud computing allows users to tap into a massive pool of shared computing resources such as servers, storage, and network. These resources are provided as a service to the users allowing them to “plug into the cloud” similar to a utility grid. The promise of the cloud is to free users from the tedious and often complex task of managing and provisioning computing resources to run applications. At the same time, the cloud brings several additional benefits including: a pay-as-you-go cost model, easier deployment of applications, elastic scalability, high availability, and a more robust and secure infrastructure. One important class of applications that users are increasingly deploying in the cloud is database management systems. Database management systems differ from other types of applications in that they manage large amounts of state that is frequently updated, and that must be kept consistent at all scales and in the presence of failure. This makes it difficult to provide scalability and high availability for database systems in the cloud. In this thesis, we show how we can exploit cloud technologies and relational database systems to provide a highly available and scalable database service in the cloud. The first part of the thesis presents RemusDB, a reliable, cost-effective high availability solution that is implemented as a service provided by the virtualization platform. RemusDB can make any database system highly available with little or no code modifications by exploiting the capabilities of virtualization. In the second part of the thesis, we present two systems that aim to provide elastic scalability for database systems in the cloud using two very different approaches. The three systems presented in this thesis bring us closer to the goal of building a scalable and reliable transactional database service in the cloud.

Thesis (MEng)--Stellenbosch University, 2014. / ENGLISH ABSTRACT: Massively Multiplayer Online Games are large virtual worlds co-inhabited by players over the Internet. As up to thousands of players can be simultaneously connected to the game, the server and network architectures are required to scale e ciently. The traditional client/server model results in a heavy nancial burden for operation of the server. Various alternative architectures have been proposed as a replacement for the traditional model, but the adoption of these alternatives are slow as they present their own set of challenges. The proposed hybrid system is based on many di erent architectures and peer-topeer concepts that were reviewed in the literature. It aims to provide a compromise for existing, commercially successful MMOGs to introduce peer-to-peer components into their systems with no requirement of modi cation to their server or client software. With the system's design presented, the middleware software is implemented and deployed in a real, controlled environment alongside an Ultima Online game server and its clients. The movement game mechanic was distributed amongst the peers while the others remained the responsiblity of the server. A number of performance experiments are performed to measure the e ects of the modi ed system over the original client/server system on bandwidth, latency, and hardware impact. The results revealed an increase in the server bandwidth usage by 35%, slave bandwidth usage by 17% and supernode bandwiwdth usage by 3111%. The latencies of distributed server mechanics were reduced by up to 94%, while the non-distributed latencies were increased by up to 6000%. These results suggested that a system with absolutely no modi cation to the server is unlikely to provide the desired bene ts. However, with 2 minor modi cations to the server, the middleware is able to reduce both server load and player latencies. The server bandwidth can be reduced by 39%, while the supernode's bandwidth is increased only by 1296%. The distributed latencies maintain their reduction while non-distributed latencies remain unchanged from the C/S system. / AFRIKAANSE OPSOMMING: Massiewe Multispeler Aanlyn Speletjies (MMAS) is groot virtuele w^erelde op die Internet wat bewoon word deur spelers. Aangesien duisende spelers gelyktydig kan inskakel op die speletjie word daar verwag van die bediener en netwerk argitektuur om e ektief te skaleer om die groot hoeveelhede spelers te kan hanteer. Die traditionele kli ent/bediener model lei tot 'n groot nansi ele las vir die operateur van die bediener. Verskeie alternatiewe argitekture is al voorgestel om die tradisionele model te vervang, maar die aanvaarding en in gebruik neem van hierdie alternatiewe (soos eweknie-netwerke) is 'n stadige proses met sy eie stel uitdagings. Die voorgestelde hibriede stelsel is gebaseer op baie verskillende argitektuur- en eweknie konsepte wat in die literatuur oorweeg is. Die doel is om 'n kompromie vir bestaande komersieel suksesvolle MMASs te verskaf om eweknie komponente te implementeer sonder om die die bediener- of kli ent sagteware aan te pas. Met hierdie stelsel se ontwerp word die middelware sagteware ge mplementeer en gebruik in 'n regte, dog gekontroleerde omgewing, tesame met 'n Ultima Online bediener en sy kli ente. Die beweging speletjie meganisme word versprei onder die eweknie netwerk en die ander meganismes bly die verantwoordelikheid van die bediener. 'n Aantal eksperimente is ingespan om die e ek van die hibriede stelsel te meet op die oorspronklike kli ent/bediener stelsel, in terme van bandwydte, vertraging en impak op hardeware. Die resultate toon 'n toename van 35% in bediener-, 17% in slaaf-, en 3111% in supernodus bandwydte gebruik. Die vertraging van verspreide bediener meganismes neem af met tot 94%, terwyl onverspreide vertragings toeneem met tot 6000%. Hierdie resultate wys dat 'n stelsel wat geen aanpassing maak aan die bediener sagteware onwaarskynlik die gewenste voordele sal lewer. Deur egter 2 klein aanpassings toe te laat tot die bediener, is dit moontlik vir die hibriede stelsel om data las van die bediener en die speler se vertraging te verminder. Die bediener bandwydte kan met 39% verminder word, terwyl die supernodus bandwydte slegs met 1296% toeneem. Die verpreide vertragings handhaaf hul vermindering, terwyl die onverspreide vertragings onveranderd bly van die C/S stelsel.

La technologie Flash arrive à ses limites de miniaturisation. Ainsi, la nécessité de nouvelles technologies mémoire augmente. Les candidats au remplacement des mémoires Flash sont les technologies non volatiles émergentes comme les mémoires à pont conducteur (CBRAM), résistives à base d'oxyde (RRAM), mémoires magnétiques (MRAM) et mémoires à changement de phase (PCRAM). En particulier, les mémoires CBRAM sont basées sur structure simple métal-isolant-métal (MIM) et présentent plusieurs avantages par rapport aux autres technologies. La CBRAM est non volatile, à savoir qu'elle garde l’information lorsque l'alimentation est coupée, ses dimensions peuvent être réduites jusqu'à nœud 10 nm, elle peut facilement être intégrée dans le Back-End d’une intégration CMOS, enfin, elle a une vitesse de fonctionnement élevée à basse tension et un faible coût de fabrication. Néanmoins, les spécifications pour l'industrialisation des CBRAM sont très strictes. Dans cette thèse, nous analysons deux générations de technologie CBRAM, chacune adressant un marché d'application spécifique. La première partie de la thèse est consacrée à l’étude électrique des structures à base de cuivre et de GdOX, qui présentent comme avantages une conservation des données très stable et une bonne résistance lors de la soudure des puces, et un bon comportement de l'endurance. Cette technologie adresse principalement les applications à haute température telle que l'automobile. Pour répondre aux spécifications, un oxyde métallique dopé ainsi que des bicouches sont intégrés pour réduire la tension de formation de la mémoire et augmenter la fenêtre de programmation. Les performances en endurance sont améliorées. La deuxième partie est dédiée à une nouvelle technologie de CBRAM, avec un empilement de type MIM. Dans ce cas, nous avons démontré des temps de commutation très rapides de 20ns à basses tensions (2V), combinés avec une endurance satisfaisante et une bonne rétention des données. Cette technologie semble être compatible avec les applications Internet des objets (IOT). En résumé, au cours de ce doctorat, l'objectif principal était d'étudier la fiabilité des dispositifs embarqués CBRAM en termes d’écriture des données, endurance et la conservation de l’information. Une méthodologie de test spécifique a été développée, afin d’évaluer les performances des technologies étudiées. Des modèles physiques ont été mis au point pour expliquer et analyser les résultats expérimentaux. Sur la base des résultats obtenus, nous démontrons que la technologie de CBRAM est très prometteuse pour les futures applications de mémoires non volatiles. / Flash technology is approaching its scaling limits, so the demand for novel memory technologies is increasing. Promising replacing candidates are the emerging non volatile technologies such as Conductive Bridge Memory (CBRAM), Oxide based Resistive RAM (OXRAM), Magnetic Random Access Memory (MRAM) and Phase Change Memory (PCRAM). In particular, CBRAM is based on a simple Metal-Insulator-Metal (MIM) structure and presents several advantages compared to the other technologies. CBRAM is non volatile, i.e. it keeps the information when the power is off, it is scalable down to 10nm technology node, it can be easily integrated into the Back-End-of-Line (BEOL), finally, it has high operation speed at low voltages and low cost per bit. Nevertheless, demands for the industrialization of CBRAM are very stringent and issues related to device reliability are still to be faced. In this thesis we analyze two generations of CBRAM technology, each one addressing a specific application market. The first part of the PhD is dedicated to the electricalstudy of Cu-based/GdOx structures, which present the advantages of a very stable data retention and resistance to soldering reflow and also good endurance behavior. This CBRAM family addresses mainly the high temperature applications as automotive. To fulfill the specification requirements, doping of metal-oxide andbilayers are integrated to decrease the forming voltage and increase the programmingwindow. Better endurance performance is also achieved. The second part isdedicated to a new CBRAM technology, with a simple MIM structure. In this case, the device showsfast operation speed of 20ns at low voltages of 2V, combined with satisfying endurance and data retention. This technology seems to be compatible with the growing Internet of Things (IOT) market. In summary, during the PhD research, the main objective was to study the reliability of the embedded CBRAM devices in terms of forming, endurance and data retention. Some methodologies were developed and the electrical set-up was modified and adapted to specific measurements. Physical models were developed to explain and better fit the experimental results. Based on the obtained results, we demonstrate that the CBRAM technology is highly promising for future NVM applications.

Les systèmes distribués, tels que les grilles, sont utilisés aujourd’hui pour l’exécution des grandes applications parallèles. Quelques caractéristiques de ces systèmes sont l’interconnexion complexe de ressources qui pourraient être présent et de la facile passage à l’échelle. La complexité d’interconnexion vient, par exemple, d’un nombre plus grand de directives de routage pour la communication entre les processus et une latence variable dans le temps. La passage à l’échelle signifie que des ressources peuvent être ajoutées indéfiniment simplement en les reliant à l’infrastructure existante. Ces caractéristiques influencent directement la façon dont la performance des applications parallèles doit être analysée. Les techniques de visualisation traditionnelles pour cette analyse sont généralement basées sur des diagrammes de Gantt que disposent la liste des composants de l’application verticalement et metent la ligne du temps sur l’axe horizontal. Ces représentations visuelles ne sont généralement pas adaptés à l’analyse des applications exécutées en parallèle dans les grilles. La première raison est qu’elles n’ont pas été conçues pour offrir aux développeurs une analyse qui montre aussi la topologie du réseau des ressources. La deuxième raison est que les techniques de visualisation traditionnels ne s’adaptent pas bien quand des milliers d’entités doivent être analysés ensemble. Cette thèse tente de résoudre les problèmes des techniques traditionnelles dans la visualisation des applications parallèles. L’idée principale est d’exploiter le domaine de la visualisation de l’information et essayer d’appliquer ses concepts dans le cadre de l’analyse des programmes parallèles. Portant de cette idée, la thèse propose deux modèles de visualisation : les trois dimensions et le modèle d’agrégation visuelle. Le premier peut être utilisé pour analyser les programmes parallèles en tenant compte de la topologie du réseau. L’affichage lui-même se compose de trois dimensions, où deux sont utilisés pour indiquer la topologie et la troisième est utilisée pour représenter le temps. Le second modèle peut être utilisé pour analyser des applications parallèles comportant un très grand nombre de processsus. Ce deuxième modèle exploite une organisation hiérarchique des données utilisée par une technique appelée Treemap pour représenter visuellement la hiérarchie. Les implications de cette thèse sont directement liées à l’analyse et la compréhension des applications parallèles exécutés dans les systèmes distribués. Elle améliore la compréhension des modes de communication entre les processus et améliore la possibilité d’assortir les motifs avec cette topologie de réseau réel sur des grilles. Bien que nous utilisons abondamment l’exemple de la topologie du réseau, l’approche pourrait être adapté, avec presque pas de changements, à l’interconnexion fourni par un middleware d’une interconnexion logique. Avec la technique d’agrégation, les développeurs sont en mesure de rechercher des patterns et d’observer le comportement des applications à grande échelle. / Sistemas distribuídos tais como grids são usados hoje para a execução de aplicações paralelas com um grande número de processos. Algumas características desses sistemas são a presença de uma complexa rede de interconexão e a escalabilidade de recursos. A complexidade de rede vem, por exemplo, de largura de banda e latências variáveis ao longo do tempo. Escalabilidade é a característica pela qual novos recursos podem ser adicionados em um grid apenas através da conexão em uma infraestrutura pré-existente. Estas características influenciam a forma como o desempenho de aplicações paralelas deve ser analisado. Esquemas tradicionais de visualização de desempenho são usualmente baseados em gráficos Gantt com uma dimensão para listar entidades monitoradas e outra para o tempo. Visualizações como essa não são apropriadas para a análise de aplicações paralelas executadas em grid. A primeira razão para tal é que elas não foram concebidas para oferecer ao desenvolvedor uma análise que mostra a topologia dos recursos e a relação disso com a aplicação. A segunda razão é que técnicas tradicionais não são escaláveis quando milhares de entidades monitoradas devem ser analisadas conjuntamente. Esta tese tenta resolver estes problemas encontrados em técnicas de visualização tradicionais para a análise de aplicações paralelas. A idéia principal consiste em explorar técnicas da área de visualização da informação e aplicá-las no contexto de análise de programas paralelos. Levando em conta isto, esta tese propõe dois modelos de visualização: o de três dimensões e o modelo de agregação visual. O primeiro pode ser utilizado para analisar aplicações levando-se em conta a topologia da rede dos recursos. A visualização em si é composta por três dimensões, onde duas são usadas para mostrar a topologia e a terceira é usada para representar o tempo. O segundo modelo pode ser usado para analisar aplicações paralelas com uma grande quantidade de processos. Ela explora uma organização hierárquica dos dados de monitoramento e uma técnica de visualização chamada Treemap para representar visualmente a hierarquia. Os dois modelos representam uma nova forma de analisar aplicação paralelas visualmente, uma vez que eles foram concebidos para larga-escala e sistemas distribuídos complexos, como grids. As implicações desta tese estão diretamente relacionadas à análise e ao entendimento do comportamento de aplicações paralelas executadas em sistemas distribuídos. Um dos modelos de visualização apresentados aumenta a compreensão dos padrões de comunicação entre processos e oferece a possibilidade de observar tal padrão com a topologia de rede. Embora a topologia de rede seja usada, a abordagem pode ser adaptada sem grandes mudanças para levar em conta interconexões lógicas de bibliotecas de comunicação. Com a técnica de agregação apresentada nesta tese, os desenvolvedores são capazes de observar padrões de aplicações paralelas de larga escala. / Highly distributed systems such as grids are used today for the execution of large-scale parallel applications. Some characteristics of these systems are the complex resource interconnection that might be present and the scalability. The interconnection complexity comes from the different number of hops to provide communication among applications processes and differences in network latencies and bandwidth. The scalability means that the resources can be added indefinitely just by connecting them to the existing infrastructure. These characteristics influence directly the way parallel applications performance must be analyzed. Current traditional visualization schemes to this analysis are usually based on Gantt charts with one dimension to list the monitored entities and the other dimension dedicated to time. These visualizations are generally not suited to parallel applications executed in grids. The first reason is that they were not built to offer to the developer an analysis that also shows the network topology of the resources. The second reason is that traditional visualization techniques do not scale well when thousands of monitored entities must be analyzed together. This thesis tries to overcome the issues encountered on traditional visualization techniques for parallel applications. The main idea behind our efforts is to explore techniques from the information visualization research area and to apply them in the context of parallel applications analysis. Based on this main idea, the thesis proposes two visualization models: the three-dimensional and the visual aggregation model. The former might be used to analyze parallel applications taking into account the network topology of the resources. The visualization itself is composed of three dimensions, where two of them are used to render the topology and the third is used to represent time. The later model can be used to analyze parallel applications composed of several thousands of processes. It uses hierarchical organization of monitoring data and an information visualization technique called Treemap to represent that hierarchy. Both models represent a novel way to visualize the behavior of parallel applications, since they are conceived considering large-scale and complex distributed systems, such as grids. The implications of this thesis are directly related to the analysis and understanding of parallel applications executed in distributed systems. It enhances the comprehension of patterns in communication among processes and improves the possibility of matching this patterns with real network topology of grids. Although we extensively use the network topology example, the approach could be adapted with almost no changes to the interconnection provided by a middleware of a logical interconnection. With the scalable visualization technique, developers are able to look for patterns and observe the behavior of large-scale applications.

Non-relational database management systems address some of the limitations relational database management systems have when storing large volumes of unstructured, user generated text-based data in distributed environments. They follow different approaches through the data model they use, their ability to scale data storage over distributed servers and the programming interface they provide. An experimental approach was followed to measure the capabilities these alternative database management systems present in their approach to address the limitations of relational databases in terms of their capability to store unstructured text-based data, data warehousing capabilities, ability to scale data storage across distributed servers and the level of programming abstraction they provide. The results of the research highlighted the limitations of relational database management systems. The different database management systems do address certain limitations, but not all. Document-oriented databases provide the best results and successfully address the need to store large volumes of user generated text-based data in a distributed environment / School of Computing / M. Sc. (Computer Science)

Big Data on cities and regions bring new opportunities and challenges to data analysts and city planners. On the one side, they hold great promise to combine increasingly detailed data for each citizen with critical infrastructures to plan, govern and manage cities and regions, improve their sustainability, optimize processes and maximize the provision of public and private services. On the other side, the massive sample size and high-dimensionality of Big Data and their geo-temporal character introduce unique computational and statistical challenges. This chapter provides overviews on the salient characteristics of Big Data and how these features impact on paradigm change of data management and analysis, and also on the computing environment. / Series: Working Papers in Regional Science

Made available in DSpace on 2015-05-14T12:36:57Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 2319492 bytes, checksum: 7322dcd074c1c36c3d7750aad585ca51 (MD5) Previous issue date: 2010-03-26 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The increasing popularity and demand of wireless networks, over the past few years, has led to new challenges that stimulate the development of technology to build better standards and protocols. In such a context arises the Wireless Mesh Networks (WMN) which are low cost, self-organized and self-configurable networks that are easy to deploy and are fault tolerant. The WMN are aiming to achieve better performance, greater reliability and flexibility when compared to other wireless networks. These networks can provide an infrastructure consisting of dedicated devices for routing, strategically placed to achieve adequate coverage, providing connectivity within the coverage area for mobile clients. Therefore, these networks need to adapt to topology changes that can occur at any time. To allow this adjustment a routing protocol is required. An important component of a proactive routing protocol is a network topology discovery protocol, which disseminates link-state updates messages over the network. The proactive protocols usually have a high message overhead. However, a high message overhead has a major impact on the performance of a routing protocol and can lead to scalability problems. In such a context, the main contribution of this work is to build a reliable link-state dissemination protocol, called MLSD specially designed for RMSF whose goals are to reduce the messages overhead and reduce the total messages sent to advertise topology changes. / O aumento da popularidade e da demanda das redes sem fio ao longo dos últimos anos tem levado a novos desafios que estimulam o desenvolvimento da tecnologia na busca de melhores padrões e protocolos. Neste contexto, surgem as Redes em Malha Sem Fio - RMSF (Wireless Mesh Networks WMN), que são redes de baixo custo, que tem a capacidade de se organizar e se configurar automaticamente, de fácil implantação e com capacidade de tolerância a falhas. As RMSFs buscam alcançar melhor desempenho, maior confiabilidade e flexibilidade quando comparadas a outras redes sem fio. Estas redes podem fornecer uma infraestrutura formada por dispositivos dedicados à tarefa de roteamento, com seu posicionamento planejado para obter cobertura satisfatória, proporcionando conectividade dentro da área de cobertura para os clientes móveis. Por isto, estas redes precisam se adaptar a alterações de topologia, que podem ocorrer a todo o momento. Para permitir esta adaptação é necessário um protocolo de roteamento. Um componente importante de um protocolo de roteamento proativo é um protocolo de descoberta de topologia da rede, no qual as modificações na topologia da rede são divulgadas através de mensagens de atualização. Entretanto, protocolos proativos possuem uma alta carga de mensagens. Uma carga elevada de mensagens tem grande impacto no desempenho do protocolo de roteamento e podem levar a problemas de escalabilidade. Neste contexto, a principal contribuição deste trabalho é a construção de um protocolo de divulgação dos estados dos enlaces com garantia de entrega, chamado Mesh Network Link State Dissemination Protocol - MLSD, projetado para RMSFs, cujos objetivos são reduzir a carga de mensagens de atualização na rede e reduzir o total de mensagens enviadas para divulgar as atualizações.

Made available in DSpace on 2014-12-17T14:56:13Z (GMT). No. of bitstreams: 1 KayoGS_DISSERT.pdf: 4975392 bytes, checksum: 5d113169a6356e5e7704aec116237caf (MD5) Previous issue date: 2013-03-25 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior / This paper analyzes the performance of a parallel implementation of Coupled Simulated Annealing (CSA) for the unconstrained optimization of continuous variables problems. Parallel processing is an efficient form of information processing with emphasis on exploration of simultaneous events in the execution of software. It arises primarily due to high computational performance demands, and the difficulty in increasing the speed of a single processing core. Despite multicore processors being easily found nowadays, several algorithms are not yet suitable for running on parallel architectures. The algorithm is characterized by a group of Simulated Annealing (SA) optimizers working together on refining the solution. Each SA optimizer runs on a single thread executed by different processors. In the analysis of parallel performance and scalability, these metrics were investigated: the execution time; the speedup of the algorithm with respect to increasing the number of processors; and the efficient use of processing elements with respect to the increasing size of the treated problem. Furthermore, the quality of the final solution was verified. For the study, this paper proposes a parallel version of CSA and its equivalent serial version. Both algorithms were analysed on 14 benchmark functions. For each of these functions, the CSA is evaluated using 2-24 optimizers. The results obtained are shown and discussed observing the analysis of the metrics. The conclusions of the paper characterize the CSA as a good parallel algorithm, both in the quality of the solutions and the parallel scalability and parallel efficiency / O presente trabalho analisa o desempenho paralelo de uma implementa??o do Simulated Annealing Acoplado (CSA, do ingl?s Coupled Simulated Annealing) para otimiza??o de vari?veis cont?nuas sem restri??es. O processamento paralelo ? uma forma eficiente de processamento de informa??o com ?nfase na explora??o de eventos simult?neos na execu??o de um software. Ele surge principalmente devido ?s elevadas exig?ncias de desempenho computacional e ? dificuldade em aumentar a velocidade de um ?nico n?cleo de processamento. Apesar das CPUs multiprocessadas, ou processadores multicore, serem facilmente encontrados atualmente, diversos algoritmos ainda n?o s?o adequados para executar em arquiteturas paralelas. O algoritmo do CSA ? caracterizado por um grupo de otimizadores Simulated Annealing (SA) trabalhando em conjunto no refinamento da solu??o. Cada otimizador SA ? executado em uma ?nica thread, e essas executadas por diferentes processadores. Na an?lise de desempenho e escalabilidade paralela, as m?tricas investigadas foram: o tempo de execu??o; o speedup do algoritmo com respeito ao aumento do n?mero de processadores; e a efici?ncia na utiliza??o de elementos de processamento com rela??o ao aumento da inst?ncia do problema tratado. Al?m disso, foi verificada a qualidade da solu??o final. Para o estudo, esse trabalho analisa uma vers?o paralela do CSA e sua vers?o serial equivalente. Ambos algoritmos foram analisados sobre 14 fun??es de refer?ncia. Para cada uma dessas fun??es, o CSA ? avaliado utilizando de 2 a 24 otimizadores. Os resultados obtidos s?o exibidos e comentados observando-se as m?tricas de an?lise. As conclus?es do trabalho caracterizam o CSA como um bom algoritmo paralelo, seja na qualidade das solu??es como na escalabilidade e efici?ncia paralela

Les systèmes distribués, tels que les grilles, sont utilisés aujourd’hui pour l’exécution des grandes applications parallèles. Quelques caractéristiques de ces systèmes sont l’interconnexion complexe de ressources qui pourraient être présent et de la facile passage à l’échelle. La complexité d’interconnexion vient, par exemple, d’un nombre plus grand de directives de routage pour la communication entre les processus et une latence variable dans le temps. La passage à l’échelle signifie que des ressources peuvent être ajoutées indéfiniment simplement en les reliant à l’infrastructure existante. Ces caractéristiques influencent directement la façon dont la performance des applications parallèles doit être analysée. Les techniques de visualisation traditionnelles pour cette analyse sont généralement basées sur des diagrammes de Gantt que disposent la liste des composants de l’application verticalement et metent la ligne du temps sur l’axe horizontal. Ces représentations visuelles ne sont généralement pas adaptés à l’analyse des applications exécutées en parallèle dans les grilles. La première raison est qu’elles n’ont pas été conçues pour offrir aux développeurs une analyse qui montre aussi la topologie du réseau des ressources. La deuxième raison est que les techniques de visualisation traditionnels ne s’adaptent pas bien quand des milliers d’entités doivent être analysés ensemble. Cette thèse tente de résoudre les problèmes des techniques traditionnelles dans la visualisation des applications parallèles. L’idée principale est d’exploiter le domaine de la visualisation de l’information et essayer d’appliquer ses concepts dans le cadre de l’analyse des programmes parallèles. Portant de cette idée, la thèse propose deux modèles de visualisation : les trois dimensions et le modèle d’agrégation visuelle. Le premier peut être utilisé pour analyser les programmes parallèles en tenant compte de la topologie du réseau. L’affichage lui-même se compose de trois dimensions, où deux sont utilisés pour indiquer la topologie et la troisième est utilisée pour représenter le temps. Le second modèle peut être utilisé pour analyser des applications parallèles comportant un très grand nombre de processsus. Ce deuxième modèle exploite une organisation hiérarchique des données utilisée par une technique appelée Treemap pour représenter visuellement la hiérarchie. Les implications de cette thèse sont directement liées à l’analyse et la compréhension des applications parallèles exécutés dans les systèmes distribués. Elle améliore la compréhension des modes de communication entre les processus et améliore la possibilité d’assortir les motifs avec cette topologie de réseau réel sur des grilles. Bien que nous utilisons abondamment l’exemple de la topologie du réseau, l’approche pourrait être adapté, avec presque pas de changements, à l’interconnexion fourni par un middleware d’une interconnexion logique. Avec la technique d’agrégation, les développeurs sont en mesure de rechercher des patterns et d’observer le comportement des applications à grande échelle. / Sistemas distribuídos tais como grids são usados hoje para a execução de aplicações paralelas com um grande número de processos. Algumas características desses sistemas são a presença de uma complexa rede de interconexão e a escalabilidade de recursos. A complexidade de rede vem, por exemplo, de largura de banda e latências variáveis ao longo do tempo. Escalabilidade é a característica pela qual novos recursos podem ser adicionados em um grid apenas através da conexão em uma infraestrutura pré-existente. Estas características influenciam a forma como o desempenho de aplicações paralelas deve ser analisado. Esquemas tradicionais de visualização de desempenho são usualmente baseados em gráficos Gantt com uma dimensão para listar entidades monitoradas e outra para o tempo. Visualizações como essa não são apropriadas para a análise de aplicações paralelas executadas em grid. A primeira razão para tal é que elas não foram concebidas para oferecer ao desenvolvedor uma análise que mostra a topologia dos recursos e a relação disso com a aplicação. A segunda razão é que técnicas tradicionais não são escaláveis quando milhares de entidades monitoradas devem ser analisadas conjuntamente. Esta tese tenta resolver estes problemas encontrados em técnicas de visualização tradicionais para a análise de aplicações paralelas. A idéia principal consiste em explorar técnicas da área de visualização da informação e aplicá-las no contexto de análise de programas paralelos. Levando em conta isto, esta tese propõe dois modelos de visualização: o de três dimensões e o modelo de agregação visual. O primeiro pode ser utilizado para analisar aplicações levando-se em conta a topologia da rede dos recursos. A visualização em si é composta por três dimensões, onde duas são usadas para mostrar a topologia e a terceira é usada para representar o tempo. O segundo modelo pode ser usado para analisar aplicações paralelas com uma grande quantidade de processos. Ela explora uma organização hierárquica dos dados de monitoramento e uma técnica de visualização chamada Treemap para representar visualmente a hierarquia. Os dois modelos representam uma nova forma de analisar aplicação paralelas visualmente, uma vez que eles foram concebidos para larga-escala e sistemas distribuídos complexos, como grids. As implicações desta tese estão diretamente relacionadas à análise e ao entendimento do comportamento de aplicações paralelas executadas em sistemas distribuídos. Um dos modelos de visualização apresentados aumenta a compreensão dos padrões de comunicação entre processos e oferece a possibilidade de observar tal padrão com a topologia de rede. Embora a topologia de rede seja usada, a abordagem pode ser adaptada sem grandes mudanças para levar em conta interconexões lógicas de bibliotecas de comunicação. Com a técnica de agregação apresentada nesta tese, os desenvolvedores são capazes de observar padrões de aplicações paralelas de larga escala. / Highly distributed systems such as grids are used today for the execution of large-scale parallel applications. Some characteristics of these systems are the complex resource interconnection that might be present and the scalability. The interconnection complexity comes from the different number of hops to provide communication among applications processes and differences in network latencies and bandwidth. The scalability means that the resources can be added indefinitely just by connecting them to the existing infrastructure. These characteristics influence directly the way parallel applications performance must be analyzed. Current traditional visualization schemes to this analysis are usually based on Gantt charts with one dimension to list the monitored entities and the other dimension dedicated to time. These visualizations are generally not suited to parallel applications executed in grids. The first reason is that they were not built to offer to the developer an analysis that also shows the network topology of the resources. The second reason is that traditional visualization techniques do not scale well when thousands of monitored entities must be analyzed together. This thesis tries to overcome the issues encountered on traditional visualization techniques for parallel applications. The main idea behind our efforts is to explore techniques from the information visualization research area and to apply them in the context of parallel applications analysis. Based on this main idea, the thesis proposes two visualization models: the three-dimensional and the visual aggregation model. The former might be used to analyze parallel applications taking into account the network topology of the resources. The visualization itself is composed of three dimensions, where two of them are used to render the topology and the third is used to represent time. The later model can be used to analyze parallel applications composed of several thousands of processes. It uses hierarchical organization of monitoring data and an information visualization technique called Treemap to represent that hierarchy. Both models represent a novel way to visualize the behavior of parallel applications, since they are conceived considering large-scale and complex distributed systems, such as grids. The implications of this thesis are directly related to the analysis and understanding of parallel applications executed in distributed systems. It enhances the comprehension of patterns in communication among processes and improves the possibility of matching this patterns with real network topology of grids. Although we extensively use the network topology example, the approach could be adapted with almost no changes to the interconnection provided by a middleware of a logical interconnection. With the scalable visualization technique, developers are able to look for patterns and observe the behavior of large-scale applications.

The objective of this thesis was to design and implement a scalable and load efficient back-end system for web game services. This is of interest since web applications may overnight gain a significant increase in user base, because of viral sharing. Therefore designing the web application to service an increasing amount of users can make or break the application, in regard to keep the user base. Because of this, testing how well the system performs during heavy load can be used as a foundation when making a decision of when and where to scale up the application. The system was to be generically accessible through an Application Programming Interface (API) by the different game services. This was done using a RESTful architecture where emphasise was put on building the system scalable and load efficient. This thesis focuses on designing and implementing such a system, and how load testing can be used to evaluate this systems performance for an increasing amount of simultaneous clients using the web application. The results from load testing the implemented system was above the expectations, considering the hardware used when running the tests and hosting the system. The conclusion of this thesis is that by following REST when designing a web service, scalability becomes a natural part of how one would design the system.