Submitted by Fabricia Fialho Reginato (fabriciar) on 2015-07-08T01:19:32Z No. of bitstreams: 1 VladimirGuerreiro.pdf: 5547701 bytes, checksum: b807e1f8091b49a5ee1e0b36e2ae4286 (MD5) / Made available in DSpace on 2015-07-08T01:19:32Z (GMT). No. of bitstreams: 1 VladimirGuerreiro.pdf: 5547701 bytes, checksum: b807e1f8091b49a5ee1e0b36e2ae4286 (MD5) Previous issue date: 2014 / Nenhuma / Em ambientes paralelos, uma das alternativas para tratar o dinamismo, tanto em nível de infraestrutura quanto de aplicação é o uso de migração, principalmente em aplicações que executam em fases utilizando BSP (Bulk Synchronous Parallel). Neste contexto, o modelo de reescalonamento MigBSP foi desenvolvido para tratar da realocação de processos em aplica- ções paralelas. Assim como o modelo BSP, ele considera as três fases de execução de uma superetapa: (i) computação local, (ii) comunicação global e (iii) uma barreira de sincroniza- ção; coletando dados localmente durante a computação para efetuar o cálculo do Potencial de Migração (PM) do processo. Com o PM e parâmetros adicionais fornecidos no inicio da execução da aplicação, o MigBSP tem condições de escolher processos candidatos a migração em uma aplicação paralela executando em um ambiente distribuído. Entretanto, as duas heurísticas possíveis de serem utilizadas hoje, dependem de informações fornecidas pelo usuário e/ou podem não selecionar uma quantidade eficiente de processos no momento do reescalonamento, podendo ser necessário várias chamadas para balancear o ambiente. Desta forma, esta disserta- ção apresenta duas novas heurísticas, MigCube e MigHull. Elas utilizam o MigBSP e efetuam a seleção automática de processos candidatos à migração sem a interferência do programador. As informações fornecidas pelo MigBSP são utilizadas nas heurísticas, a combinação das três métricas mensurados, posicionadas em um plano tridimensional, define cada processo como um ponto no espaço que possui as coordenadas x, y e z, onde cada eixo representa uma mé- trica para tomada de decisão. A heurística MigCube monta um cubo a partir das médias das distâncias entre os pontos, utilizando o processo com o maior PM como centro do cubo. A heurística MigHull segue a definição da Envoltória Convexa, tentando envolver todos os pontos, porém utilizando duas adaptações que se fazem necessárias para a aplicação neste trabalho. O MigBSP foi desenvolvido no simulador SimGrid, e este segue sendo utilizado para a criação das duas heurísticas apresentadas nesta dissertação. Nos testes realizados neste simulador, foi possível verificar um ganho de até 45% no tempo de execução da aplicação utilizando a heurística MigHull, e até 42% utilizando a MigCube, quando comparado a aplicação sem o modelo de migração. Porém, em simulações com um maior número de processos, este ganho tende a cair, já que um dos maiores problemas do BSP e aplicações que executam em grades é o tempo de sincronização de tarefas, ou seja, quanto mais processos, maior a necessidade de sincronização, e mesmo o balanceamento dos processos acaba tendo um resultado prejudicado. / In a parallel environment, one of the alternatives to address the dynamism, both at the infrastructure and application levels, is the use of migration, mostly with applications that execute in steps using BSP (Bulk Synchronous Parallel). In this context, the rescheduling model MigBSP was developed to deal with processes reallocation in parallel applications. As BSP model, MigBSP uses the three steps of a superstep: (i) computation, (ii) communication and (iii) a synchronization barrier; collecting local data during the computation step, to compute the processes’ Potential of Migration (PM). With the PM and additional parameters provided in the beginning of the application’s execution, MigBSP have conditions to choose the processes candidate to migrate in a parallel application running in a distributed system. However, the two heuristics possible to be used today depend of information provided by the user and/or may not select the proper quantity of processes in the rescheduling moment, being necessary many executions to balance the environment. This way, this dissertation present two new heuristics, MigCube and MigHull. They make use of MigBSP, and automatically will choose the processes to migrate without user interference. The information provided by MigBSP are used in the heuristics, the combination of the three measured metrics, positioned in a three-dimensional space, defines each process as a point in space and has the coordinates x, y e z, where each axis represents a metric for decision making. The MigCube heuristic build a cube from the average of the distances between points, using the process with the highest PM as the center of the cube. The MigHull follows the definition of a Convex Hull, trying to involve all points, but using two adaptations that are necessary to implement this work. The MigBSP was developed using SimGrid simulator, and it keeps being used to creation of the two heuristics presented in this dissertation. In the conducted tests in this simulator, was possible to achieve a gain of until 45% on application execution time using MigHull, and until 42% using MigCube, when compared with the application without the migration model. However, simulations with a bigger number of processes, this gain tends to fall, since one of the bigger problems of BSP and applications that run in grid is the time of tasks synchronization, that is, as more processes, more need of synchronization, and even the processes balancing ends up having an impaired outcome.

No description available.

最近，金融科技(FinTech)和行動金融服務，吸引越來越多的目光。新的創新金融科技服務，改變了金融服務的消費行為。行動網路的發展使人們能夠隨時隨地的享受行動銀行的服務已經是個不爭的事實。然而，由於無線網路先天的特性以及行動裝置的移動性，使行動金融的服務品質受到網路不穩定的影響。而且，隨著Bank 3.0時代的來臨，將會有大量的使用者同時使用行動金融服務，特別是在股市開盤以及重大訊息揭露的時候。因著大量使用者瞬間湧入，以及無線網路不穩定的影響，交易系統的效能很可能會時好時壞，所以無法滿足即時金融市場的需求。 本論文中，我們提出「行動銀行訊息即服務」的框架，使系統能夠很容易的水平擴充，並且能夠輕易的實現雙向通訊和雙向交易等多項行動金融服務。為了達到最少成本追求最大利益的目的，我們發展了能夠適應突發流量以及網路不穩定性的任務分配演算法，使得不用增加額外硬體成本的前提下改善系統效能。然後，為了實驗欲模擬大量行動裝置的使用者，我們觀察真實網路的特性並發明了網路延遲自相關模型來驗證我們提出的任務分配演算法。結果顯示，透過此任務分配演算法，確實能夠有效改善系統資源管理的能力。最後，本研究將系統佈署於真實網路環境當中，並且發現進行同樣實驗的結果與採用網路延遲自相關模型的實驗結果一致。因此，本研究間接驗證了網路自相關模型的正確性，以及證明本任務分配演算法，在突發流量和網路不穩定的即時行動金融服務環境下，能有效降低系統響應時間。 / FinTech (financial technology) and mobile financial services are getting more and more attention recently. New innovative FinTech services change the consumption behavior for financial services. It is an indisputable fact development of the mobile Internet allows people to enjoy mobile banking everywhere and anytime. However, due to the nature of wireless networking and the mobility of the mobile device, the quality of mobile financial service will be affected by network instability. Moreover, with the coming of Bank 3.0, a huge amount of users would be in the mobile service of finance simultaneously, especially when the instances of the stock market opening or disclosure of highly important financial message. As the result of bursty traffic and network instability, the performance of transaction system is up and down, making it tough to satisfy the demand of real-time financial markets. In this thesis, we propose a “Mobile Banking Messaging as a Service Framework” that can easily scale out and fulfill functions comprising Bilateral Communication, Bilateral Trading, and many other mobile financial services. To pursuit of the greatest benefit along with investment of the least resources, we develop the task assignment algorithm which can adapt the system to bursty traffic and unstable networks to improve performance for free. Then, in order to simulate a large number of mobile users, we observe the characteristic of real-world network delay and propose a network delay autocorrelation model to verify our task assignment algorithm. The results of experiment show that we could actually use our task assignment algorithm to improve the ability of the system to manage resource. Finally, we deploy our system in a real-world network delay environment and find that the results obtained in the real condition are the same with our simulation results. Therefore, this research can indirectly verify the correctness of the network delay autocorrelation model, and prove that our task assignment algorithm can effectively reduce the system response time for real-time mobile financial service system under bursty traffic and unstable networks.

Submitted by Silvana Teresinha Dornelles Studzinski (sstudzinski) on 2016-11-01T15:35:45Z No. of bitstreams: 1 Clebio Dossa_.pdf: 1252617 bytes, checksum: 784b95c29ee09e2a922686b26cb7aa51 (MD5) / Made available in DSpace on 2016-11-01T15:35:45Z (GMT). No. of bitstreams: 1 Clebio Dossa_.pdf: 1252617 bytes, checksum: 784b95c29ee09e2a922686b26cb7aa51 (MD5) Previous issue date: 2016-08-18 / Nenhuma / Atualmente, muitos serviços distribuem a carga entre diversos nós computacionais direcionando as conexões com alguma estratégia de balanceamento para divisão da carga. O advento do uso de redes definidas por software (SDN) está mudando paradigmas da administração de redes, absorvendo serviços especializados, automatizando processos e gerando inteligência para regras estáticas com uma grande variedade de opções de implementação. O balanceamento de carga é um dos serviços especializados que pode usufruir dos conceitos de SDN, sem definições e processos estáticos como ocorre muitas vezes nos atuais modelos usados de balanceamento de carga. A definição dos protocolos que suportam SDN usualmente permitem soluções alternativas e eficientes para este problema, desta forma, neste trabalho, é apresentada uma proposta de metodologia para balanceamento de carga entre distintos servidores de um pool com a troca do destino de tráfego realizada pela rede. Esta solução é chamada Core-based load balance (CoreLB), pois o serviço especializado de balanceamento de carga é realizado pela rede onde a administração de pacotes é nativamente realizada. A metodologia faz uso do protocolo SNMP para análise de recursos dos servidores com o objetivo de avaliar a situação de carga de cada nó computacional e de estatísticas de consumo de rede através do protocolo OpenFlow. Este trabalho avaliou o balanceamento de carga em serviços Web e a união de estatísticas de rede e da carga dos servidores, para a tomada de decisão de balanceamento, mostra-se uma metodologia eficiente e com melhores tempos de resposta ao usuário comparado com outras metodologias de avaliadas. Também melhorou a distribuição de consumo de recursos entre os servidores. / Currently, most services balance the load between distinct hosts forwarding connections with a load balance strategy in front. Usually, a dedicated appliance is responsible to performthe balance and may be a fault point and become expensive. The new concepts of computer network architecture with Software-Defined Networking (SND) are changing the network management, absorving specialist services, automating process and building intelligence to statics rules with loads of delivery options. The load balance is a specialized service that can enjoy in a positive way of SDN concepts, with low costs, in a flexible way as per the process needs instead of a plastered process definitions that occurs in many actual models. The OpenFlow protocol definition allow us to use a new solution to address this issue. This work shows a load balance purpose between distinct hosts with the destination change of connections made by the network core. It calls Core-based load balance (CoreLB) because the specialized load balance service move to the network core where the package forwarding is naturally made. This solution intend to use the SNMP protocol to analyse the hosts resources to evaluate server’s load. Using the network forwarding statistics and OS load informations, an efficient solution of load balance, the metodology proved to be efficient with better users’ response times average of 19% than no balanced scenario as well as around 9% better than others load balance strategies and a properly balance consumption of resources from hosts side. This process can be inhered in distinct models, however, this research intend to evaluate Web Services.

<p>One of the most parts of VOIP management is fault management and, in having a good fault management, finding good mechanisms to detect faults in the network have to be considered.</p><p>The main focus of this project is to implement different types of fast fault recovery protocols in networks<strong>,</strong> especially networks that carry IP telephony. Having a complete understanding of some common link failure detection and fault recovery protocols<strong>,</strong> such as spanning tree protocol (STP), rapid spanning tree protocol (RSTP) and per-VLAN spanning tree protocol (PVSTP)<strong>,</strong> and also having a complete understanding of three other common techniques for fault detection and fault recovery, such as hot standby routing protocol (HSRP), virtual router redundancy protocol (VRRP) and gateway load balancing protocol (GLBP) will be regarded in the project. We are going to test some fault recovery protocols which can be used in IP telephony networks and choose the best. We intend to focus on this issue in LAN environment in theoretical descriptions and practical implementations.</p><p>The final outcome of the thesis is implementation in the Halmstad University’s lab environment to obtain the final result. For doing our thesis, we are going to use some technical tools as hardware tools (Cisco L3 and L2 switches, Routers, IP Phones) and tools which are used for network performance monitoring<strong>,</strong> like as CommVeiw.</p>

One of the most parts of VOIP management is fault management and, in having a good fault management, finding good mechanisms to detect faults in the network have to be considered. The main focus of this project is to implement different types of fast fault recovery protocols in networks, especially networks that carry IP telephony. Having a complete understanding of some common link failure detection and fault recovery protocols, such as spanning tree protocol (STP), rapid spanning tree protocol (RSTP) and per-VLAN spanning tree protocol (PVSTP), and also having a complete understanding of three other common techniques for fault detection and fault recovery, such as hot standby routing protocol (HSRP), virtual router redundancy protocol (VRRP) and gateway load balancing protocol (GLBP) will be regarded in the project. We are going to test some fault recovery protocols which can be used in IP telephony networks and choose the best. We intend to focus on this issue in LAN environment in theoretical descriptions and practical implementations. The final outcome of the thesis is implementation in the Halmstad University’s lab environment to obtain the final result. For doing our thesis, we are going to use some technical tools as hardware tools (Cisco L3 and L2 switches, Routers, IP Phones) and tools which are used for network performance monitoring, like as CommVeiw.