Thesis (M.S.)--University of Delaware, 2009. / Principal faculty advisor: Daniel J. Leathers, Dept. of Geography. Includes bibliographical references.

Submitted by Luiz Felipe Barbosa (luiz.fbabreu2@ufpe.br) on 2015-03-12T13:23:25Z No. of bitstreams: 2 Dissertaçao Rafael de Souza.pdf: 5024167 bytes, checksum: d0690bd2c55d5a6ca5e7bbe243777c13 (MD5) license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) / Approved for entry into archive by Daniella Sodre (daniella.sodre@ufpe.br) on 2015-03-13T13:14:31Z (GMT) No. of bitstreams: 2 Dissertaçao Rafael de Souza.pdf: 5024167 bytes, checksum: d0690bd2c55d5a6ca5e7bbe243777c13 (MD5) license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) / Made available in DSpace on 2015-03-13T13:14:31Z (GMT). No. of bitstreams: 2 Dissertaçao Rafael de Souza.pdf: 5024167 bytes, checksum: d0690bd2c55d5a6ca5e7bbe243777c13 (MD5) license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Previous issue date: 2013-08-30 / FACEPE / Os data centers estão em constante crescimento, a fim de atender às demandas de novas tecnologias, como cloud computing e e-commerce. Em tais paradigmas, períodos de inatividade podem levar a perdas financeiras de milhões de dólares e danificar permanentemente a reputação de uma empresa. Vários fatores afetam a disponibilidade de sistemas de TI em data center, entre eles, as variações de temperatura ambiente. Este trabalho propõe modelos para contemplar o efeito de variação de temperatura nas infraestruturas do data center. Além destes modelos, também é proposta uma metodologia para auxiliar na elaboração e avaliação dos diferentes cenários. Esta metodologia permite a análise através de vários modelos intermediários que ajudam a encontrar o efeito de variação de temperatura na disponibilidade das infraestruturas de TI do data center. Nesta abordagem, a avaliação é realizada com modelos de rede de Petri estocásticas, modelo de Arrhenius, modelo de energia, e diagrama de blocos de confiabilidade. Por fim, três estudos de casos real, bem como, exemplos são apresentados com a finalidade de mostrar a aplicabilidade deste trabalho.

Made available in DSpace on 2018-08-02T00:01:58Z (GMT). No. of bitstreams: 1 tese_3847_Tese - Gilmar Luiz Vassoler.pdf: 9249571 bytes, checksum: 33940b4014472a53a7362b39a8c9af3a (MD5) Previous issue date: 2015-05-22 / sta tese apresenta duas contribuições para as redes de data center centrado em servidores. A primeira, intitulada Twin Datacenter Interconnection Topology, foca nos aspectos topológicos e demostra como o uso de Grafos Gêmeos podem potencialmente reduzir o custo e garantir alta escalabilidade, tolerância a falhas, resiliência e desempenho. A segunda, intitulada TRIIIAD TRIple-Layered Intelligent and Integrated Architecture for Datacenters, foca no acoplamento entre a orquestração da nuvem e o controle da rede. A TRIIIAD é composta por três camadas horizontais e um plano vertical de controle, gerência e orquestração. A camada superior representa a nuvem do data center. A camada intermediária fornece um mecanismo leve e eficiente para roteamento e encaminhamento dos dados. A camada inferior funciona como um comutador óptico distribuído. Finalmente, o plano vertical alinha o funcionamento das três camadas e as mantem agnósticas entre si. Este plano foi viabilizado por um controlador SDN aumentado, que se integrou à dinâmica da orquestração, de forma a manter a consistência entre as informações da rede e as decisões tomadas na camada de virtualização

Energy efficient infrastructures or green IT (Information Technology) has recently become a hot button issue for most corporations as they strive to eliminate every inefficiency from their enterprise IT systems and save capital and operational costs. Vendors of IT equipment now compete on the power efficiency of their devices, and as a result, many of the new equipment models are indeed more energy efficient. Various studies have estimated the annual electricity consumed by networking devices in the U.S. in the range of 6 - 20 Terra Watt hours. Our research has the potential to make promising solutions solve those overuses of electricity. An energy-efficient data center network architecture which can lower the energy consumption is highly desirable. First of all, we propose a fair bandwidth allocation algorithm which adopts the max-min fairness principle to decrease power consumption on packet switch fabric interconnects. Specifically, we include power aware computing factor as high power dissipation in switches which is fast turning into a key problem, owing to increasing line speeds and decreasing chip sizes. This efficient algorithm could not only reduce the convergence iterations but also lower processing power utilization on switch fabric interconnects. Secondly, we study the deployment strategy of multicast switches in hybrid mode in energy-aware data center network: a case of famous Fat-tree topology. The objective is to find the best location to deploy multicast switch not only to achieve optimal bandwidth utilization but also minimize power consumption. We show that it is possible to easily achieve nearly 50% of energy consumption after applying our proposed algorithm. Finally, although there exists a number of energy optimization solutions for DCNs, they consider only either the hosts or network, but not both. We propose a joint optimization scheme that simultaneously optimizes virtual machine (VM) placement and network flow routing to maximize energy savings. The simulation results fully demonstrate that our design outperforms existing host- or network-only optimization solutions, and well approximates the ideal but NP-complete linear program. To sum up, this study could be crucial for guiding future eco-friendly data center network that deploy our algorithm on four major layers (with reference to OSI seven layers) which are physical, data link, network and application layer to benefit power consumption in green data center.

Environmental monitoring using wireless sensors plays a key role in detecting hotspots or over-cooling conditions in a data center (DC). Despite a myriad of Data Center Wireless Sensor Network (DCWSN) solutions in literature, their adoption in DCs is scarce due to four challenges: low reliability, short battery lifetime, lack of adaptability, and labour intensive deployment. The main objective of this research is to address these challenges in our specifically designed hierarchical DCWSN, called Low Energy Monitoring Network (LEMoNet). LEMoNet is a two-tier protocol, which features Bluetooth Low Energy (BLE) for sensors communication in the first tier. It leverages multi-gateway packet reception in its second tier to mitigate the unreliability of BLE. The protocol has been experimentally validated in a small DC and evaluated by simulations in a midsize DC. However, since the main application of DCWSNs is in colocation and large DCs, an affordable and fast approach is still required to assess LEMoNet in large scale. As the first contribution, we develop an analytical model to characterize its scalability and energy efficiency in a given network topology. The accuracy of the model is validated through extensive event-driven simulations. Evaluation results show that LEMoNet can achieve high reliability in a network of 4800 nodes at a duty cycle of 15s. To achieve the network adaptability, we introduce and design SoftBLE, a Software-Defined Networking (SDN) based framework that provides controllability to the network. It takes advantages of advanced control knobs recently available in BLE protocol stacks. SoftBLE is complemented by two orchestration algorithms to optimize gateway and sensor parameters based on run-time measurements. Evaluation results from both an experimental testbed and a large-scale simulation study show that using SoftBLE, sensors consume 70% less power in data collection compared to those in baseline approaches while achieving the Packet Reception Rate (PRR) no less than 99.9%. One of its main steps of DCWSN commissioning is sensor localization, which is labour-intensive if is driven manually. To streamline the process, we devise a novel approach for automated sensor mapping. Since Radio Frequency (RF) alone is not a reliable data source for sensor localization in harsh and multi-path rich environments such as a DCs, we investigate using non-RF alternatives. Thermal Piloting is a classification model to correlate temperature sensor measurements with the expected thermal values at their locations. It achieves an average localization error of 0.64 meters in a modular DC testbed. The idea is further improved by a multimodal approach that incorporates pairwise Received Signal Strength (RSS) measurements of RF signals. The problem is formulated as Weighted Graph Matching (WGM) between an analytical graph and an experimental graph. A parallel algorithm is proposed to find heuristic solutions to this NP-hard problem, which is 30% more accurate than the baselines. The evaluation in a modular DC testbed shows that the localization errors using multi-modality are less than one-third of that of using thermal data alone. / Thesis / Candidate in Philosophy

Live migration of Virtual Machines (VMs) has significantly improved the flexibility of modern Data Centers (DCs). Ideally, live migration ought to be seamless which in turn raises challenges on how to minimize service disruption and avoid performance degradation. To address these challenges, a comprehensive support from the underlying network is required. However, legacy DC networks fall short to help as they take a reactive approach to live migration procedure. Moreover, the complexity and inflexibility of legacy DC networks make it difficult to deploy, manage, and improve network technologies that DC providers may need to use for migration. In this thesis, we explore the application of Software Defined Networking (SDN) paradigm for making live VM migration more seamless. Exploiting the characteristics of SDN such as its centralized view on network states, we contribute to the body of knowledge by enhancing the quality of intra- and inter-DC live migration. Firstly, for intra-DC migration, we provide an SDN-based solution which minimizes the service disruption by employing OpenFlow-based resiliency mechanisms to prepare a DC network for migration proactively. Secondly, we improve the inter-DC live migration by accelerating the network convergence through announcing the migration in the control plane using MP-BGP protocol. Further, our proposed framework resolves the sub-optimal routing problem by conducting the gateway functionality at the SDN controller. Finally, with the ultimate goal of improving the inter-DC migration, we develop an SDN-based framework which automates the deployment, improves the management, enhances the performance, and increases the scalability of interconnections among DCs. / Live migration of Virtual Machines (VMs) has significantly improved the flexibility of modern Data Centers (DCs). Ideally, live migration ought to be seamless which requires a comprehensive support from the underlying network. However, legacy DC networks fall short to address the challenges of migration due to their inflexible and decentralized characteristics. In contrast, Software Defined Networking (SDN) is a new networking paradigm, which has the potential to improve the live migration thanks to its comprehensive view over the network, flexible structure, and its close integration with DC management infrastructures. This thesis investigates networking challenges of short and long-haul live VM migration in SDN-based DCs. We propose solutions to make the intra- and inter-DC live migration procedures more seamless. Furthermore, our proposed SDN-based framework for inter-DC migration improves the management, enhances the performance, and increases the scalability of interconnections among DCs. / HITS, 4707

La congestion dans les réseaux est un phénomène qui peut influer sur la qualité de service ressentie par les utilisateurs. L’augmentation continue du trafic sur l’internet rend le phénomène de congestion un problème auquel l’opérateur doit répondre pour satisfaire ses clients. Les solutions historiques à la congestion pour un opérateur, comme le surdimensionnement des liens de son infrastructure, ne sont plus aujourd’hui viables. Avec l’évolution de l’architecture des réseaux et l’arrivée de nouvelles applications sur l’internet, de nouveaux paradigmes de contrôle de congestion sont à envisager pour répondre aux attentes des utilisateurs du réseau de l’opérateur. Dans cette thèse, nous examinons les nouvelles approches proposées pour le contrôle de congestion dans le réseau d’un opérateur. Nous proposons une évaluation de ces approches à travers des simulations, ce qui nous permet d’estimer leur efficacité et leur potentiel à être déployés et opérationnels dans le contexte d’internet, ainsi que de se rendre compte des défis qu’il faut relever pour atteindre cet objectif. Nous proposons également des solutions de contrôle de congestion dans des environnements nouveaux tels que les architectures Software Defined Networking et le cloud déployé sur un ou plusieurs data centers, où la congestion est à surveiller pour maintenir la qualité des services cloud offerts aux clients. Pour appuyer nos propositions d’architectures de contrôle de congestion, nous présentons des plateformes expérimentales qui démontrent le fonctionnement et le potentiel de nos solutions. / Network congestion is a phenomenon that can influence the quality of service experienced by the users. The continuous increase of internet traffic makes this phenomenon an issue that should be addressed by the network operator to satisfy its clients. The usual solutions to congestion, such as overdimensioning the infrastructure, are not viable anymore. With the evolution of the network architecture and the emergence of new internet applications, new paradigms for congestion control have to be considered as a response to the expectations of network users. In this thesis, we examine new approaches to congestion control in an operator’s network. We propose an evaluation of these approaches through simulations, which allows us to estimate their potential to be deployed and used over the internet, and allows us to be aware of the challenges in order to achieve this goal. We also provide solutions for congestion control in new environments such as Software- Defined Networking architectures and cloud computing deployed over many data centers, where congestion is to be monitored to maintain the quality of cloud services to its users. To support our proposals for congestion control architectures, we present experimental platforms that demonstrate the feasibility of our solutions.

Today, data centers host a wide variety of applications which generate a mix of diverse internal data center traffic. In a data center environment 90% of the traffic flows, though they constitute only 10% of the data carried around, are short flows with sizes up to a maximum of 1MB. The rest 10% constitute long flows with sizes in the range of 1MB to 1GB. Throughput matters for the long flows whereas short flows are latency sensitive. This thesis studies various data center transport mechanisms aimed at either improving flow completion time for short flows or throughput for long flows. Thesis puts forth two data center transport mechanisms: (1) for improving flow completion time for short flows (2) for improving throughput for long flows. The first data center transport mechanism proposed in this thesis, FA-DCTCP (Flow Aware DCTCP), is based on Data Center Transmission Control Protocol (DCTCP). DCTCP is a Transmission Control Protocol (TCP) variant for data centers pioneered by Microsoft, which is being deployed widely in data centers today. DCTCP congestion control algorithm treats short flows and long flows equally. This thesis demonstrate that, treating them differently by reducing the congestion window for short flows at a lower rate compared to long flows, at the onset of congestion, 99th percentile of flow completion time for short flows could be improved by up to 32.5%, thereby reducing their tail latency by up to 32.5%. As per data center traffic measurement studies, data center internal traffic often exhibit predefined patterns with respect to the traffic flow mix. The second data center transport mechanism proposed in this thesis shows that, insights into the internal data center traffic composition could be leveraged to achieve better throughput for long flows. The mechanism for the same is implemented by adopting the Software Defined Networking paradigm, which offers the ability to dynamically adapt network configuration parameters based on network observations. The proposed solution achieves up to 22% improvement in long flow throughput, by dynamically adjusting network element’s QoS configurations, based on the observed traffic pattern.

No description available.

Computational fluid dynamics and heat transfer (CFD/HT) models have been employed as the dominant technique for the design and optimization of both new and existing data centers. Inviscid modeling has shown great speed advantages over the full Navier-Stokes CFD/HT models (over 20 times faster), but is incapable of capturing the physics in the viscous regions of the domain. A coupled inviscid-viscous solution method (CIVSM) for bounded domains has been developed in order to increase both the solution speed and accuracy of CFD/HT models. The methodology consists of an iterative solution technique that divides the full domain into multiple regions consisting of at least one set of viscous, inviscid, and interface regions. The full steady, Reynolds-Averaged Navier-Stokes (RANS) equations with turbulence modeling are used to solve the viscous domain, while the inviscid domain is solved using the Euler equations. By combining the increased speed of the inviscid solver in the inviscid regions, along with the viscous solver’s ability to capture the turbulent flow physics in the viscous regions, a faster and potentially more accurate solution can be obtained for bounded domains that contain inviscid regions which encompass more than half of the domain, such as data centers.