ITC/USA 2008 Conference Proceedings / The Forty-Fourth Annual International Telemetering Conference and Technical Exhibition / October 27-30, 2008 / Town and Country Resort & Convention Center, San Diego, California / A common concern in telemetry post-processing environments is adequate disk storage capacity to house captured and post-processed telemetry data. In today's network environments there are many storage solutions that can be deployed to address storage needs. Recent trends in storage systems reveal movement to implement security services in storage systems. After reviewing storage options appropriate for telemetry post-processing environments; the security services such systems typically offer will also be discussed and contrasted with other third party security services that might be implemented directly on top of a networked storage system.

Cette thèse porte sur l'influence d'un jet plan pariétal sur une nappe de retour (backlayering). Cette étude peut être divisée en deux parties : une étude expérimentale et une étude numérique. Les expériences ont permis d'identifier le comportement de la nappe de retour en présence du jet. L'étude numérique débute par une validation du modèle numérique à partir des résultats de l'étude expérimentale. Une étude paramétrique est réalisée afin de déterminer l'influence du jet sur les caractéristiques et la stratification de la nappe de retour. La présence du jet plan pariétal induit une diminution de la longueur de la nappe de retour et une variation de l'épaisseur de la nappe de retour du fait de l'introduction du jet :pour une vitesse débitante fixée, l'augmentation du rapport entre la vitesse d'éjection, du jet et la vitesse débitante induit une augmentation de l'épaisseur de la nappe de retour ; pour un rapport entre la vitesse d'éjection du jet et la vitesse débitante constant, l'augmentation de la vitesse débitante conduit à un affinement de la nappe de retour. La stratification de la nappe de retour est conservée en présence du jet plan pariétal / The aim of this study is to determine the influence of a wall attached plane jet on the backlayering. This study is composed of two parts. The experimental part consists in identifying the behaviour of the backlayering in presence of the jet. The numerical part started with a validation of the numerical model using experimental study results. Then, a parametric study enables to determine the influence of the jet on the backlayering. The presence of the parietal plan jet induces a decrease of the backlayering length. Two main conclusions are achieved for the backlayering thickness change due to the jet : for a constant air flow velocity, the increase of the ratio between jet velocity and air flow velocity induces an increase of the backlayering thickness ; for a constant ratio between jet velocity and air flow velocity, the increase of the air flow velocity leads to an increase of the backlayering thickness. Finally, the backlayering stratification is maintained on the presence of the jet

Este trabalho se propôs a avaliar o desempenho de um reator com leito de biomassa aderida de fluxo ascensional em escala plena (159 \'M POT.3\') e com aeração contínua a 7 m acima da base do reator para o tratamento do esgoto sanitário. O leito do reator era formado por carvão ativado granular, carvão antracitoso e areia. O trabalho foi dividido em duas etapas. Na primeira etapa foram feitos ensaios hidrodinâmicos que consistiram da curva de expansão/fluidificação para as biopartículas do leito do reator em escala plena e do ensaio de estímulo-resposta com o traçador \'NA\'CL\'. Na segunda etapa monitorou-se o comportamento do reator com leito de biomassa aderida operado com setor 1 (sem aeração) e setor 2 (com aeração) sobrepostos e foi utilizada a recirculação para expandir o leito. As curvas de expansão/fluidificação das biopartículas constituintes do leito do reator mostraram que o leito do reator no setor 1 encontrava-se sem expansão e no setor 2 estava expandido, sob as condições de operação determinadas para esse trabalho. Do ensaio com o traçador \'NA\'CL\', obtiverem-se tempo de detenção hidráulica médio (TDHM) de 18,45 h para o ensaio sem recirculação e com compressor desligado; e de 16,90 h para o ensaio com recirculação e compressor ligado. Os TDHM obtidos foram maiores que o TDH teórico (igual a 12 h), provavelmente devido à difusão do traçador no biofilme. A Etapa II foi dividida em duas fases, na fase 1 o TDH teórico médio foi de 12 h e na fase 2 foi de 16 h. Verificou-se baixa retenção de sólidos no sistema, visto que o efluente apresentou, durante a maior parte do período de monitoramento, concentrações de sólidos totais e sólidos suspensos totais superiores as do afluente. A maior eficiência de remoção de DQO de amostra bruta aconteceu na fase 2 e foi de 32% sob taxa de carregamento de 0,86kgDQObruta/\'M POT.3\'.d. Os valores médios de DQO de ) amostras filtradas do efluente foram iguais 119 e 116 mg \'O IND.2\'/L, o que equivale a remoções médias de 54% e 63%, respectivamente para a fase 1 e 2. A maior remoção DBO de amostra filtrada no sistema aconteceu na fase 1 e foi de 91%, com o efluente apresentando concentração de 19 mg \'O IND.2\'/L. Com relação à concentração de NTK, o sistema foi pouco eficiente em sua remoção e apresentou concentrações muito baixa de nitrogênio oxidado (menores que 1 mg \'NO IND.3\'POT.-1\'/L). O grau de amonificação foi de 89%, em média, quando considerou a concentração de NTK em amostra filtrada. Concluiu-se que o sistema não é eficiente na remoção de material orgânico e de nutrientes devido ao arraste de sólidos e material biológico no efluente. / This work aimed to evaluate the performance of a full scale up-flow attached biomass bed reactor (159 \'M POT.3\') and with continuous aeration at 7 m above bottom of the reactor to sanitary sewage treatment. The bed of the reactor was constituted for activated granular carbon, anthracite carbon and sand. The work was divided into two stages. In the first stage, hydrodynamics tests were made which consisted of the expansion/fluidization curve for the bioparticles of the bed of the full scale reactor and of the test of stimulus-response with tracer \'NA\'CL\'. In the second stage, it monitored the performance of the attached biomass bed reactor operated with overlapping sector 1 (without aeration) and sector 2 (with aeration) and it utilized the recirculation to expand the bed. The expansion/fluidization curves of the bioparticles of the bed of the reactor showed that the bed of the reactor in the sector 1 was without expansion and in the sector 2 was expanded, under the operating conditions determined for this work. Of the test with \'NA\'CL\' tracer obtained mean hydraulic detention time (HDTM) of 18,45 h for the test without recirculation and compressor off; and of 16,90 h for the test with recirculation and compressor on. The obtained HDTM were bigger than the theoretical HDT (equal at 12 h), probably due to at the diffusion of trace into biofilm. The stage II was divided into two phases, in the phase 1 mean theoretical HDT was of 12 h, and in the phase 2 was of 16 h. It verified low retention of solids in the system, since the effluent presented, during most of the monitoring period, concentrations of total solids and total suspended solids higher than the influent. The greater efficiency of COD removal of raw sample occurred in phase 2 and was of 32% under loading rate of 0,86kgCODraw/\'M POT.3\'.d. The mean amounts COD of filtered samples in the effluent were equal to 119 and 116mg\'O IND.2\'/L, which is equivalent ) to mean removals of 54% and 63%, respectively for phase 1 and 2. The higher efficiency of removal BOD filtered sample in the system was in phase 1 and was of 91%, with effluent presenting concentration of 19mg \'O IND.2\'/L. With respect to the TKN concentration, the system was little efficient in its removal and presented very low concentration of oxidized nitrogen (lower than 1mg \'NO IND.3\'POT.-1\'). The degree of ammonification was 89%, on average, when considered the TKN concentration in the filtered sample. It concluded that the system is not effective removing organic material and nutrients due to drag of solids and biological material in the effluent.

<p>Problems in the area of land/biosphere-atmosphere interaction, hydrology, climate modeling etc. can be systematically organized as a study of turbulent flow in presence of boundary conditions in an increasing order of complexity. The present work is an attempt to study a few subsets of this general problem of turbulence in natural environments- in the context of neutral and thermally stratified atmospheric surface layer, the presence of a heterogeneous vegetation canopy and the interaction between air flow and a static water body in presence of flexible protruding vegetation. The main issue addressed in the context of turbulence in the atmospheric surface layer is whether it is possible to describe the macro-states of turbulence such as mean velocity and turbulent velocity variance in terms of the micro-states of the turbulent flow, i.e., a distribution of turbulent kinetic energy across a multitude of scales. This has been achieved by a `spectral budget approach' which is extended for thermal stratification scenarios as well, in the process unifying the seemingly different and unrelated theories of turbulence such as Kolmogorov's hypothesis, Heisenberg's eddy viscosity, Monin Obukhov Similarity Theory (MOST) etc. under a common framework. In the case of a more complex scenario such as presence of a vegetation canopy with edges and gaps, the question that is addressed is in what detail the turbulence is needed to be resolved in order to capture the bulk flow features such as recirculation patterns. This issue is addressed by a simple numerical framework and it has been found out that an explicit prescription of turbulence is not necessary in presence of heterogeneities such as edges and gaps where the interplay between advection, pressure gradients and drag forces are sufficient to capture the first order dynamics. This result can be very important for eddy-covariance flux calibration strategies in non-ideal environments and the developed numerical model can be used in related dispersion studies and coupled land atmosphere interaction models. For other more complex biosphere atmosphere interactions such as greenhouse gas emissions from wetlands, the interplay between air and water, often in presence of flexible aquatic vegetation, controls turbulence in water, which in turn affect the gas transfer processes. This process of wind shear induced wave-turbulent-vegetation interaction is studied for the first time in the laboratory and the state of turbulence as well as the bulk flow is found to be highly sensitive to environmental controls such as water height, wind speed, vegetation density and flexibility. This dissertation describes and gradually develops these concepts in an increasing order of complexity of boundary conditions. The first three chapters address the neutral and thermally stratified boundary layers and the last two chapters address the canopy edge problem and the air-water-vegetation experiments respectively.</p> / Dissertation

Immunohistologic investigations of wound healing in human oral mucosa require specific cell biological markers as well as consecutive small biopsies. Small specimens are ideally embedded in plastic (methylmethacrylate, MMA) resin due to their miniature size. This limits the use of antibodies for these markers. In this immunohistochemical study, the distribution of wound healing markers, e.g. cytokeratin (CK), laminin, collagen IV, vimentin, vinculin and fibronectin, were compared between semithin sections of plastic-embedded tissue and frozen sections of mucosal tissue in order to assess their use for future investigations. The antibodies against laminin, collagen IV and CK 1/2/10/11, 5/6, 13, 14, 17, 19 gave comparable staining patterns on cryostat sections of attached mucosa and on semithin sections of MMA-embedded attached mucosa. In the epithelial cell layers, the following distribution of CK immunostaining was observed: The basal cell layer was positive for CK 5/6, CK 14 and CK 19; the intermediate cell layer for CK 13, CK 17 and CK 1/2/10/11, and the superficial cell layer for CK 13 and CK 1/2/10/11. For most of these antibodies, enzyme digestion with 0.1% trypsin was adequate for demasking the antigens, except for anti-CK 14, anti-CK 17 and anti-laminin; predigestion with 0.4% pepsin in 0.01 N HCl gave similar staining results. The antibodies against vimentin, vinculin, fibronectin and CK 4 showed no affinity or a reciprocal reaction on the semithin sections. Therefore, the antibodies against CK 1/2/10/11; 5/6; 13; 14; 17, and 19, as well as the basement proteins laminin and collagen IV are deemed markers suitable on semithin sections of plastic-embedded attached oral mucosa. / Dieser Beitrag ist mit Zustimmung des Rechteinhabers aufgrund einer (DFG-geförderten) Allianz- bzw. Nationallizenz frei zugänglich.

Thesis (M.T.S.C.)--Miami University, Dept. of English, 2003. / Title from first page of PDF document. Includes bibliographical references (p. 105).

Made available in DSpace on 2015-05-14T12:09:28Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 3492986 bytes, checksum: cb918a3cba1ca29e84bfc06dbb5b6d70 (MD5) Previous issue date: 2014-03-27 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / The efficiency of removal of the biochemical oxygen demand (BOD5), chemical oxygen demand (COD), total suspended solids (TSS) and volatile suspended solid (SSV) were studied in two anaerobic reactors which received semi-settled effluent (flow limited around 3.2 times the average flow) of a septic tank, one being adhered growth (R1, with bed of gravel Nº. 4) and the other dispersed growth (R3 without support material for microbial growth). This system was installed in the University Residence at UFPB in João Pessoa-PB. The treated sewage came from two bathrooms that catered to a theoretical population of 4 inhabitants. Eleven samples of the effluent from the septic tank and the effluent from the reactors 1 and 3 were performed in the period between 28/02/2013 and 10/12/2013. The temperature and pH of the reactors were presented in the range 24.5 to 28.1 ° C and 7.3 to 8.1, respectively. Those values are from the anaerobic treatment of household sewage. The BOD5 removal efficiencies (32.8 and 27.1% in R1 and R3, respectively), COD (43.4 and 33.0% in R1 and R3, respectively), SST (56.2 and 48.0% in R1 and R3, respectively) and VSS (54.7 and 46.2% in R1 and R3, respectively) were high. It demonstrates a good applicability of these low-cost housing units which are devoid of collective sewage system in the reduction of pollution load released in water bodies systems. An analysis of variance showed no significant difference of 5% between the average of the corresponding parameters of the two reactors, suggesting that reactor that receives discharges from the attenuation peaks, there is no need of support means, which may represent a reduction in costs involved. / Estudou-se a eficiência de remoção de demanda bioquímica de oxigênio (DBO5), demanda química de oxigênio (DQO), sólidos suspensos totais (SST) e sólidos suspensos voláteis (SSV) em dois reatores anaeróbios que recebiam efluentes semi-regularizados (vazão limitada a algo em torno de 3,2 vezes a vazão média) de um tanque séptico, sendo um deles de crescimento aderido (R1, com leito de brita n°4) e o outro de crescimento disperso (R3, sem material suporte para crescimento microbiano). Este sistema foi instalado na Residência Universitária da UFPB, em João Pessoa-PB. O esgoto a ser tratado era proveniente de dois banheiros que atendiam a uma população teórica de 4 habitantes. Foram realizadas 11 coletas do efluente do tanque séptico e dos efluentes dos reatores 1 e 3, no período compreendido entre 28/02/2013 e 10/12/2013. A temperatura e o pH nos reatores se apresentaram na faixa 24,5 - 28,1 °C e 7,3 - 8,1, respectivamente, valores estes, propícios ao tratamento anaeróbio de esgotos domiciliares. As eficiências de remoção de DBO5 (32,8 e 27,1%, em R1 e R3, respectivamente), DQO (43,4 e 33,0%, em R1 e R3, respectivamente), SST (56,2 e 48,0%, em R1 e R3, respectivamente) e SSV (54,7 e 46,2%, em R1 e R3, respectivamente) foram elevadas. Isto demonstra a aplicabilidade desses sistemas de baixo custo a unidades habitacionais desprovidas de sistema coletivo de esgotamento sanitário, na redução da carga poluidora lançada em corpos aquáticos. Um teste de análise de variância demonstrou não haver diferenças significativas ao nível de 5% entre as médias dos parâmetros correspondentes nos dois reatores, sugerindo que, para reatores que recebam vazões com atenuação dos picos, não há necessidade do meio suporte, podendo representar diminuição nos custos envolvidos.

Este trabalho se propôs a avaliar o desempenho de um reator com leito de biomassa aderida de fluxo ascensional em escala plena (159 \'M POT.3\') e com aeração contínua a 7 m acima da base do reator para o tratamento do esgoto sanitário. O leito do reator era formado por carvão ativado granular, carvão antracitoso e areia. O trabalho foi dividido em duas etapas. Na primeira etapa foram feitos ensaios hidrodinâmicos que consistiram da curva de expansão/fluidificação para as biopartículas do leito do reator em escala plena e do ensaio de estímulo-resposta com o traçador \'NA\'CL\'. Na segunda etapa monitorou-se o comportamento do reator com leito de biomassa aderida operado com setor 1 (sem aeração) e setor 2 (com aeração) sobrepostos e foi utilizada a recirculação para expandir o leito. As curvas de expansão/fluidificação das biopartículas constituintes do leito do reator mostraram que o leito do reator no setor 1 encontrava-se sem expansão e no setor 2 estava expandido, sob as condições de operação determinadas para esse trabalho. Do ensaio com o traçador \'NA\'CL\', obtiverem-se tempo de detenção hidráulica médio (TDHM) de 18,45 h para o ensaio sem recirculação e com compressor desligado; e de 16,90 h para o ensaio com recirculação e compressor ligado. Os TDHM obtidos foram maiores que o TDH teórico (igual a 12 h), provavelmente devido à difusão do traçador no biofilme. A Etapa II foi dividida em duas fases, na fase 1 o TDH teórico médio foi de 12 h e na fase 2 foi de 16 h. Verificou-se baixa retenção de sólidos no sistema, visto que o efluente apresentou, durante a maior parte do período de monitoramento, concentrações de sólidos totais e sólidos suspensos totais superiores as do afluente. A maior eficiência de remoção de DQO de amostra bruta aconteceu na fase 2 e foi de 32% sob taxa de carregamento de 0,86kgDQObruta/\'M POT.3\'.d. Os valores médios de DQO de ) amostras filtradas do efluente foram iguais 119 e 116 mg \'O IND.2\'/L, o que equivale a remoções médias de 54% e 63%, respectivamente para a fase 1 e 2. A maior remoção DBO de amostra filtrada no sistema aconteceu na fase 1 e foi de 91%, com o efluente apresentando concentração de 19 mg \'O IND.2\'/L. Com relação à concentração de NTK, o sistema foi pouco eficiente em sua remoção e apresentou concentrações muito baixa de nitrogênio oxidado (menores que 1 mg \'NO IND.3\'POT.-1\'/L). O grau de amonificação foi de 89%, em média, quando considerou a concentração de NTK em amostra filtrada. Concluiu-se que o sistema não é eficiente na remoção de material orgânico e de nutrientes devido ao arraste de sólidos e material biológico no efluente. / This work aimed to evaluate the performance of a full scale up-flow attached biomass bed reactor (159 \'M POT.3\') and with continuous aeration at 7 m above bottom of the reactor to sanitary sewage treatment. The bed of the reactor was constituted for activated granular carbon, anthracite carbon and sand. The work was divided into two stages. In the first stage, hydrodynamics tests were made which consisted of the expansion/fluidization curve for the bioparticles of the bed of the full scale reactor and of the test of stimulus-response with tracer \'NA\'CL\'. In the second stage, it monitored the performance of the attached biomass bed reactor operated with overlapping sector 1 (without aeration) and sector 2 (with aeration) and it utilized the recirculation to expand the bed. The expansion/fluidization curves of the bioparticles of the bed of the reactor showed that the bed of the reactor in the sector 1 was without expansion and in the sector 2 was expanded, under the operating conditions determined for this work. Of the test with \'NA\'CL\' tracer obtained mean hydraulic detention time (HDTM) of 18,45 h for the test without recirculation and compressor off; and of 16,90 h for the test with recirculation and compressor on. The obtained HDTM were bigger than the theoretical HDT (equal at 12 h), probably due to at the diffusion of trace into biofilm. The stage II was divided into two phases, in the phase 1 mean theoretical HDT was of 12 h, and in the phase 2 was of 16 h. It verified low retention of solids in the system, since the effluent presented, during most of the monitoring period, concentrations of total solids and total suspended solids higher than the influent. The greater efficiency of COD removal of raw sample occurred in phase 2 and was of 32% under loading rate of 0,86kgCODraw/\'M POT.3\'.d. The mean amounts COD of filtered samples in the effluent were equal to 119 and 116mg\'O IND.2\'/L, which is equivalent ) to mean removals of 54% and 63%, respectively for phase 1 and 2. The higher efficiency of removal BOD filtered sample in the system was in phase 1 and was of 91%, with effluent presenting concentration of 19mg \'O IND.2\'/L. With respect to the TKN concentration, the system was little efficient in its removal and presented very low concentration of oxidized nitrogen (lower than 1mg \'NO IND.3\'POT.-1\'). The degree of ammonification was 89%, on average, when considered the TKN concentration in the filtered sample. It concluded that the system is not effective removing organic material and nutrients due to drag of solids and biological material in the effluent.

This study aimed to test the use of residual meat attached to bone structure of the chicken as raw material for the preparation of chicken pickles, using low cost and accessible ingredients. In addition, this study aimed to use different types of antioxidants in the preparation of the pickles and to evaluate the sensory acceptance and preference for these products and to estimate their shelf-life. Frozen chicken were used, with an average weight of 2.5 kg, and they were thawed and subsequently boneless to get the boneless structure with attached meat. These structures passed for a treatment with immersion and boiling in a solution of 2% sodium tripolyphosphate to meat could be more easily removed from bones. On the obtained meat, it was added spices, the broth obtained after meat boiling, vinegar and antioxidants. The product that was not the addition of antioxidants was kept as control, whereas treatments 1, 2 and 3 were the products added 0.01% BHT antioxidant, 0.5% potato peel extract and 0.5% yerba mate extract, respectively. Tests were conducted to evaluate the sensory acceptability (flavour, color, aroma and overall appearance of the products) and preference, in addition to microbiological analysis to evaluate the food security of the pickles and even physical-chemical analysis for composition and to evaluate the deterioration, during the 98 days of storage tested. The results indicated that it was possible to use this meat as raw material of chicken pickles and have a good acceptability. Pickles prepared without the addition of antioxidant and natural antioxidant of potato peel extract were more preferred than the other treatments. The average of protein, fat, moisture, ash and calcium, respectively, were 6.82%, 18.66%, 62.38%, 3.66%, 0.3%. The addition of vinegar and citric acid promoted an acid characteristic in all treatments produced, during the whole storage period. This, combined with the fact that the products were stored at a temperature of 5 °C, contributed to a good preservation of the microbiological chicken pickles. It can be observed because there was no significant increase in the counting of analyzed microorganisms (total aerobic mesophiles, yeasts and molds and coliforms at 45 °C). However, due to oxidative reaction, the estimated time of storage, at 5 °C, to pickles without added antioxidants was 28 days, to pickles which have been added BHT antioxidant, 42 days, and for pickles produced with the addition of extracts of natural antioxidants, 56 days. / Este trabalho teve como objetivo testar o aproveitamento da carne residual aderida à estrutura óssea do frango como matéria-prima para a elaboração do pickles de frango, utilizando ingredientes de baixo custo e acessíveis. Além disso, este trabalho objetivou utilizar diferentes tipos de antioxidantes na elaboração dos pickles e avaliar a aceitação e preferência sensorial destes produtos e estimar a vida-de-prateleira dos mesmos. Foram utilizados frangos congelados, com peso médio de 2,5 Kg, que foram descongelados e, posteriormente, desossados para que pudesse obter a estrutura óssea dos mesmos. Estas estruturas passaram por um tratamento de imersão e cozimento em uma solução de tripolifosfato de sódio 2% para que a carne pudesse ser removida mais facilmente dos ossos. A esta carne obtida, adicionou-se temperos, o caldo obtido no cozimento, vinagre e antioxidantes. O produto que não teve a adição de antioxidantes foi mantido como controle, enquanto que os tratamentos 1, 2 e 3 foram os produtos adicionados de 0,01% de antioxidante BHT, 0,5% de extrato de casca de batata e 0,5% de extrato de erva mate, respectivamente. Foram realizados testes sensoriais de aceitabilidade (em relação ao sabor, cor, aroma e aparência geral dos produtos) e de preferência, além de análises microbiológicas para avaliar a segurança alimentar dos pickles, e ainda análises físico-químicas de composição e de acompanhamento do grau de deterioração do produto, durante os 98 dias de armazenamento estudados. Os resultados indicaram que foi possível utilizar a carne aderida aos ossos como matéria-prima para o pickles de frango e ter uma boa aceitação sensorial. Os pickles elaborados sem adição de antioxidante e com antioxidante natural de extrato de casca de batata foram os mais preferidos. A média dos teores de proteína, lipídios, umidade, cinzas e cálcio foi, respectivamente, 6,82%, 18,66%, 62,38% 3,66% e 300 mg/100g de amostra. A adição de vinagre e ácido cítrico promoveu uma característica acidez em todos os tratamentos elaborados, durante todo o período de armazenamento. Isto, somado ao fato de que os produtos foram armazenados a uma temperatura de 5 oC, contribuiu para que não houvesse crescimento significativo na contagem dos microrganismos analisados (aeróbios mesófilos totais, bolores e leveduras e coliformes a 45 oC). Porém, devido a reação oxidativa, o tempo estimado de armazenamento, a 5 oC, para pickles de frango sem adição de antioxidantes foi de 28 dias, para pickles que tenham sofrido a adição de antioxidante BHT, 42 dias e para pickles elaborados com adição de extratos de antioxidantes naturais, 56 dias.

The sheer increase in the volume of data over the last decade has triggered research in cluster computing frameworks that enable web enterprises to extract big insights from big data. While Apache Spark defines the state of the art in big data analytics platforms for (i) exploiting data-flow and in-memory computing and (ii) for exhibiting superior scale-out performance on the commodity machines, little effort has been devoted to understanding the performance of in-memory data analytics with Spark on modern scale-up servers. This thesis characterizes the performance of in-memory data analytics with Spark on scale-up servers.Through empirical evaluation of representative benchmark workloads on a dual socket server, we have found that in-memory data analytics with Spark exhibit poor multi-core scalability beyond 12 cores due to thread level load imbalance and work-time inflation (the additional CPU time spent by threads in a multi-threaded computation beyond the CPU time required to perform the same work in a sequential computation). We have also found that workloads are bound by the latency of frequent data accesses to the memory. By enlarging input data size, application performance degrades significantly due to the substantial increase in wait time during I/O operations and garbage collection, despite 10% better instruction retirement rate (due to lower L1cache misses and higher core utilization).For data accesses, we have found that simultaneous multi-threading is effective in hiding the data latencies. We have also observed that (i) data locality on NUMA nodes can improve the performance by 10% on average,(ii) disabling next-line L1-D prefetchers can reduce the execution time by upto14%. For garbage collection impact, we match memory behavior with the garbage collector to improve the performance of applications between 1.6xto 3x and recommend using multiple small Spark executors that can provide up to 36% reduction in execution time over single large executor. Based on the characteristics of workloads, the thesis envisions near-memory and near storage hardware acceleration to improve the single-node performance of scale-out frameworks like Apache Spark. Using modeling techniques, it estimates the speed-up of 4x for Apache Spark on scale-up servers augmented with near-data accelerators. / <p>QC 20171121</p>