Leung Wai Tak. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2002. / Includes bibliographical references (leaves 58-60). / Abstracts in English and Chinese. / Acknowledgement --- p.I / Abstract --- p.II / 摘要 --- p.III / Chapter Chapter 1 --- Introduction --- p.1 / Chapter Chapter 2 --- Related Works --- p.5 / Chapter 2.1 --- Previous Works --- p.5 / Chapter 2.2 --- Contributions of this Study --- p.7 / Chapter Chapter 3 --- Architecture --- p.9 / Chapter 3.1 --- Data Placement Policy --- p.10 / Chapter 3.2 --- Retrieval and Transmission Scheduling --- p.13 / Chapter 3.3 --- Fault Tolerance --- p.20 / Chapter Chapter 4 --- Performance Modeling --- p.22 / Chapter 4.1 --- Storage Requirement --- p.22 / Chapter 4.2 --- Network Bandwidth Requirement --- p.23 / Chapter 4.3 --- Buffer Requirement --- p.24 / Chapter 4.4 --- System Response Time --- p.27 / Chapter Chapter 5 --- System Reliability --- p.29 / Chapter 5.1 --- System Failure Model --- p.29 / Chapter 5.2 --- Minimum System Repair Capability --- p.32 / Chapter 5.3 --- Redundancy Configuration --- p.35 / Chapter Chapter 6 --- System Dimensioning --- p.37 / Chapter 6.1 --- Storage Capacity --- p.38 / Chapter 6.2 --- Network Capacity --- p.38 / Chapter 6.3 --- Disk Access Bandwidth --- p.39 / Chapter 6.4 --- Buffer Requirement --- p.41 / Chapter 6.5 --- System Response Time --- p.43 / Chapter Chapter 7 --- Multiple Parity Groups --- p.45 / Chapter 7.1 --- System Failure Model --- p.47 / Chapter 7.2 --- Buffer Requirement --- p.47 / Chapter 7.3 --- System Response Time --- p.49 / Chapter 7.4 --- Redundancy Configuration --- p.49 / Chapter 7.5 --- Scalability --- p.51 / Chapter Chapter 8 --- Conclusions and Future Works --- p.53 / Appendix --- p.55 / Chapter A. --- Derivation of the Artificial Admission Delay --- p.55 / Chapter B. --- Derivation of the Receiver Buffer Requirement --- p.56 / Bibliography --- p.58

Modern computer networks allow server resources to be shared. While this multiplexing is the unsung hero of scalability and performance, the fact that clients are sharing resources and each client’s network traffic is transmitted in a larger pool of the total network traffic, poses distinct challenges for security. By adopting multiplexing so broadly, the networking and systems communities have implicitly favored performance over security. When servers multiplexing clients are compromised, the attack is able to spread by exploiting unsuspecting clients sharing the resource. Drive-by-downloads are an example of an attack where a Web server is compromised and begins distributing malware to connecting clients. As a result of using today’s many-to-one client-server network model, current approaches are inadequate at protecting the network and its resources. We propose a redesign of the modern network infrastructure. Our approach involves moving from the current many-to-one client-server model to a one-to-one client-server model. In redesigning the network, we provide a means of better accountability for traffic between clients and servers. With accountability, we enable the ability to quickly determine which client is responsible for an attack. This allows us to quickly repair the affected entities. To accomplish this accountability, we separate each client’s communication into separate flows. A flow is identified by various network features, such as IP addresses and ports. Further, instead of allowing multiple clients to be multiplexed at the same server, we use a technique that allows each client to communicate with a server that is logically separate from all other clients. Accordingly, a server compromise only effects a single client. We create a one-to-one client-server model using virtualization techniques and OpenFlow, a software-defined network (SDN) protocol. We complete our model in three phases. In the first, we deploy a physical SDN using physical machines and a commodity network switch that supports OpenFlow to gain an initial understanding of SDNs. The next phase involves implementation of Choreographer, a DNS access control mechanism, in a virtualized SDN environment for better scalability over our physical configuration. Finally, we leverage Choreographer to dynamically instantiate a server for each client and create network flows that allow a client to reach the requested server.

Denna rapport har genomförts för att utvärdera om svetsgardiner som luftflödesstyrandeelement i datorhallar ger någon energieffektivisering. Vilket utslag systemet ger dåsvetsgardiner används jämförs med utslaget när inga svetsgardiner finns med och påverkarsystemet. Får servrarna tillräckligt mycket kalluft för att kylas och behöver man tre styckenrumskylare för att kyla rummet tillräckligt? Rapporten har också genomförts för att ge enutvärdering av svetsgardinernas form och funktion. Volvo Personvagnar vill veta om det ärekonomiskt hållbart att använda svetsgardiner i sina datorhallar.Det är också önskvärt att bevisa att det blir ett bättre klimat runt servrarna när svetsgardineranvänds. Resultatet visar att temperaturskillnaden, före och efter rumskylarna samt före ochefter servrarna, ökar för luften i pilotanläggningen. Det leder i sin tur till atttemperaturskillnaden på köldbäraren för kompressorkylmaskinen kommer att minska.Eftersom rumskylarnas funktion, vid specifika fall när svetsgardiner inte existerar, leder tillbegränsning av massflödet på köldbäraren kommer massflödet att vara större under längreperioder vid tilldragna svetsgardiner. Massflödet ökar ofta mer än vad temperaturskillnadenminskar vilket leder till att mer värme kyls bort från pilotanläggningen vid tilldragnasvetsgardiner. Alltså är inte systemet rätt kalibrerat för att utnyttja svetsgardinernas funktion.Vid stor belastning på beräkningsklustren är ofta begränsningen av massflödet förköldbäraren svag. Det leder till att flödena är ungefär lika stora vid både till- och fråndragnasvetsgardiner. Då får man en minskad driftkostnad. Det går inte att bevisa en minskaddriftkostnad genom minskat kompressionsarbete vid alla belastningar och under kortareperioder p.g.a. att rumskylarna inte är korrekt kalibrerade. Inte heller en nämnvärd ökning idriftkostnad inträffar hos kompressorerna. Det verkar som att svetsgardinernas utformninginte är optimalt då kalluften från rumskylarna blåser förbi runt kanterna på gardinerna ochdärmed kyler delar av den varma luften i det varma området. För att detta problem inte skalluppstå kan man t.ex. använda längre gardiner, fästa gardinerna högre upp, fästa gardinernaäven på sidorna och vid golvet. Slutligen bör man få ett bättre resultat av att inte användauppdelad svetsgardin. Kalluftflödet till servrarna är större än vad som behövs men intetillräckligt stort att man bör stänga av rumskylaren som står i det kalla området. Ökar manrumskylarnas reglervärde på temperaturen (för när rumskylana kommer begränsa massflödetpå köldbäraren genom sig) med förslagsvis någon till några grader Celsius, (djupare analyskrävs) vid användandet av svetsgardiner, kommer systemet kunna utnyttja svetsgardinernaspotential. En mindre driftkostnad i form av att mer värme kommer kunna kylas bort tillsamma arbeta för kompressorn kommer då troligtvis inträffa. / <p>This project has been conducted to evaluate if welding curtains function as an airflow-guiding</p><p>element within a data centers will increase the energy efficiency of the system. What result</p><p>does the system show when welding curtains are being used? Compared with what the system</p><p>show when no welding curtains are being used and doesn’t affect the system is something</p><p>that’s look into. Does the servers get enough cold air to sustain its usage? Do you need three</p><p>cooling systems to cool the room sufficiently? The report has also been conducted to provide</p><p>an evaluation of the welding curtains function as well as its design. Volvo Cars wants to know</p><p>if it’s economically viable to use welding curtains in their data centers. It’s also desirable to</p><p>prove that there is a better climate around the servers when welding curtains are being used.</p><p>The result shows that the temperature difference of the air, before and after the cooling</p><p>systems and before and after the servers, increases in the data center. This also leads to that</p><p>the temperature difference of the coolant for compressor chiller will decrease. Because of the</p><p>room coolers limiting of the mass flow (in certain situations) on the coolant to a lesser extent</p><p>when welding curtains are being used, the mass flow rate will be greater for longer periods.</p><p>The mass flow often increases more than the temperature difference decreases and that leads</p><p>to that more heat is cooled from the data center when welding curtains are being used. Thus,</p><p>the system isn’t properly calibrated to utilize welding curtains to a full extent. At large loads</p><p>on the computing clusters, the room coolers affect the mass flow of the coolant to a minimum.</p><p>The result is that the mass flow is approximately equal both with and without welding</p><p>curtains in use. This leads to an increase in efficiency of compressor chiller because more heat</p><p>is being chilled with the same work of the compressor as without any welding curtains. When</p><p>welding curtains are being used you can’t prove that a reduced operating cost will occur due</p><p>to reduction of the compressor work. This is because of the room coolers are not properly</p><p>calibrated. Neither a significant increase in operating cost occurs in the compressor chiller.</p><p>An indication occurs that the design of the welding curtain is not optimal because of that the</p><p>cold air from the cooler systems blows past the edges of the curtains and thus cools some of</p><p>the hot air in the heated area. To solve this problem one can e.g. use taller welding curtains,</p><p>replace the curtains to a higher position and attach the curtains on both sides as well as on the</p><p>floor. Also the use of a welding curtain that isn’t divided in two or more parts is an idea that</p><p>can be implemented. The cold airflow to the servers flows at a larger rate than necessary but</p><p>not that sufficient that one, in the cold area, should put the cooling system in idle mode. If you</p><p>increase the cooling systems temperature set-value (temperature set-value for when the cooler</p><p>limits the mass flow through it) by a few degrees Celsius when welding curtains are being</p><p>used (a deeper analysis is needed), the system will hopefully be able to utilize the potential of</p><p>the welding curtains. I.e. increased efficiency of the compressor chiller because more heat is</p><p>being chilled to the same work of the compressor.</p><p>Program: Energiingenjörsprogrammet med inriktning mot eldistribution</p>

Ip Tak Shun. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2005. / Includes bibliographical references (leaves 95-101). / Abstracts in English and Chinese. / Abstract --- p.i / Acknowledgement --- p.iv / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Background --- p.1 / Chapter 1.1.1 --- Media Streaming --- p.1 / Chapter 1.1.2 --- Incentive Mechanism --- p.2 / Chapter 1.2 --- Cooperative and Incentive-based Proxy-and-Client Caching --- p.4 / Chapter 1.2.1 --- Cooperative Proxy-and-Client Caching --- p.4 / Chapter 1.2.2 --- Revenue-Rewarding Mechanism --- p.5 / Chapter 1.3 --- Thesis Contribution --- p.6 / Chapter 1.4 --- Thesis Organization --- p.7 / Chapter 2 --- Related Work --- p.9 / Chapter 2.1 --- Media Streaming --- p.9 / Chapter 2.2 --- Incentive Mechanism --- p.11 / Chapter 2.3 --- Resource Pricing --- p.14 / Chapter 3 --- Cooperative Proxy-and-Client Caching --- p.16 / Chapter 3.1 --- Overview of the COPACC System --- p.16 / Chapter 3.2 --- Optimal Cache Allocation (CAP) --- p.21 / Chapter 3.2.1 --- Single Proxy with Client Caching --- p.21 / Chapter 3.2.2 --- Multiple Proxies with Client Caching --- p.24 / Chapter 3.2.3 --- Cost Function with Suffix Multicast --- p.26 / Chapter 3.3 --- Cooperative Proxy-Client Caching Protocol --- p.28 / Chapter 3.3.1 --- Cache Allocation and Organization --- p.29 / Chapter 3.3.2 --- Cache Lookup and Retrieval --- p.30 / Chapter 3.3.3 --- Client Access and Integrity Verification --- p.30 / Chapter 3.4 --- Performance Evaluation --- p.33 / Chapter 3.4.1 --- Effectiveness of Cooperative Proxy and Client Caching --- p.34 / Chapter 3.4.2 --- Robustness --- p.37 / Chapter 3.4.3 --- Scalability and Control Overhead --- p.38 / Chapter 3.4.4 --- Sensitivity to Network Topologies --- p.40 / Chapter 4 --- Revenue-Rewarding Mechanism --- p.43 / Chapter 4.1 --- System Model --- p.44 / Chapter 4.1.1 --- System Overview --- p.44 / Chapter 4.1.2 --- System Formulation --- p.47 / Chapter 4.2 --- Resource Allocation Game --- p.50 / Chapter 4.2.1 --- Non-Cooperative Game --- p.50 / Chapter 4.2.2 --- Profit Maximizing Game --- p.52 / Chapter 4.2.3 --- Utility Maximizing Game --- p.61 / Chapter 4.3 --- Performance Evaluation --- p.74 / Chapter 4.3.1 --- Convergence --- p.76 / Chapter 4.3.2 --- Participation Incentive --- p.77 / Chapter 4.3.3 --- Cost effectiveness --- p.85 / Chapter 5 --- Conclusion --- p.87 / Chapter A --- NP-Hardness of the CAP problem --- p.90 / Chapter B --- Optimality of the Greedy Algorithm --- p.92 / Bibliography --- p.95

Leung Siu Kuen. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2005. / Includes bibliographical references (leaves 67-68). / Abstracts in English and Chinese. / Abstract --- p.i / Acknowledgement --- p.iii / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Motivation --- p.1 / Chapter 1.1.1 --- A Story --- p.1 / Chapter 1.2 --- Overview --- p.2 / Chapter 1.2.1 --- Introduction of SQL Injection --- p.4 / Chapter 1.3 --- The importance of SQL Injection --- p.6 / Chapter 1.4 --- Thesis organization --- p.8 / Chapter 2 --- Background --- p.10 / Chapter 2.1 --- Flow of web applications using DBMS --- p.10 / Chapter 2.2 --- Structure of DBMS --- p.12 / Chapter 2.2.1 --- Tables --- p.12 / Chapter 2.2.2 --- Columns --- p.12 / Chapter 2.2.3 --- Rows --- p.12 / Chapter 2.3 --- SQL Syntax --- p.13 / Chapter 2.3.1 --- SELECT --- p.13 / Chapter 2.3.2 --- AND/OR --- p.14 / Chapter 2.3.3 --- INSERT --- p.15 / Chapter 2.3.4 --- UPDATE --- p.16 / Chapter 2.3.5 --- DELETE --- p.17 / Chapter 2.3.6 --- UNION --- p.18 / Chapter 3 --- Details of SQL Injection --- p.20 / Chapter 3.1 --- Basic SELECT Injection --- p.20 / Chapter 3.2 --- Advanced SELECT Injection --- p.23 / Chapter 3.2.1 --- Single Line Comment (--) --- p.23 / Chapter 3.2.2 --- Guessing the number of columns in a table --- p.23 / Chapter 3.2.3 --- Guessing the column name of a table (Easy one) --- p.26 / Chapter 3.2.4 --- Guessing the column name of a table (Difficult one) . --- p.27 / Chapter 3.3 --- UPDATE Injection --- p.29 / Chapter 3.4 --- Other Attacks --- p.30 / Chapter 4 --- Current Defenses --- p.32 / Chapter 4.1 --- Causes of SQL Injection attacks --- p.32 / Chapter 4.2 --- Defense Methods --- p.33 / Chapter 4.2.1 --- Defensive Programming --- p.34 / Chapter 4.2.2 --- hiding the error messages --- p.35 / Chapter 4.2.3 --- Filtering out the dangerous characters --- p.35 / Chapter 4.2.4 --- Using pre-complied SQL statements --- p.36 / Chapter 4.2.5 --- Checking for tautologies in SQL statements --- p.37 / Chapter 4.2.6 --- Instruction set randomization --- p.38 / Chapter 4.2.7 --- Building the query model --- p.40 / Chapter 5 --- Proposed Solution --- p.43 / Chapter 5.1 --- Introduction --- p.43 / Chapter 5.2 --- Natures of SQL Injection --- p.43 / Chapter 5.3 --- Our proposed system --- p.44 / Chapter 5.3.1 --- Features of the system --- p.44 / Chapter 5.3.2 --- Stage 1 - Checking with current signatures --- p.45 / Chapter 5.3.3 --- Stage 2 - SQL Server Query --- p.45 / Chapter 5.3.4 --- Stage 3 - Error Triggering --- p.46 / Chapter 5.3.5 --- Stage 4 - Alarm --- p.50 / Chapter 5.3.6 --- Stage 5 - Learning --- p.50 / Chapter 5.4 --- Examples --- p.51 / Chapter 5.4.1 --- Defensing BASIC SELECT Injection --- p.52 / Chapter 5.4.2 --- Defensing Advanced SELECT Injection --- p.52 / Chapter 5.4.3 --- Defensing UPDATE Injection --- p.57 / Chapter 5.5 --- Comparison --- p.59 / Chapter 6 --- Conclusion --- p.62 / Chapter A --- Commonly used table and column names --- p.64 / Chapter A.1 --- Commonly used table names for system management --- p.64 / Chapter A.2 --- Commonly used column names for password storage --- p.65 / Chapter A.3 --- Commonly used column names for username storage --- p.66 / Bibliography --- p.67

Esta dissertação de mestrado apresenta a implementação e validação de um modelo de servidor web que divide o funcionamento de o servidor web em quatro módulos onde cada um desses módulos é responsável por uma etapa que a requisição percorre ao longo de seu processamento. Esses módulos são: atendimento da requisição (módulo 1), transferência do arquivo para a memória principal (módulo 2), processamento de requisições dinâmicas (módulo 3) e envio do arquivo ao cliente (módulo 4). Esses quatro módulos são interligados e são alimentados primeiramente por uma carga inicial gerada pelo gerador de cargas W4Gen e passa obrigatoriamente, nessa ordem, pelo módulo 1, módulo 2 e módulo 4. O módulo 3 só é utilizado quando se trata de uma requisição dinâmica. Ao ser atendido por um dos módulos, é atribuído um tempo de execução (leia-se tempo que a requisição toma para ser processada por esse módulo). Esses tempos foram baseados em trabalhos que fizeram benchmarks em servidores web reais. Os resultados alcançados com o desenvolvimento deste trabalho visam principalmente integrar-se aos trabalhos de simulação envolvendo servidores web do grupo de Sistemas Distribuídos e Programação Concorrente (LaSDPC) e com isso alcançar resultados próximos a resultados aplicados em servidores web reais / This work presents the implementation and validation of a web server model that divides the web server functions into four modules and each one of these is responsible for an execution step, in which a request goes through during processing. These modules are: request serving (module 1); file transferring to the main memory (module 2); dynamic request processing (module 3); and client file sending (module 4). These four modules are linked and the W4Gen generates the initial loads. These loads run in the order as follows: modules 1, 2 and 4. The module 3 is used only when the request is dynamic. When a request is served by a module, an execution time is defined (i.e. the time used by the request in the module), which is based on real world web servers benchmark\'s. The results obtained in the work aims to be integrated to other projects conduced in the Distributed System and Concurrent Programming group (LaSDPC) in order to reach results close to real world web servers

Diplomová práce popisuje vývoj softvéru na platformě Visual Studio 2005 Team System (VSTS). Práce klade důraz především na popis funkcionality umožňující efektivní týmovou spolupráci a na popis nových inovativních rysů. Práce nepopisuje pouze pozitivní rysy produktu, identifikuje také základní nedostatky současné verze, na základě zkušeností vývojových týmů používajících VSTS a subjektivních názorech autora.

Abstract This thesis deals with the process of server consolidation, which is tightly connected to IT infrastructure of virtually any organization. Moreover it is dealt in some respect with the way the organization runs and with linking these two particular areas together. The goal of this thesis is to identify consequences or impacts of server consolidation process on the way the organization functions and on methods how to measure these impacts. One of the steps towards this goal is unmistakably analyzing the planning and deployment of server consolidation including virtualization technology. Server consolidation is characterized as a process leading to simplified and more effective IT infrastructure of an organization. To analyze the IT infrastructure in complete and systematic way it is used framework called IT Infrastructure Library respectively part of it called ICT Infrastructure Management. Because of that linking of the server consolidation process progress to organization's global goals is granted. Contribution of this thesis can be seen in complex analysis of the consequences of the server consolidation process to the functioning of an organization. Especially in areas that are usually not being considered in connection with server consolidation, which is considered to be IT matter, such as human resources and social enviroment. Key words: server consolidation, server consolidation consequences, ICT Infrastructure Management, Virtualization

This diploma thesis is focused on operation and maintenance of BI systems based on Microsoft's SQL Server 2008 R2 platform. This work covers regular activities related to BI system management that means monitoring, performance tuning, maintenance, administration, backup and recovery tasks. System management is based mainly on SQL Server's native management tools along with Microsoft's applications System Center: Essentials and System Center: Data Protection Manager. The main goal of this work is to create BI system operations and maintenance methodology in cooperation with Clever Decision which is specialized in BI area. Partial goals are to design a way of monitoring components of BI system, including to outline backup and recovery posibilities, maintenance and performance tuning of the SQL Server platform. Each chapter contains methods, best practices, summary tables and appropriate tools which help DBA to achieve the specified goals. The main contribution of this work is operations and maintenance methodology created according to Clever Decision's needs. Among the other contributions of this thesis also belongs the definition of using tools for monitoring and maintenance and their relations to the SQL Server's components. The work also contains simple backup and performance tuning methodology for SQL Server. At least, this diploma thesis represents comprehensive material which is strictly focused on OLAP databases operation and maintenance.

Esta dissertação apresenta a construção de um protótipo de servidor Web distribuído, baseado no modelo de servidor Web com diferenciação de serviços (SWDS) e a implementação e avaliação de algoritmos de seleção, utilizando o conceito de controle de congestionamento para requisições HTTP. Com isso, além de implementar uma plataforma de testes, este trabalho também avalia o comportamento de dois algoritmos de controle de congestionamento. Os dois algoritmos estudados são chamados de Drop Tail e RED (Random Early Detection), no qual são bastante difundidos na literatura científica e aplicados em redes de computadores. Os resultados obtidos demostram que, apesar das particularidades de cada algoritmo, existe uma grande relação entre tempo de resposta e a quantidade de requisições aceitas / This MSc dissertation presents the implementation of a prototype for a distributed web server based on the SWDS, a model for a web server with service differentiation, and the implementation and evaluation of selection algorithms adopting the concept of congestion control for HTTP requests. Thus, besides implementing a test platform this work also evaluates the behavior of two congestion control algorithms. The two algorithms studied are the Drop Tail and the RED (Random Early Detection), which are frequently discussed in the scientific literature and widely applied in computer networks. The results obtained show that, although the particularities of each algorithm, there is a strong relation between the response times and the amount of requests accepted in the server