In the modern computing landscape, the challenge of tuning software systems is exacerbated by the necessity to accommodate multiple divergent execution environments and stakeholders. Achieving optimal performance requires a different configuration for every combination of hardware setups and business requirements. In addition, the state of the art in system tuning can involve complex statistical models and tools which require deep expertise not commonly possessed by the average software engineer. As an alternative approach to performance tuning, this thesis puts forward the use of sonification-conveying information via non-speech audio-to aid software engineers in tuning complex systems. In particular, this thesis designs, develops, and evaluates a tuning system that interactively (i.e., in response to user actions) sonifies the performance metrics of a computer system. This thesis demonstrates that interactive sonification can effectively guide software engineers through performance tuning of a computer system. To that end, a scientific survey determined which sound characteristics (e.g., loudness, panning, pitch, tempo, etc.) are best suited to express information to the engineer. These characteristics were used to create a proof-of-concept tuning system that was applied to tune the parameters of a real world enterprise application server. Equipped with the tuning system, engineers-not experts in enterprise computing nor performance tuning-were able to tune the server, so that its resulting performance surpasses that exhibited under the standard configuration. The results indicate that sound-based tuning approaches can provide valuable solutions to the challenges of configuring complex computer systems. / Master of Science

Automatic physical database design tools rely on ``what-if'' interfaces to the query optimizer to estimate the execution time of the training query workload under different candidate physical designs. The tools use these what-if interfaces to recommend physical designs that minimize the estimated execution time of the input training workload. Minimizing estimated execution time alone can lead to designs that are not robust to query optimizer errors and workload changes. In particular, if the optimizer makes errors in estimating the execution time of the workload queries, then the recommended physical design may actually degrade the performance of these queries. In this sense, the physical design is risky. Furthermore, if the production queries are slightly different from the training queries, the recommended physical design may not benefit them at all. In this sense, the physical design is not general. We define Risk and Generality as two new measures aimed at evaluating the robustness of a proposed physical database design, and we show how to extend the objective function being optimized by a generic physical design tool to take these measures into account. We have implemented a physical design advisor in PostqreSQL, and we use it to experimentally demonstrate the usefulness of our approach. We show that our two new metrics result in physical designs that are more robust, which means that the user can implement them with a higher degree of confidence. This is particularly important as we move towards truly zero-administration database systems in which there is not the possibility for a DBA to vet the recommendations of the physical design tool before applying them.

Automatic physical database design tools rely on ``what-if'' interfaces to the query optimizer to estimate the execution time of the training query workload under different candidate physical designs. The tools use these what-if interfaces to recommend physical designs that minimize the estimated execution time of the input training workload. Minimizing estimated execution time alone can lead to designs that are not robust to query optimizer errors and workload changes. In particular, if the optimizer makes errors in estimating the execution time of the workload queries, then the recommended physical design may actually degrade the performance of these queries. In this sense, the physical design is risky. Furthermore, if the production queries are slightly different from the training queries, the recommended physical design may not benefit them at all. In this sense, the physical design is not general. We define Risk and Generality as two new measures aimed at evaluating the robustness of a proposed physical database design, and we show how to extend the objective function being optimized by a generic physical design tool to take these measures into account. We have implemented a physical design advisor in PostqreSQL, and we use it to experimentally demonstrate the usefulness of our approach. We show that our two new metrics result in physical designs that are more robust, which means that the user can implement them with a higher degree of confidence. This is particularly important as we move towards truly zero-administration database systems in which there is not the possibility for a DBA to vet the recommendations of the physical design tool before applying them.

Esant dideliems saugomų duomenų kiekiams, yra svarbus našus jų apdorojimas, taigi, vartotojams reikia vis didesnio duomenų bazių našumo. Šiame darbe sprendžiama problema, kaip paskatinti duomenų bazes veikti greičiau, kai duomenų bazių lentelės turi labai daug įrašų. Todėl skiriamas dėmesys duomenų bazių spartos derinimui, ar duomenų bazių spartos optimizavimui. Išnagrinėjus duomenų bazių esamus spartinimo metodus ir priežastis, kurios mažina našumą, yra siūlomas metodas, kuris leidžia sparčiau apdoroti ir filtruoti duomenis bei greičiau pateikti vartotojui užklausos rezultatą. Darbui atlikti pasirinkta MS SQL Server duomenų bazių valdymo sistema. Eksperimento metu atliktas užklausų greičio tyrimas, palyginant sudarytą metodą su virtualių lentelių metodu. / When the amount of stored data is growing, it is very important to get them fast and users are expecting to see how database performance is rising. Using database performance tuning, or database performance optimization, it is possible to make a database system run faster. In this paper after analysis of database performance optimization and performance tuning methods was suggested a method which enables to process data from database more quick and to user to get query result faster. To perform the research the MS SQL Server Database Management System was chosen. The experiment was performed in order to evaluate how method works. The experiment results show that compared with views, this method has better query performance.

Duomenų kubo konstravimas yra laikui ir kompiuteriniams resursams imlus procesas. Nepaisant to, šis darbas turi būti atliktas norint pasinaudoti greitų užklausų iš ypatingai didelių OLAP kubų teikiamais privalumais . Telekomunikacijų bendrovės surenka didelius duomenų kiekius apie įvykius telekomunikaciniuose tinkluose. Kiekviena duomenų porcija aprašo daug informacijos (pavyzdžiui: paslaugos tipą, iniciatorių, gavėją, pradžios laiką, trukmę, perduotų duomenų kiekį, skambučio kryptį, kainą, tinklo sąsajos adresą ir t.t.). Mobiliojo ryšio rinkoje yra įprasta apdovanoti kiekvieną abonentą tam tikru prizu (pinigais, nuolaidomis ar nauju mobiliuoju telefonu) mainais į 24 mėnesių sutartį naudotis konkretaus operatoriaus paslaugomis. Taigi kas 24 mėnesius abonentas turi galimybę pakeisti paslaugos teikėją. Tam, kad ryšio operatorius išlaikytų savo klientus, už sutarties pratęsimą taip pat turi pasiūlyti dovaną. Kad būtų galima tai atlikti nepatiriant finansinių nuostolių – mobiliojo ryšio operatorius privalo žinoti kiekvieno abonento naudojimosi paslaugomis statistiką. Šiame dokumente aprašoma pora būtų kaip pakeisti duomenų pirminį vaizdą (struktūrą ir sudėtį) siekiant pagreitinti duomenų kubų konstravimo procesą. Vienas šių metodų – duomenų agregavimas iki didžiausio, vis dar tinkamo analizei, lygio. Kitas metodas – tai lėtai kintančių kubo dimensijų sintezavimas taip sumažinant kubo dydį ir pagreitinant jo kūrimą. / Data cube pre computing is time and computer resources consuming task. In spite of this it needs to be done in order to construct an OLAP cube to take advantage of fast querying in data sets enormous in its sizes. Telecommunication industries collect huge amount of data about events in its networks. Every data portion holds a lot of information (i.e. service type, originator, receiver, time for start, duration, data volume, calling direction, cost, network interface address, etc.). In mobile telecommunication industries it is common to award each customer / subscriber by some prize (money, cell phone, discount to services and so on) in return of 24 month obligation to use one’s services. So, every 24 months subscriber gains ability to choose another telecommunication network. In order to maintain stable amount of subscribers’ service provider must offer something in return. In order to do that correctly, without financial loses, one must know exact usage statistics of each subscriber. This paper covers couple tips to arrange data in data warehouses (data marts) in order to achieve greater data cube pre processing speed. One of these methods covers partial data aggregation to highest degree, still sufficient to answer specific queries. Another method shows the ability to synthesize data cube dimensions in order to lower data volumes, that data cube pre calculation could take less time.

With more and more integration of analog and RF circuits in scaled CMOS technologies, process variation is playing a critical role which makes it difficult to achieve all the performance specifications across all the process corners. Moreover, at scaled technology nodes, due to lower voltage and current handling capabilities of the devices, they suffer from reliability issues that reduce the overall lifetime of the system. Finally, traditional static style of designing analog and RF circuits does not result in optimal performance of the system. A new design paradigm is emerging toward digitally assisted analog and RF circuits and systems aiming to leverage digital correction and calibration techniques to detect and compensate for the manufacturing imperfections and improve the analog and RF performance offering a high level of integration. The objective of the proposed research is to design digital friendly and performance tunable adaptive analog/RF circuits and systems with digital enhancement techniques for higher performance, better process variation tolerance, and more reliable operation and developing strategy for testing the proposed adaptive systems. An adaptation framework is developed for process variation tolerant RF systems which has two parts – optimized test stimulus driven diagnosis of individual modules and power optimal system level tuning. Another direct tuning approach is developed and demonstrated on a carbon nanotube based analog circuit. An adaptive switched mode power amplifier is designed which is more digital-intensive in nature and has higher efficiency, improved reliability and better process resiliency. Finally, a testing strategy for adaptive RF systems is shown which reduces test time and test cost compared to traditional testing.

Context. Today’s parallel systems are widely used in different computational tasks. Developing parallel programs to make maximum use of the computing power of parallel systems is tricky and efficient tuning of parallel programs is often very hard. Objectives. In this study we present a performance prediction and visualization tool named VPPB for a Linux environment, which had already been introduced by Broberg et.al, [1] for a Solaris2.x environment. VPPB shows the predicted behavior of a multithreaded program using any number of processors and the behavior is shown on two different graphs. The prediction is based on a monitored uni-processor execution. Methods. An experimental evaluation was carried out to validate the prediction reliability of the developed tool. Results. Validation of prediction is conducted, using an Intel multiprocessor with 8 processors and PARSEC 2.0 benchmark suite application programs. The validation shows that the speed-up predictions are +/-7% of a real execution. Conclusions. The experimentation of the VPPB tool showed that the prediction of VPPB is reliable and the incurred overhead into the application programs is low. / contact: +46(0)736368336

When considering the addition of a mobile presentation channel to an existing web-based application, project managers should know how the mobile channel|s characteristics will impact the user experience and the cost of using the application, even before development begins. The PETTICOAT (Performance Tuning and cost discovery of mobile web-based Applications) approach presented here provides decision-makers with indicators on the economical feasibility of mobile channel development. In a nutshell, it involves analysing interaction patterns on the existing stationary channel, identifying key business processes among them, measuring the time and data volume incurred in their execution, and then simulating how the same interaction patterns would run when subjected to the frame conditions of a mobile channel. As a result of the simulation, we then gain time and volume projections for those interaction patterns that allow us to estimate the costs incurred by executing certain business processes on different mobile channels.

Today's trend in microprocessor design is increasing the number of execution cores within one single chip. Increasing the processor's clock speed reached its limit with growing power consumption. This trend brings new opportunities to software developers, as they can take advantage of real multithreading in their applications. But a lot of new problems to solve appear with threading compared to sequential programming. With proper design, threading can enhance performance by making better use of hardware resources. However, the improper use of threading can lead to performance degradation, unpredictible behavior, or error conditions that are difficult to solve. For this reason Intel developed a suite of tools, that can help software developers to analyze performance and detect coding errors in thread interactions. This thesis focuses on the examination of ways that this tools can be used in multithreaded application development.

Canvas is a LMS used by many colleges, universities, and K-12 schools. The primary purpose of Canvas is for the instructors to organize courses, create assignments, and enter students’ grades. Canvas LMS is a web application. The response time performance of the Canvas is essential for its users. Both instructors and students claim that they experience slow response times from this system. Given the vast numbers of users, the effect of delayed responses is multiplied by a very large number; hence, improving the performance of such a system has a large economic and social impact. Moreover, many other applications utilize a set of services that communicate to realize the overall web service; hence, the results could also positively impact other services. This thesis focuses on the Canvas when running in a Docker container environment and proposes solutions to tune the system to optimize the performance of Canvas. This thesis uses experiments on the following aspects of the Canvas: Canvas RESTful API, Canvas GUI, the underlying system of the Docker containers, Canvas webserver, VM configurations, etc. This thesis provides tuning and optimization suggestions that could benefit Canvas developers and Canvas administrators. / Canvas är en Lärplattform (eng. Learning Management System(LMS)) används av många högskolor, universitet och K-12-skolor. Canvas huvudsakliga syfte är att instruktörerna ska organisera kurser, skapa uppdrag och ange elevernas betyg. Canvas LMS är en webbapplikation. Svarstider prestanda för Canvas är avgörande för användarna. Både instruktörer och studenter hävdar att de upplever långsamma svarstider från detta system. Med tanke på det stora antalet användare multipliceras effekten av fördröjda svar med ett mycket stort antal; följaktligen har förbättring av prestanda för ett sådant system en stor ekonomisk och social inverkan. Dessutom använder många andra applikationer en uppsättning tjänster som kommunicerar för att förverkliga den övergripande webbtjänsten; därför kan resultaten också påverka andra tjänster positivt. Denna avhandling fokuserar på Canvas när den körs i en Docker-behållare miljö och föreslår lösningar för att justera systemet för att optimera prestandan för Canvas. Denna avhandling använder experiment på följande aspekter av Canvas: Canvas RESTful applikationsprogrammeringsgränssnitt (API), Canvas grafiskt användargränssnitt (GUI), det underliggande systemet för Docker-behållare, Canvas-webbserver, virtuella maskin (VM)-konfigurationer, osv. Denna avhandling ger prestanda justering och optimering förslagsom kan gynna Canvas-utvecklare och Canvas-administratörer.