This thesis deals with the design and implementation of an optimizer and a transformer of relational queries. Firstly, the thesis describes the theory of the relational query compilers. Secondly, we present the data structures and algorithms used in the implemented tool. Finally, the important implementation details of the developed tool are discussed. Part of the thesis is the selection of used relational algebra operators and design of an appropriate input. Input of the implemented software is a query written in a XML file in the form of relational algebra. Query is optimized and transformed into physical plan which will be executed in the parallelization framework Bobox. Developed compiler outputs physical plan written in the Bobolang language, which serves as an input for the Bobox. Powered by TCPDF (www.tcpdf.org)

Diploma thesis focuses on the selection of suitable HW solution for parallell processing of multiple audio sources. Compares several different platforms based on architectures ARM, x86 and Epiphany, compares their performance in serial and parallel data processing, their energy consumption and price.

Big Data on cities and regions bring new opportunities and challenges to data analysts and city planners. On the one side, they hold great promise to combine increasingly detailed data for each citizen with critical infrastructures to plan, govern and manage cities and regions, improve their sustainability, optimize processes and maximize the provision of public and private services. On the other side, the massive sample size and high-dimensionality of Big Data and their geo-temporal character introduce unique computational and statistical challenges. This chapter provides overviews on the salient characteristics of Big Data and how these features impact on paradigm change of data management and analysis, and also on the computing environment. / Series: Working Papers in Regional Science

This work deals with the design and implementation of a system for parallel execution of tasks in the Knowledge Technology Research Group. The goal is to create a web application that allows to control their processing and monitor runs of these tasks including the use of system resources. The work first analyzes the current method of parallel data processing and the shortcomings of this solution. Then the work describes the existing tools including the problems that their test deployment revealed. Based on this knowledge, the requirements for a new application are defined and the design of the entire system is created. After that the selected parts of implementation and the way of the whole system testing is described together with the comparison of the efficiency with the original system.

Der exponentielle Anstieg elektronisch gespeicherter Daten erfordert leistungsfähige Systeme zur Verarbeitung und Analyse großer Datenmengen. Parallel relationale Datenbanksysteme (PRDBMS) waren lange Zeit der Standard für analytische Abfragen. Neuere Systeme, wie Apache Flink, Tez und Spark, nutzen erweiterte Ansätze zur Analyse und trennen logische Spezifikationen von physischen Ausführungen. Ein weit verbreitetes Optimierungsverfahren in der analytischen Verarbeitung ist die partielle Aggregation, bei der Aggregation in zwei Stufen erfolgt: Zunächst werden partielle Aggregatgruppen erstellt, die dann zusammengeführt werden, um das Endergebnis zu berechnen. Dieses Verfahren ermöglicht eine parallele Verarbeitung und reduziert die Größe der Zwischenergebnisse. Bisherige Ansätze konzentrieren sich auf ordnungsunabhängige Gruppierungsfunktionen, bei denen Elemente ohne Berücksichtigung der Reihenfolge gruppiert werden können. In der Praxis gibt es jedoch auch ordnungsabhängige Gruppierungsfunktionen, die von der Reihenfolge der Eingaben abhängen und komplexer in der parallelen Ausführung sind. Derzeit existieren nur begrenzte Ansätze für eine effiziente Parallelisierung solcher Funktionen. Diese Dissertation präsentiert einen neuen Ansatz zur Parallelisierung von Aggregationsanfragen für drei ordnungsabhängige Gruppierungsfunktionen: Sessionization, Regular Expression Matching (REM) und Complex Event Recognition (CER). Unsere Methode nutzt zerlegbare Aggregationsfunktionen, um eine effiziente parallele Ausführung in modernen Shared-Nothing-Compute-Umgebungen zu ermöglichen. Die stufenweise Ausführung dieser Funktionen eröffnet neue Optimierungsmöglichkeiten. Unser Ansatz erlaubt es Optimierungsalgorithmen, zwischen sequentiellen und stufenweisen Verfahren zu wählen. Zusätzlich schlägt die Arbeit ein Schema vor, wie weitere Gruppierungsfunktionen zerlegt und in die partielle Aggregation integriert werden können. / Advances in information technologies and decreasing cost for storage and compute capacities lead to exponential growth of data being available electronically worldwide. Systems capable of processing these large amounts of data with the goal of analyzing and extracting information are essential for both: research and businesses. Analytical data processing systems employ various optimizations to execute queries efficiently. Partial Aggregation (PA) using GroupBy and decomposable aggregation functions is a common optimization approach in analytical query processing. Analytical systems execute PA in two stages: During the first stage, they create partial groups to compute partial aggregates. During the second stage, the partial aggregates are grouped and aggregated again to produce the final result. The main benefits of PA are an increased potential of parallel execution during the first stage and a reduction of intermediate result sizes by aggregating over the partial groups. So far, existing approaches to PA only use an order-agnostic grouping function on sets to create groups. There are grouping functions that depend on ordered input and information on previously processed input items to associate a given input item to its group. Staged execution of order-dependent grouping functions is more difficult than for order-agnostic grouping functions. Systems must compute correct partial states during the first stage and combine them during the final stage. Approaches for efficient parallel execution only exist in a limited way despite the high practical relevance. In this thesis, we present a novel approach for parallelizing aggregation for three order-dependent grouping functions: Sessionization, Regular Expression Matching (REM), and Complex Event Recognition (CER). Our approach of computing the three grouping functions in stages combined with decomposable aggregation functions allows for efficient parallel execution in state-of-the-art shared-nothing compute environments.