05 Fakultät Informatik, Elektrotechnik und Informationstechnik
Permanent URI for this collectionhttps://elib.uni-stuttgart.de/handle/11682/6
Browse
Search Results
Item Open Access Cost optimization for data placement strategies in an analytical cloud service(2016) Saleem, Muhammad UsmanAnalyzing a large amount of business-relevant data in near-realtime in order to assist decision making became a crucial requirement for many businesses in the last years. Therefore, all major database system vendors offer solutions that assist customers in this requirement with systems that are specially tuned for accelerating analytical workloads. Before the decision is made to buy such a huge and expensive solution, customers are interested in getting a detailed workload analysis in order to estimate potential benefits. Therefore, a more agile solution is desirable having lower barriers to entry that allows customers to assess analytical solutions for their workloads and lets data scientists experiment with available data on test systems before rolling out valuable analytical reports on a production system. In such a scenario where separate systems are deployed for handling transactional workloads of daily customers business and conducting business analytics on either a cloud service or a dedicated accelerator appliance, data management and placement strategies are of high importance. Multiple approaches exist for keeping the data set in-sync and guaranteeing data coherence with unique characteristics regarding important metrics that impact query performance, such as the latency when data will be propagated, achievable throughputs for larger data volumes, or the amount of required CPU to detect and deploy data changes. So the important heuristics are analyzed and evolved in order to develop a general model for data placement and maintenance strategies. Based on this theoretical model, a prototype is also implemented that predicts these metrics.Item Open Access Serialization of foreign types with SKilL(2016) Weißer, ConstantinSKilL provides a language independent means to specify serialisable data types. Bindings for these types can be generated automatically for multiple supported languages based on this specification. However, using these bindings is only hassle-free for newly written code, because existing types must be replaced by the controlled generated bindings. This impedes the usefulness of SKilL in old projects. We investigate a new approach. SKilL is to be extended in order to support the reuse of existing Java classes, so-called foreign types. The tool shall be able to analyse classes, associate them with specified SKilL types, verify the type correctness of this association and generate the required code in order to serialise objects of these types. This thesis points out the occurring challenges and discusses potential solutions. We experiment with several alternatives and provide insight into their pros and cons, as well as justification for our design. Functional and performance tests assess our implementation and shortcomings are addressed in detail.Item Open Access Multi-view stereo with inverse depth parameterization(2016) Goroll, OliverEstimating accurate depth maps by solving the stereo problem is an important step in reconstructing real world surfaces. Variational methods that minimize a global energy functional are considered as especially precise throughout the literature. By using the depth as parameterization directly, the approach is easily extended to multiple views, allowing to considerably enhance the quality of the resulting depth maps. Further improvement can be achieved by adapting the depth parameterization to the regularizer used. In this thesis second-order regularization is used with an \textsc{inverse} depth parameterization and compared against the direct depth parameterization to examine its benefits. Several extensions that improve upon the naiive multi-view approach are suggested. The proposed method and its extensions are evaluated by experiments, using artificial as well as real world test scenes.Item Open Access Exploring cloud-based sharing of community recipes for smart environments(2016) Olp, DominikOur environments have become continuously smarter. With the advancement of technology, devices in our daily lives become controllable via smartphones or configurable to one's demand. Letting users define the behaviour of their devices yields a new type of interaction and creates a smart environment. Once a helpful behaviour was created, sharing such work with others, leads to the benefit of even more users. Therefore, a sharing platform was created that lets users simply create and exchange new solutions for their smart environment. To this date, the sharing of a smart environment's behaviour is a novel approach. The built sharing platform was used to research the implications of sharing behaviour between smart environments. In addition, the study was carried out to gather experience and feedback from real users. Creating such a platform required developing a data format to easily provide and distribute the necessary information to apply behaviour in another environment.Item Open Access Kooperative Vorhersage der minimalen Anwendungsausführungszeit(2016) Kuhn, JulianCode Offloading Frameworks verbessern durch Auslagern von Programmteilen - auch Offloadingkandidaten genannt - auf Server die Leistung oder den Energieverbrauch von Geräten mit limitierten Ressourcen. Offloadingkandidaten werden dann ausgelagert, wenn mit Inbetrachtnahme der Übertragung des Kandidaten eine Einsparung im Vergleich zur rein lokalen Ausführung vorliegt. Die Entscheidung, ob Offloading stattfindet, hängt stark von der Ausführungszeit des Kandidaten ab. Im Fall von Methoden kann die Ausführungszeit je nach aktueller Parameterkonfiguration stark variieren. Da es in vielen Fällen unpraktikabel ist, für jede Parameterkombination Aufzeichnungen durchzuführen, ist die Verwendung von einfachen, historienbasierten Modellen zur Bestimmung der Ausführungszeit ungeeignet. Eine möglichst genaue Angabe der Ausführungszeit wird aber benötigt, um die Offloadingentscheidung korrekt zu treffen. Ziel der Arbeit war, die Vorhersage von Ausführungszeiten mit Hilfe von Machine Learning Modellen anhand verschiedener Testanwendungen- und Szenarien im Kontext des Code Off-loadings zu untersuchen. Außerdem wurde ein kooperativer Systementwurf vorgestellt und implementiert, der zur Verwaltung von Datensätzen, Vorhersagemodellen und deren Erstellung, sowie zur Vorhersage von Ausführungszeiten verwendet werden kann. Der Entwurf erweitert dabei bestehende Offloadingframeworks. Es konnte festgestellt werden, dass sich Machine Learning Algorithmen zur Vorhersage und insbesondere zum Verbessern der Offloadingentscheidung eignen.Item Open Access Konzept und Implementierung einer Komponente zur Kommunikation TOSCA-basierter Anwendungen(2016) Zimmermann, MichaelCloud Computing ist ein häufig eingesetztes Paradigma zur Realisierung moderner IT-Systeme. Es ermöglicht die flexible Nutzung von professioneller IT-Infrastruktur ohne die Verwaltung dieser zwingend selbst übernehmen zu müssen. Jedoch besteht das Problem des Vendor-Lock-Ins, also die Abhängigkeit von einem bestimmten Provider. Die Topology and Orchestration Specification for Cloud Applications (TOSCA) bietet jedoch eine Möglichkeit zur Modellierung von portablen und interoperablen Cloud-Anwendungen. Weiterhin wurde an der Universität Stuttgart eine Open Source Implementierung einer TOSCA-Laufzeitumgebung namens OpenTOSCA für diesen Standard entwickelt. Allerdings bietet diese bisher keine Möglichkeit zur direkten Kommunikation verschiedener, durch OpenTOSCA verwalteter Anwendungen an. Im Rahmen dieser Arbeit wurde aus diesem Grund eine Komponente für Open-TOSCA konzipiert und anschließend implementiert, um eine Kommunikation zwischen verschiedenen durch OpenTOSCA deployten Anwendungen auf eine einheitliche Art und Weise zu ermöglichen. Dafür werden in dieser Arbeit, neben benötigten Erweiterungen des TOSCA-Standards, Anforderungen an die zu entwickelnde Komponente aufgestellt und daraus folgend Entwurfsentscheidungen getroffen. Weiterhin wird ein Konzept zur Problemlösung entworfen und darauf aufbauend eine Architektur erarbeitet, welche die Kommunikationskomponente in OpenTOSCA integriert. Eine prototypische Implementierung der konzipierten Komponente sowie deren Integration in OpenTOSCA ist ebenfalls Teil dieser Arbeit.Item Open Access Automatic splitting in data-parallel complex event processing systems(2016) Sanwald, TimParallel Complex Event Processing (CEP) systems handle today’s heavy loaded event streams from smart homes, network traffic systems or stock trading systems by distributing the incoming event stream to several pattern detection systems. The correct splitting is currently done by CEP experts which ensure the consistent splitting without generating false-positive or false-negative complex events in comparison with centralized CEP systems. In this work an approach is developed which automatically generates a splitting model from the pattern definition which ensures the consistent distribution without generating false positives or false negatives. This approach enables a parallel CEP system to be configured and used the same way as a centralized CEP system. Further, a method which combines window based splitting and key based splitting is presented to reduce the network load and the CPU load on pattern detection operators. The functionality of the automatic splitting and the optimization is validated with common CEP scenarios based on generated and real world data to ensure a wide applicability of the approach.Item Open Access Introducing performance awareness in an integrated specification environment(2016) Keller, FabianWith an increase in software complexity and modularization to create large software systems and software product lines it is increasingly difficult to ensure all requirements are met by the built system. Performance requirements are an important concern to software systems and research has developed approaches being capable of predicting software performance from annotated software architecture descriptions, such as the Palladio tool suite. However, the tooling when moving between specification, implementation and verification phase has a gap as the tools are commonly not linked, leading to inconsistencies and ambiguities in the produced artifacts. This thesis introduces performance awareness into the Integrated Specification Environment for the Specification of Technical Software Systems (IETS3), which is a specification environment aiming to close the tooling gap between the different lifecycle phases. Performance awareness is introduced by integrating existing approaches for software performance prediction from the Palladio tool suite and extending them to cope with variability-aware system models for software product lines. The thesis includes an experimental evaluation showing that the developed approach is able to provide performance predictions to users of the specification environment within 2000 ms for systems of up to 20 components and within 8000 ms for systems of up to 30 components.Item Open Access Recognition of resource patterns in human-centric processes : a case study(2016) Song, MoziBusiness experts need to improve business processes by increasing process efficiency and reducing process costs. To achieve this, a common method is to capture a series of repeatedly conducted process activities and their structure, i.e. the business logic of the process, and then enact process execution based on it. However, there exist informal processes, which are human-centric processes that are highly dynamic. Since this approach assumes the existence of predictable business logic of the process, it does not apply for management of informal processes. The Informal Process Essentials (IPE) model is a modeling approach for informal processes. This model depicts informal processes by documenting resources used in these process. Through this approach, we are able to retain best practice and knowledge accumulated in the processes. Based on this approach, there is also the InProXec method to enable the application of the IPE approach in organizations. In this thesis work, we want to validate the concepts introduced in the InProXec method using a case study on the jclouds project. To achieve this aim, we introduce the concept of a generic mapping mechanism and an evolving correlation coefficient function. Based on these concepts, we present the Informal Process Discoverer (IPD) service. The IPD service is an implementation of the discovery of IPE models. The test results of the IPD service have shown that the service is successful in discovering the IPE model and giving process recommendations. For example, using an informal process model with includes 7 human resources and 2 GitHub repositories as input, we are able to discover 74 other resources that participate in the process including 65 human resources and 9 Git repositories.Item Open Access Testabdeckungsmaße für den Integrationstest von E/E-Systemen im Fahrzeug(2016) Walter, JaasielDie vorliegende Masterarbeit beschäftigt sich mit Testabdeckungsmaßen für den Integrationstest von E/E-Systemen. E/E-Systeme bestehen aus E/E-Komponenten (z.B. Steuergeräte, Sensoren und Aktoren) und sind fester Bestandteil bzw. Zukunftstrend in der Automobilindustrie. Beispielsweise befinden sich in einer modernen Mercedes-Benz S-Klasse je nach Ausstattung ca. 100 Steuergeräte. Komponenten der E/E-Systeme werden i.d.R. von Zulieferern isoliert entwickelt, getestet und als Black-Box an den Fahrzeughersteller (OEM) übergeben. Letzterer integriert die Komponenten zu Sub-/Systemen und testet deren Zusammenwirken. Oft wird zur Bewertung der Testgüte lediglich die Anforderungsabdeckung gemessen. Ziel dieser Arbeit war die Entwicklung von ergänzenden Testabdeckungsmaßen für den black-box-orientierten Integrationstest. Weiterhin sollte ein Konzept zur Nutzung der entwickelten Maße in einer erweiterten Teststrategie entworfen werden. Da es sich bei den E/E-Komponenten um Black-Boxes handelt, sind auch die beim Softwareintegrationstest einsetzbaren Testabdeckungsmaße wie Anweisungs-, Zweig- und Pfadüberdeckung nicht direkt einsetzbar. In der Literatur konnten keine veröffentlichten Testabdeckungsmaße zu diesem Thema ermittelt werden. Daher wurden aus dem Softwaretest bekannte Testabdeckungsmaße auf den Integrationstest von E/E-Systemen übertragen. Diese basieren auf Architektur- und Schnittstellenbeschreibungen, Zustandsautomaten und Kommunikationsmatrizen. Zur Evaluierung wurde der Einsatz der Testabdeckungsmaße anhand von Fallbeispielen demonstriert und durch Expertenbefragungen bewertet.