Each software project requires a specialized management to handle software development activities throughout the project life cycle successfully and efficiently. Software governance structures provide easy and efficient ways to handle software development activities. Software configuration management is an important software development activity, and while selecting the right strategy in configuration management, its conformity to the software governance should be considered as well. Software configuration management patterns are aligned with the software governance structures to increase the success in development and management of the projects. Companies running large and inter-dependent projects, should adapt their governance structures to the changing characteristics and dependencies of projects. In this thesis, we propose a method to dynamically manage software configuration management, as a result of the changing specifications in the software governance representation.

This thesis addresses the problem of setting up a surface mount placement machine for production. The objective is to minimize the number of machine changeovers made during a production run consisting of a number of circuit cards. The solution to the problem involves two separate decisions. The first decision considers determining how to combine feeders together in "bays" or groups of feeders, and how to assign the bays to the circuit cards. The second decision considers the circuit card production sequence. A mathematical programming formulation is given, however, its solution is very difficult for problems of a realistic size. Several heuristic approaches are suggested and used to solve actual and test problems. The heuristic for bay design uses clustering techniques used in Group Technology while the sequencing problem is solved using heuristics based on solution techniques for the Traveling Salesman problem.

ITC/USA 2007 Conference Proceedings / The Forty-Third Annual International Telemetering Conference and Technical Exhibition / October 22-25, 2007 / Riviera Hotel & Convention Center, Las Vegas, Nevada / The Instrumentation Hardware Abstraction Language (IHAL) has been developed to be a neutral language that is focused on the description and control of instrumentation systems and networks. This paper describes the various instrumentation configuration management tools we have designed that make use of IHAL’s neutral specification of instrumentation networks. We discuss the features currently present in prototypes as well as future enhancements.

The ever-increasing complexity of software systems makes them hard to comprehend, predict and tune due to emergent properties and non-deterministic behaviour. Complexity arises from the size of software systems and the wide variety of possible operating environments: the increasing choice of platforms and communication policies leads to ever more complex performance characteristics. In addition, software systems exhibit different behaviour under different workloads. Many software systems are designed to be configurable so that policies (e.g. communication, concurrency and recovery strategies) can be chosen to meet the needs of various stakeholders. For complex software systems it can be difficult to accurately predict the effects of a change and to know which configuration is most appropriate. This thesis demonstrates that it is useful to run automated experiments that measure a selection of system configurations. Experiments can find configurations that meet the stakeholders' needs, find interesting behavioural characteristics, and help produce predictive models of the system's behaviour. The design and use of ACT (Automated Configuration Tool) for running such experiments is described, in combination a number of search strategies for deciding on the configurations to measure. Design Of Experiments (DOE) is discussed, with emphasis on Taguchi Methods. These statistical methods have been used extensively in manufacturing, but have not previously been used for configuring software systems. The novel contribution here is an industrial case study, applying the combination of ACT and Taguchi Methods to DC-Directory, a product from Data Connection Ltd (DCL). The case study investigated the applicability of Taguchi Methods for configuring complex software systems. Taguchi Methods were found to be useful for modelling and configuring DC-Directory, making them a valuable addition to the techniques available to system administrators and developers.

This research is based on real issues that have been recognised within the global organisation, Rolls-Royce. The first aim concerned an issue that many companies face, the difficulty that employees face in locating the knowledge and information they require, especially in larger organisations. The developed solution of an Information Map proved to be a success in providing people within the Submarines business with the location of' Configuration Management information. The concept of the Information Map is one that can be adopted by any business as the stages in the tools development have been well documented within Chapters Four, Five and Six. Analysis of the success of the Information Map led to the derivation of 10 lessons learned. These were then verified in a second case study of an intranet development. The second aim of the research was to create a Knowledge Management framework that could be adapted by companies looking to invest in Knowledge Management and provide them with a guide to use. This framework was built from the lessons learned from the Information Map and from other best practice derived from available literature and within Rolls-Royce. The work conducted within the Support business tries to fill in gaps in current research by offering companies a new approach to Knowledge Management, which was based upon the way that industries work today. The creation of the Knowledge Management framework simplifies the work conducted and offers practitioners an easy, high-level approach to the adoption of Knowledge Management by grouping the process into ten steps. This is presented in a fashion that is easy to follow and ultimately offers a guide to make the best use of the resources and budget available to Knowledge Management practitioners. Overall the research addresses the 'real' issues faced by Knowledge Management practitioners. The main contributions to the Knowledge Management domain are the Information Map, action research approach, implementation of Knowledge Management tools for the users needs and a Framework as a guide for industry.

Open Source Design (OSD) é uma estratégia de desenvolvimento adotada por comunidades que desenvolvem produtos manufaturados colaborativamente, em um regime copyleft e rede peer to peer virtual. O resultado é um conjunto de informações capazes de permitir a fabricação remota em que o usuário pode construir o seu produto, ao invés de um sistema produtivo específico previamente definido, como no desenvolvimento tradicional. A tangibilidade e a distância física entre desenvolvedores impõem barreiras à colaboração, mas avanços como a tecnologia de impressão 3D fez surgir plataformas virtuais para o compartilhamento de arquivos que tornam esta realidade possível. O número deste tipo de plataformas é significativo e há problemas na delimitação teórica do fenômeno e questões em aberto. No desenvolvimento OSD faz-se necessário o uso de objetos de fronteira (Boundary Objects – BOs). Qual o papel destes objetos? Eles estão presentes nas plataformas citadas? As plataformas existentes possuem recursos para a gestão destes objetos? Quais as características que estas plataformas devem apresentar para viabilizar o desenvolvimento OSD? Este trabalho reúne um conjunto de investigações para elucidar as questões acima. Por meio de um estudo de caso inicial e revisões sistemáticas, os conceitos foram analisados e propostos teoricamente. Por meio de um levantamento de 686 projetos de seis comunidades OSD diferentes hospedadas na plataforma Thingiverse, auxiliado por uma análise de conteúdo sobre a relação entre plataforma e colaboração, em seis projetos de diferentes plataformas, foi possível discutir o papel das plataformas utilizadas por comunidades OSD. Discute-se a relação com o desempenho do projeto e com a colaboração alcançada, identificando-se problemas potenciais e soluções latentes. Os resultados indicam que as plataformas estudadas, entre as mais difundidas, apresentam limitações. Finalmente, por meio da análise conjunta dos resultados, à luz das práticas de gestão de configuração existentes no desenvolvimento tradicional, foi possível identificar uma lista de propriedades latentes, que poderiam ser incorporadas nas plataformas, como: estrutura do produto, versionamento, ferramentas de armazenamento e rastreabilidade, auditoria, controle da configuração e governança da interface. As propriedades identificadas são uma contribuição importante para a inserção em ferramentas de design colaborativo. / Open Source Design (OSD) is a strategy adopted by communities that develop collaboratively manufactured products in a copyleft conditions and on a peer to peer network. The result is a set of information that allow the remote manufacturing where the users can build their own products, despite of a specific productive system previously defined as the traditional development. The tangibility and the geographic distance among developers impose barriers on the collaboration, on the other hand technology breakthroughs in the 3D printer technology made arise virtual platforms to share files that made this reality achievable. The number of these types of platforms is increasing and there are difficulties to delimitating of the phenomenon and there are gaps that need investigation. Is required the use of BOs in OSD development? What is the role of these BOs? Are they used in the mentioned platforms? Do the current platforms have features to the management of BOs? Which features must these platforms have to make feasible the OSD development? This thesis gathers an amount of investigations to elucidate the questions above. Through an initial study of case and systematic revisions, the concepts were analyzed and theoretically proposed. There was collected information of 686 projects from 6 different OSD communities hosted on the Thingiverse platform; in complement a content analysis about the relation between the platform and collaboration in six projects of three different platforms, to discuss the role of the platforms used for OSD communities. It is discussed the relation between the performance project with the collaboration, identifying the potential problems and the latent solutions. The result indicates that the investigated platforms, even the most widespread, present limitations. Finally, using a combined analysis of the results, according to the practices of management configuration existent in the traditional development, it was possible to identify a list of features that could be incorporated in the platforms, such: product structure, versioning, storage tools and traceability, audits, control configuration, governance. The identified features are important contributions to the insertion in tools of collaborative design.

Open Source Design (OSD) é uma estratégia de desenvolvimento adotada por comunidades que desenvolvem produtos manufaturados colaborativamente, em um regime copyleft e rede peer to peer virtual. O resultado é um conjunto de informações capazes de permitir a fabricação remota em que o usuário pode construir o seu produto, ao invés de um sistema produtivo específico previamente definido, como no desenvolvimento tradicional. A tangibilidade e a distância física entre desenvolvedores impõem barreiras à colaboração, mas avanços como a tecnologia de impressão 3D fez surgir plataformas virtuais para o compartilhamento de arquivos que tornam esta realidade possível. O número deste tipo de plataformas é significativo e há problemas na delimitação teórica do fenômeno e questões em aberto. No desenvolvimento OSD faz-se necessário o uso de objetos de fronteira (Boundary Objects – BOs). Qual o papel destes objetos? Eles estão presentes nas plataformas citadas? As plataformas existentes possuem recursos para a gestão destes objetos? Quais as características que estas plataformas devem apresentar para viabilizar o desenvolvimento OSD? Este trabalho reúne um conjunto de investigações para elucidar as questões acima. Por meio de um estudo de caso inicial e revisões sistemáticas, os conceitos foram analisados e propostos teoricamente. Por meio de um levantamento de 686 projetos de seis comunidades OSD diferentes hospedadas na plataforma Thingiverse, auxiliado por uma análise de conteúdo sobre a relação entre plataforma e colaboração, em seis projetos de diferentes plataformas, foi possível discutir o papel das plataformas utilizadas por comunidades OSD. Discute-se a relação com o desempenho do projeto e com a colaboração alcançada, identificando-se problemas potenciais e soluções latentes. Os resultados indicam que as plataformas estudadas, entre as mais difundidas, apresentam limitações. Finalmente, por meio da análise conjunta dos resultados, à luz das práticas de gestão de configuração existentes no desenvolvimento tradicional, foi possível identificar uma lista de propriedades latentes, que poderiam ser incorporadas nas plataformas, como: estrutura do produto, versionamento, ferramentas de armazenamento e rastreabilidade, auditoria, controle da configuração e governança da interface. As propriedades identificadas são uma contribuição importante para a inserção em ferramentas de design colaborativo. / Open Source Design (OSD) is a strategy adopted by communities that develop collaboratively manufactured products in a copyleft conditions and on a peer to peer network. The result is a set of information that allow the remote manufacturing where the users can build their own products, despite of a specific productive system previously defined as the traditional development. The tangibility and the geographic distance among developers impose barriers on the collaboration, on the other hand technology breakthroughs in the 3D printer technology made arise virtual platforms to share files that made this reality achievable. The number of these types of platforms is increasing and there are difficulties to delimitating of the phenomenon and there are gaps that need investigation. Is required the use of BOs in OSD development? What is the role of these BOs? Are they used in the mentioned platforms? Do the current platforms have features to the management of BOs? Which features must these platforms have to make feasible the OSD development? This thesis gathers an amount of investigations to elucidate the questions above. Through an initial study of case and systematic revisions, the concepts were analyzed and theoretically proposed. There was collected information of 686 projects from 6 different OSD communities hosted on the Thingiverse platform; in complement a content analysis about the relation between the platform and collaboration in six projects of three different platforms, to discuss the role of the platforms used for OSD communities. It is discussed the relation between the performance project with the collaboration, identifying the potential problems and the latent solutions. The result indicates that the investigated platforms, even the most widespread, present limitations. Finally, using a combined analysis of the results, according to the practices of management configuration existent in the traditional development, it was possible to identify a list of features that could be incorporated in the platforms, such: product structure, versioning, storage tools and traceability, audits, control configuration, governance. The identified features are important contributions to the insertion in tools of collaborative design.

Reusability of assets within a family of products is the major goal of Software Product Line Engineering (SPLE), therefore managing variability is an important task in SPLs. Configuration management in the context of software product line engineering is more complicated than that in single systems engineering due to &rdquo / variability in space&rdquo / in addition to &rdquo / variability in time&rdquo / of core assets. In this study, a method for documenting variability in executable configuration items, namely components, is proposed by associating them with the Orthogonal Variability Model (OVM) which introduces variability as a separate model. The main aim is to trace variability in dierent configurations by explicitly documenting variability information for components. The links between OVM elements and components facilitate tool support for product derivation as the components matching the selected variations can be gathered by following the links. The proposed scheme is demonstrated on a case study about a radar GUI variability model.

Uma rede local necessita ter a sua configuração gerenciada para manter a sua utilidade dentro da organização a qual pertence. Esta gerencia a complicada pela heterogeneidade da rede e pela sua importância dentro da organização. A tentativa de solucionar este problema passa pela utilização de um padrão de gerencia de rede e pelo desenvolvimento de ferramentas que permitam a gerencia eficiente da rede pelo administrador. Este trabalho apresenta uma proposta de gerencia de configuração de redes locais. São apresentados os objetos gerenciados necessários a gerencia de configuração. Parte destes objetos são previamente definidos pela IAB (Internet Activities Board) em normas; e parte 6 definida neste trabalho. Os objetos são definidos de acordo com o formato proposto pela IAB. Entretanto, para que os objetos sejam definidos, é necessário, antes, que as funções de gerencia de configuração_ sejam delineadas. Os objetos são definidos a partir das funções que se deseja executar. Portanto, antes da apresentação dos objetos, são definidas funções de gerencia de configuração. Também é apresentada a implementação de um protótipo de gerente que realiza varias das funções definidas no trabalho, utilizando o padrão SNMP (Simple Network Management Protocol). Este protótipo utiliza o pacote de programas CMU (Carnegie Mellon University) SNMP para a implementação das funções. A partir do protótipo pode-se perceber como se realizam as funções de gerencia e como deve ser a interface de , um gerente de configuração, a fim de que este seja •realmente útil nas tarefas do administrador da rede. Por ultimo, e feita uma modelagem dos objetos gerenciados de acordo com a norma OSI (Open Systems Interconnection). Várias são as modificações decorrentes delta modelagem diferenciada. Silo apresentadas as diferenças que ocorrem na gerencia de configuração quando se utiliza o modelo de dados OSI, em vez do modelo de dados Internet. / In order to keep a LAN useful to its organization, its configuration has to be managed. Heterogenity and importance of networks makes management more difficult. An attempt to solve this problem has to consider the management standards and the development of tools which allow an efficient network management. This work presents a proposal of local area network configuration management. Managed objects necessary to manage configuration are presented. Part of these objects are, previously, define in IAB (Internet Activities Board) standards, and part of them are defined here. Objects are defined according to the format proposed by IAB. However, before the definition of objects, it is necessary to define configuration management functions. Objects are defined according to the functions one wants to implement. Thus, before the presentation of objects, configuration management functions are defined. It is also presented a manager prototype which executes many of the functions defined in this work, using SNMP (Simple Network Management Protocol). The prototype uses CMU (Carnegie Mellon University) SNMP software package to implement functions. It is possible to notice, from the prototype, how functions are realized and which characteristics the interface of a configuration manager must have to be useful in the tasks performed by human manager. Finally, modelling of managed objects is done according to OSI (Open Systems Interconnection) standards. Many changes occur because of this diferent modelling. The changes in configuration management which occurs when one uses OSI data model, instead of Internet data model, are presented.

Nesta dissertação é apresentada uma avaliação da viabilidade de obtenção de operações atômicas para a configuração de múltiplos dispositivos em uma rede de computadores. Em especial, é tratada a configuração de dispositivos a partir do gerenciamento baseado em políticas. Nestes casos, a necessidade de operações atômicas vem do fato de que uma aplicação de política que falha em um determinado dispositivo pode levar a um estado inconsistente em uma rede da qual se deseja um comportamento uniforme e global. Assim, este trabalho analisa requisitos, limitações, implicações e alternativas de implementação para a obtenção de atomicidade na aplicação de políticas. Como uma proposta de solução para este problema é definido um protocolo de consenso a ser utilizado entre os participantes da aplicação da política na rede. Tal protocolo é implementado utilizando Web Services e integrado junto ao sistema de gerenciamento baseado em políticas denominado QAME. Por fim, a solução é avaliada através de testes em um cenário de testes composto por um cluster de computadores. / This work presents an investigation of the viability of performing atomic operations during the configuration of multiple devices in a computer network. Specially, the configuration of devices using policy-based network management is addressed. In these cases, the need of atomic operations comes from the fact that a policy application that fails in a certain device can lead to an inconsistent state in a network from which an uniform and global behavior is desired. Thus, this work analyses requirements, limitations, implications, and alternatives for obtaining atomicity in policy deployment. As a solution for this problem, this work defines a consensus protocol to be used by the participants in the policy deployment process. Such protocol was implemented using Web Services and integrated to a policy-based network management system named QAME. At last, the proposed solution is evaluated through a set of tests performed over a testing scenario composed of a high-performance cluster.