Submitted by PPG Ci?ncia da Computa??o (ppgcc@pucrs.br) on 2018-07-16T17:26:53Z No. of bitstreams: 1 ELISA COSTA DIEL_DIS.pdf: 2030569 bytes, checksum: 97dac6ad532e10d4b88caec8f69d282e (MD5) / Approved for entry into archive by Sheila Dias (sheila.dias@pucrs.br) on 2018-07-23T11:12:32Z (GMT) No. of bitstreams: 1 ELISA COSTA DIEL_DIS.pdf: 2030569 bytes, checksum: 97dac6ad532e10d4b88caec8f69d282e (MD5) / Made available in DSpace on 2018-07-23T11:31:48Z (GMT). No. of bitstreams: 1 ELISA COSTA DIEL_DIS.pdf: 2030569 bytes, checksum: 97dac6ad532e10d4b88caec8f69d282e (MD5) Previous issue date: 2017-03-16 / Apesar de o ?gil buscar colabora??o com todos as partes envolvidas, a maioria dos projetos ?geis n?o extende essa colabora??o para o pessoal de opera??es. Problemas de comunica??o s?o um problema recorrente em equipes ?geis que tamb?m ? eminente na rela??o entre desenvolvedores e opera??es. Esta pesquisa visa compreender como a comunica??o acontece em DevOps a partir das percep??es dos praticantes. Para alcan?ar nosso objetivo, realizou-se uma Revis?o de Literatura sobre DevOps e Comunica??o, e conduziu-se um Estudo de Campo com dados qualitativos sendo coletados atrav?s de entrevistas. Os resultados indicam que hoje existem pelo menos tr?s configura??es diferentes de DevOps sendo aplicadas na ind?stria: profissionais Devs e Ops alocados na mesma equipe; uma equipe de DevOps com um conjunto de habilidades compartilhadas; e uma equipe separada de Dev e Ops trabalhando juntos. Apesar dessas configura??es, n?o foram encontradas nenhuma particularidade. Em resumo, os resultados indicam que os membros de equipes co-alocadas e multi-funcionais se comunicam melhor; ? importante trabalhar em conjunto e compartilhar conhecimentos t?cnicos; o poder de decis?o variar? de acordo com a situa??o enfrentada; entre outros. Nossas descobertas ajudam a diminuir a lacuna apontada por Erich, Amrit e Daneva entre Dev e Ops, avan?ando para uma melhor compreens?o de como profissionais DevOps colaboram, ajudando eles a melhorar suas pr?ticas de comunica??o em seu trabalho di?rio. / Even though agile actively seeks collaboration with all kinds of stakeholders, most agile projects do not extend toward the operations people. Issues in communication are a recurring issue in agile teams. Such issues are also eminent in the relationship between developers and operations. This study aims to understand how communication happens in DevOps from the perspective of practitioners. To achieve our goal, a Literature Review on DevOps and Communication was performed, and a Field Study was conducted with qualitative data being collected through interviews. Results revealed that today there are at least three different DevOps configurations being applied in the industry, being: Dev and Ops professionals allocated to the same team; a team of DevOps with a shared skill set; and a separate team of Dev and Ops working together. Despite the configurations, we did not find any particularities. In summary, results show that co-allocated and cross functional team members communicate better; it is important to work together as a single team and share technical knowledge; decision power changes based on the situation that is being faced; among others. Our findings help to narrow the gap pointed out by Erich, Amrit, and Daneva between Dev and Ops, moving towards a better understanding of how DevOps team collaborates by helping practitioners to improve their communication practices in their daily work.

This thesis gives an example of how decision analytic methods can be applied to choices in the adoption of cloud computing. The lifecycle of IT systems from planning to retirement is rapidly changing. Making a technology decision that can be justified and explained in terms of outcomes and benefits can be increasingly challenging without a systematic approach underlying the decision making process. It is proposed that better, more informed cloud adoption decisions would be taken if organisations used a structured approach to frame the problem to be solved and then applied trade-offs using an additive utility model. The trade-offs that can be made in the context of cloud adoption decisions are typically complex and rarely intuitively obvious. A structured approach is beneficial in that it enables decision makers to define and seek outcomes that deliver optimum benefits, aligned with their risk profile. The case study demonstrated that proven decision tools are helpful to decision makers faced with a complex cloud adoption decision but are likely to be more suited to the more intractable decision situations.

What is measured can be improved. That's why Red Hat and the PnT DevOps team are calling for a graphical visualization of metrics and key performance identifiers (KPIs) that will be available to customers, engineers, and stakeholders to review the progress of the PnT DevOps team.

In today’s knowledge-based economy, modern organisations understand the importance of technology in their quest to be considered global leaders. South African markets like others worldwide are regularly flooded with the latest technology trends which can complicate the acquisition, use, management and maintenance of software. To achieve a competitive edge, companies tend to leverage agile methods with the best possible combination of innovative supporting tools as a key differentiator. Software technology firms are in this light faced with determining how to leverage technology and efficient development processes for them to consistently deliver quality software projects and solutions to their customer base. Previous studies have discussed the importance of software development processes from a project management perspective. African academia has immensely contributed in terms of software development and project management research which has focused on modern frameworks, methodologies as well as project management techniques. While the current research continues with this tradition by presenting the pertinence of modern agile methodologies, it additionally further describes modern agile development processes tailored in a sub-Saharan context. The study also aims novelty by showing how innovative sometimes disruptive technology tools can contribute to producing African software solutions to African problems. To this end, the thesis contains an experimental case study where a web portal is prototyped to assist firms with the management of agile project management and engineering related activities. Literature review, semi-structure interviews as well as direct observations from the industry use case are used as data sources. Underpinned by an Activity Theory analytical framework, the qualitative data is analysed by leveraging content and thematic oriented techniques. This study aims to contribute to software engineering as well as the information systems body of knowledge in general. The research hence ambitions to propose a practical framework to promote the delivery of quality software projects and products. For this thesis, such a framework was designed around an information system which helps organizations better manage agile project management and engineering related activities. / Information Science / Ph. D. (Information Systems)

No description available.

Denna uppsats studerar förutsättningarna för att implementera mjukvaruutvecklings - metoden Continuous Delivery (CDE). Problemställningen som lade grunden för studien, var att det inte finns någon enhetlig standard för CDE. Studien ämnade att undersöka om detta innebar att metoden har varierande innebörd inom olika företag och om de således, i viss mån tillämpar skilda tillvägagångssätt med metoden. Ytterligare en aspekt var att se vilka utmaningar företagen upplevde vid övergången till CDE. Att undersöka om det var främst organisatoriska eller utvecklingsrelaterade problem som upplevts. Samt hur de hanterade kommunikation och tillit till medarbetarna och arbetet inom verksamheten under förändringen. För att belysa problemen, beskrevs teori med fokus på organisatoriska och tekniska utmaningar med Continuous - metoderna: Continuous Integration (CI), Continuous Delivery (CDE) och Continuous Deployment (CD). Teorikapitlet samt tidigare studier inkluderade även forskning om kringliggande koncept som DevOps och LEAN. Metoder, vilka kan underlätta implementationen av CDE. Datainsamlingen genomfördes med öppna individuella intervjuer med representanter från sex stycken företag, där de delade med sig av deras erfarenheter av och syn på CDE. Studien visar att anledningen till att företag väljer att arbeta med CDE, är att de vill gå från utvecklingsmetoder, vilka kräver många beslut inför varje förändring, till ett mer flexibelt arbetssätt där de funnit fördelar som: bättre kvalitet på det som levereras, snabbare leverans av affärsvärde till kunder samt kortare feedback - loopar. Företag som gör en övergång till CDE väljer dessutom ofta att inte automatisera hela vägen ut till produktion, enligt CD, då de ser utmaningar med att säkra kvalitén. Studien har identifierat ett antal faktorer som viktiga för en framgångsrik implementering av CDE, samt faktorer som kan resultera i en svår övergång. / This thesis studies the conditions needed for implementing the software development method Continuous Delivery (CDE). The problem identified for the study, is that there is no standardized approach for CDE as of today. The intentions of the study were to determine whether this means that the method will have a shifting tenor within different companies, and if so, will these companies implement the method with different approaches. Another aspect was to determine which types of challenges the companies were faced with during the transition towards CDE. To review whether the challenges were foremost organisational or development related. And how the organisations handled the communication and trust towards the co-workers and the development work within the organisation during the change towards the method. To highlight these issues, we presented theories with focus on organisational and technical challenges with the different Continuous practices were made. The practices being: Continuous Integration (CI), Continuous Delivery (CDE) and Continuous Deployment (CD). The theory chapter and former studies also contains research about surrounding concepts such as DevOps and LEAN methods, which can aim to facilitate the implementation of CDE. The empirical data collection was performed using open individual interviews with informants from six different companies, where they shared their experience and views on the method CDE. The study demonstrates that the reason organisations chose to implement CDE, is that they want to transform from software development methods, which requires a lot of decision making for any change, to a more flexible work procedure, in order to experience benefits such as: better quality of what is delivered, faster deliveries of business value to the customers and faster feedback-loops. Organisations that make the transition towards CDE also tend not to automate all the way to production, as in agreement with CD, this because the organisations identify challenges with assuring that the quality is sufficient. The study has identified a number of factors that are essential for a successful implementation of CDE, along with factors that may result in a less successful implementation.

The last couple of years the trend within the software industry has been to releasesmaller software updates more frequent, to overcome challenges and increase flexibility, to alignwith the swiftly changing industry environment. As an effect, we now see companies moving over tocapitalizing on subscriptions and incremental releases instead of charging for upgrades. By utilizingneo-institutional theory and Oliver’s (1991) strategic response theory, an enterprise systemsvendor’s response to the emerging DevOps trend can be determined.

Migration to the cloud has been a popular topic in industry and academia in recent years. Despite many benefits that the cloud presents, such as high availability and scalability, most of the on-premise application architectures are not ready to fully exploit the benefits of this environment, and adapting them to this environment is a non-trivial task.Therefore, many organizations consider a gradual process of moving to the cloud with Hybrid Cloud architecture. In this paper, the author is making an effort of analyzing particular enterprise case in cloud migration topics like cloud deployment, cloud architecture and cloud management.This paper aims to identify, classify, and compare existing challenges in cloud migration, illustrate approaches to resolve these challenges and discover the best practices in cloud adoption and process of conversion teams to the cloud.

Die Digitalisierung von Produkten und Dienstleistungen durch die Integration moderner Informations- und Kommunikationstechnologien wird in Wissenschaft und Praxis intensiv und anhand diverser Modelle diskutiert. Smart Service Systeme, als Gegenstand dieser Diskussion, skizzieren den Einsatz „intelligenter“ Produkte (sog. „Smart Products“) in flexibilisierten Dienstleistungssystemen, mit dem Ziel, sensorisch erfasstes Kontextwissen zur automatisierten Anpassung von Leistungen an individuelle Kundenbedürfnisse zu nutzen. Dieses ambitionierte Ziel birgt eine Reihe von Forschungsaufgaben für die Wirtschaftsinformatik (WI). Denn neben den diversen gestalterischen Herausforderungen, die für die Entwicklung einzelner Bestandteile zu lösen sind, ist gerade die Verknüpfung einzelner Forschungsperspektiven, wie dem Service Systems Engineering oder der Data Science, eine Herausforderung, in der die WI als Vermittlerin auftreten kann. Die vorliegende Dissertation untersucht vor diesem Hintergrund Smart Service Systeme als Modell für datengetriebene Wertschöpfung im Maschinen- und Anlagenbau und erarbeitet Wissen in vier Bereichen: (1) der Identifikation von domänenspezifischen Anforderungen an die zugrundeliegenden Informationssysteme, (2) der Integration von produkt- und dienstleistungsbezogenen Datenströmen, (3) der Verfolgung von herstellerübergreifenden Plattformlösungen für den kontinuierlichen Betrieb und das Management von Smart Services und (4) dem Kontexttransfer entwickelter Lösungen in weitere Domänen. Die Arbeit generiert auf diese Weise anwendungsorientiertes Wissen über die Entwicklung und den Einsatz von Smart Service Systemen im Maschinen- und Anlagenbau. Auf übergeordneter Ebene entsteht ein Beitrag zur bislang ausstehenden Etablierung einer gemeinsamen Gestaltungsgrundlage zwischen beteiligten Disziplinen, in dem demonstriert wird, wie eine gemeinsame Entwicklung zukünftiger, datengetriebener Wertschöpfungssysteme gelingen kann.

The machine learning (ML) industry has taken great strides forward and is today facing new challenges. Many more models are developed, used and served within the industry. Datasets that models are trained on, are constantly changing. This demands that modern machine learning processes can handle large number of models, extreme load and support recurring updates in a scalable manner. To handle these challenges, there is a concept called model serving. Model serving is a relatively new concept where more efforts are required to address both conceptual and technical challenges. Existing ML model serving solutions aim to be scalable for the purpose of serving one model at a time. The industry itself requires that the whole ML process, the number of served models and that recurring updates are scalable. That is why this thesis presents an open-source framework for large-scale ML model serving that aims to meet the requirements of today’s ML industry. The presented framework is proven to handle a large-scale ML model serving environment in a scalable way but with some limitations. Results show that the number of parallel requests the framework can handle can be optimized. This would make the solution more efficient in the sense of resource utilization. One avenue for future improvements could be to integrate the developed framework as an application into the open-source machine learning platform STACKn.