CoordenaÃÃo de AperfeÃoamento de Pessoal de NÃvel Superior / O crescente debate sobre mediÃÃo de desempenho entre profissionais e entre pesquisadores da Ãrea de gestÃo organizacional ainda nÃo foi suficiente para consolidar as prÃticas de mediÃÃo na indÃstria da construÃÃo. Assim, o presente trabalho tem por objetivo identificar as oportunidades para o aprimoramento dos Sistemas de MediÃÃo de Desempenho (SMD) nas Empresas de ConstruÃÃo Civil. Para tanto, realizou-se uma anÃlise da literatura de mediÃÃo de desempenho, a partir da literatura de disciplinas como contabilidade, estratÃgia, recursos humanos, produÃÃo e gestÃo de operaÃÃes, marketing e comportamento organizacional. Esta revisÃo propiciou a elaboraÃÃo de um instrumento de auditoria para os SMDs de empresas construtoras. Este mesmo instrumento foi validado por cinco especialistas com grande experiÃncia com o tema e serviu tambÃm como protocolo para anÃlise de 22 modelos identificados na literatura especÃfica da construÃÃo, nacional e internacional, nos Ãltimos 25 anos, o que permitiu mapear as principais caracterÃsticas, papÃis e processos das prÃticas no setor. Tal diagnÃstico foi expresso quantitativamente e explorado em profundidade, identificando gaps e oportunidades de melhoria. ApÃs este estudo teÃrico realizou-se um estudo de caso mÃltiplo com trÃs empresas construtoras cearenses. Como resultados do estudo de caso, foi possÃvel analisar as sistemÃticas de mediÃÃo de desempenho das empresas estudadas, realizar proposiÃÃes de melhorias e aplicar o instrumento de auditoria proposto nesta pesquisa. Elaborou-se tambÃm uma matriz que relaciona os papÃis do sistema com os seus processos, a partir da visÃo dos mesmos especialistas que validaram o instrumento de auditoria. Por fim propÃs-se um framework que distingue um Sistema de Indicadores (medir) de um Sistema de MediÃÃo de Desempenho (medir e gerir o desempenho) e que classifica os Sistemas de MediÃÃo a partir do instrumento de auditoria e da matriz de papÃis versus processos. Esta classificaÃÃo permite que as empresas verifiquem se seus SMDâs sÃo capazes de suportar o uso para o qual foram designados. Conclui-se que, apesar das inÃmeras iniciativas de melhoria analisadas, a indÃstria da construÃÃo ainda necessita de modelos de mediÃÃo que ofereÃam diretrizes operacionais mais contundentes. Se a evoluÃÃo dos SMDs no setor da ConstruÃÃo Civil à lenta, como aponta a literatura, alguns dos aspectos que confirmam essa deficiÃncia foram mapeados, podendo, portanto, ser aprimorados e discutidos no ambiente acadÃmico e empresarial. / The growing debate on measuring performance among professionals and researchers from organizational management is still not enough to consolidate the measurement practices in the construction industry. Thus, this study aims to identify the opportunities for improvement of the Performance Measurement Systems (PMS) in construction management. To do so, an analyze of the performance measurement literature was carried out from disciplines as accounting, strategy, human resources, production and operations management, marketing and organizational behavior. This analysis provides a proposition of an audit tool to evaluate the PMS in Construction Companies. The proposed tool was validated by expert researchers with great experience with the PMS theme. The audit tool was used to analyze 22 models identified in the specific literature of PMS in construction over the last 25 years, which provided a mapping of the key features, roles and processes of measurement practices in that industry. This diagnostic use quantitative techniques and explore in depth the qualitative aspects, identifying gaps and opportunities of improvement. After the theoretical study, a multiple case study was carried out in three construction companies from CearÃ, northeast of Brazil. As results of the case studies, an analyzing of the PMS of the companies was carried out. Moreover, propositions of improvements for the PMS were carried out and the audit tool was applied. In following, a matrix that correlates the roles and the processes of a PMS was formulated based in the experience of the same experts that validated the audit tool. With experience acquired, a framework that presents a holistic view of the measurement process was proposed, emphasizing the difference between indicators system and performance measurement system. This framework classify the PMS based on the audit tool and the matrix that correlates the roles and the processes of a PMS. The classification enable that construction companies evaluates if their PMS are able to support the uses to which they were designated. As conclusion, despite of the great number of initiatives to improve the PMS in Construction that were analyzed, the sector need to models of PMS that containing more scathing operational guidelines. If the evolution of PMS in Building Construction it is slow, as pointed out in literature, some evidences that confirm this thesis were mapped that enable propositions of improvement by academics and professionals. A research agenda was suggested, from the gaps and improvement opportunities found.

Com o crescimento das redes de computadores e, principalmente, de sua importância para as organizações, o gerenciamento de redes tornou-se fundamental. Contudo, o gerenciamento de redes é um processo difícil dada sua complexidade e mudanças freqüentes em sua configuração. O ideal seria que sistemas pudessem fazer o trabalho de administradores de redes, reduzindo com isso a sobrecarga de trabalho dos administradores, ou seja, essas pessoas poderiam realizar outras tarefas enquanto o sistema ficaria gerenciando a rede. O principal objetivo deste trabalho é propor um paradigma que, à medida que seja constatada a ocorrência de algum problema na rede, se tenha um módulo com inteligência suficiente para diagnosticá-lo, determinando porque aquele problema ocorreu. O trabalho descreve um sistema especialista para a gerência de redes que integrado com um sistema de registro de problemas. Módulos especializados, orientados à análise de aspectos específicos do comportamento da rede, efetuam uma análise das características e do status da mesma, filtrando eventos e tentando prover resposta automatizada e/ou recomendações sobre cursos de ação para as anormalidades percebidas. Portanto, com esse objeto, definiu-se um sistema denominado MAD (Módulo de Automatização de Diagnóstico), no qual são definidas regras para diagnosticar os problemas ocorridos e também tentar prover ao usuário a maior qualidade de serviço possível. Este sistema efetua monitorações diárias sobre objetos da MIB II (Management Information Base) para tentar localizar os problemas da rede e então gerar alertas ao administrador da rede dependendo de sua gravidade. Este trabalho é um sub-projeto do projeto CINEMA (Cooperative Integrated Network Management). No projeto CINEMA, foram especificados módulos de manuseio de registros de problemas numa base de dados (Sistema de Trouble Ticket) e uma plataforma básica para configurar a obtenção de informações sobre a rede - acesso a objetos gerenciados (Sistema de Alertas). O MAD atua como um integrador entre esses dois sistemas, filtrando eventos e gerando alertas ao administrador da rede na forma de um trouble ticket, isto é, um registro de problemas. A validação deste sistema foi realizada através da implementação de um protótipo. O protótipo utiliza o protocolo SNMP (Simple Network Management Protocol) para fazer polling em objetos da MIB II de determinados componentes da rede. Sua implementação foi 661 para que se pudesse fazer o refinamento das regras, tornando-as mais de acordo com a rede monitorada. / With the growth of computer networks and, mainly, of its importance to organizations, the management of networks has become fundamental. However, network management is a hard process given its complexity and its frequent configuration changes. The ideal would be if the systems could play the role of network managers, thus reducing the job overload from the managers. It means that these managers could perform other tasks while the system manages the network itself. The main objective of this work is to propose a paradigm so that, as any trouble is found in the network, a module intelligent enough to diagnose this problem will identify the reason why that has occurred. The work describes an intelligent system for network management which is integrated to a trouble ticket system. Specialized modules, directed to the analysis of specific aspects of the network behavior, bring about an analysis of its characteristics and status, filtering events and trying to supply an automated answer and/or advices about action paths to deal with the abnormalities. Therefore, with this goal, a system called MAD (Diagnosis Automatization Module) has been defined, in which rules to diagnose the most common troubles are specified and that tries to supply the user with a quality of services as higher as possible. This system executes daily monitoration on MIB II (Management Information Base) objects trying to locate the troubles in the network and then generating alerts to the network manager depending on its severity. This work is a sub-project for the project CINEMA (Cooperative Integrated Network Management). The project CINEMA specified modules of trouble ticket handling in a database (Trouble Ticket System) and a basic platform to configure the acquisition of information about the network - access to managed objects (Alerts System). MAD acts as an integrator between these two systems, filtering events and generating alerts to the network administrator in the form of a trouble ticket. The validation of this system has been done through the implementation of a prototype. The prototype uses the SNMP protocol (Simple Network Management Protocol) to do polling in MIB II objects located in certain network components. Its implementation has been useful to refine the rules, making them fit the network under management.

Local Government in South Africa has undergone much transformation since 2000. Although much of the change has been to correct imbalances, inequities and disparities within our local communities as a result of Apartheid, change has also been motivated by National Government’s realisation that, as with governments throughout the world, there is a need to modernise all spheres of Government. Part of this transformation process at a local government level in South Africa has been to ensure that municipalities become more responsive to the communities’ needs. The guiding principles for this transformation are contained in the White Paper on the Transformation of the Public Service (1995) and the Batho Pele White Paper (1997). This has informed the Municipal Systems Act: Act 32 of 2000 of which Chapter 6 determines that municipalities will have a performance management system to promote a culture of performance management amongst the political structures, political office bearers, councillors and administration. The performance management system must ensure that the municipality administers its affairs in an economical, effective, efficient and accountable manner. A literature review contained in this research, indicates that internationally, implementing performance management systems at a local government level is impact upon by a number of factors such as the organizational culture of an institution. This research, which has been grounded within a constructivist paradigm, describes the impact organizational culture has had on the implementation process of the performance management system at Buffalo City Municipality. Interviews were conducted amongst the Section 57 employees (i.e. the Directors) and those employees directly responsible for implementing performance management. The four Directors, two General Managers and the portfolio councilor were interviewed. Semi-structured interviews were conducted with the interviewees and this culminated in a total of 10 hours of interviewing. The protocols were analyzed using the guidelines suggested by Boyatzis (1998) and the findings are detailed in two chapters. The findings of the research were that the implementation of the performance management system at Buffalo City Municipality has been impacted on by the dichotomy between the political and administrative leadership, resulting in non-implementation of council resolutions, the lack of an organizational strategy, poor institutional arrangements and inadequate resource allocation, are reported. However, the most important finding was the impact that a culture of fear has had on the implementation process. This culture of fear and its impact on the implementation of performance management system is described as the most pervasive and insidious of all the findings to have negatively impacted on the implementation process. This research ends with recommendations for further research and it is argued that each organization has its own unique organizational culture. The conclusion is that no single typology, as contained in the literature, which can account for the specific impact organizational culture will have on the implementation process of a performance management system at local government level in South Africa. Consequently, implementers of performance management systems must assess the unique characteristics of each organization’s culture prior to implementation, in order to evaluate its impact that the organizational culture can have on the process.

This study evaluated how the communication of the new performance management system by the City of Windhoek influenced the beliefs of its employees and consequently to determine the effects of the communication process on the degree to which buy-in and readiness is created. Armenakis, Harris and Field's (1999) five element model of examining the change recipient's beliefs, namely, discrepancy, appropriateness, principal support, efficacy, and valence was used as a framework for this study. These beliefs play a major role in affecting behaviours of the change recipients (employees) toward the implementation of a new change initiative in an organisation. Discrepancy involves assessing if employees believe that a need for change does really exist in the organization. Appropriateness seeks to find out from the change recipients whether the performance management system being introduced in the organization is an appropriate reaction to the need and the vision of the organisation. Principal support describes the support from change agents and opinion leaders for the organizational change. Efficacy refers to a belief in one's capabilities to organize and execute the courses of action required to produce given attainments. Valence refers to the intrinsic and extrinsic attractiveness (from the change recipient's perspective) associated with the perceived outcome of the change. Literature on organizational change, organizational communication and change implementation were used, to gain and highlight insights regarding the role of change communication during change implementation, and to provide the conceptual framework for the research design and analysis. A single case study research method was used under a mixed research paradigm. The study used a survey questionnaire that was sent to 711 employees; a sufficient and representative 202 questionnaires were returned from all levels of the organization, yielding a response rate of 22.4%. In addition, three focus group interviews of a representative cross section of staff were conducted as also document analysis to collect data for the study. The summative average percentages of levels of agreements and disagreements of the respondents for each of the five change recipient's beliefs indicate a fairly high degree of agreement (buy-in) and considerably low degree of disagreement (resistance) among the respondents to nearly all five change recipient's beliefs. Discrepancy received the highest degree of agreement (buy-in) 84%; efficacy received 73%, appropriateness 66% and valence 59% whereas principal support scored the lowest degree of agreement of about 40% of the respondents. Despite the high degree of agreement (buy-in) for these elements, there was also a notable number of the respondents who neither agreed nor disagreed to some of the five change recipient's beliefs. In particular, principal support had 38%, valence had 28%, and appropriateness had 22% who neither agreed nor disagreed. The degree of disagreement (resistance) is relatively low in almost all five change recipient‟s beliefs except for principal support which indicated slightly higher levels of disagreement (19%). The results of the study indicate that the communication methods e.g., the road show, leaflets and brochures, discussions and training sessions used during the pre-implementation phase of the new performance management system influenced the change recipient‟s beliefs on the discrepancy for the new performance management system. The evidence is that these communication methods and actions helped to create a sense of urgency and the need to implement a new performance management system among employees. The new performance management system was also believed to be appropriate and its appropriateness was justified through the use of various communication methods which provided detailed explanations on how the new performance management system was different to the earlier performance appraisal system. Notwithstanding the belief in the appropriateness of the new performance management system, the study found that the absence of such a performance management system over a long period of time had cultivated a hidden but unifying shared set of values, beliefs and assumptions among employees that seem not to fit or be suitable to the demands of the new performance management system and the current challenges facing the organization. The employees themselves noted this discrepancy. Hence, they argued, there is a need to bring about new culture with new sets of values, beliefs and assumptions that will better suit and support the current demands facing the organisation and its members. The study found that although employees significantly believe that they have high levels of personal self-efficacy, the change message that was communicated to them did not in itself appear to help to create and further develop this employee self-efficacy. The change message mainly focussed on the need for training, and the design of job descriptions and technical aspects of the performance management system. The study revealed that the employees' belief in principal support of the new performance management system was that principal support was seen as less than sufficient or not as evident as they would have liked. Employees remarked that principal support was only conveyed during the initial stage of communication, thereafter, principal support from senior management was seen to have declined throughout the pre-implementation phase. Nonetheless, immediate supervisors and managers were believed to more supportive than senior management. The literature points to a lack of belief in principal support as an important cause of a loss momentum in change initiatives, which in turn may result in cynicism and greater resistance, especially by those who have vested interest in maintaining the status quo. Valence refers to the intrinsic and extrinsic attractiveness (from the change recipient's perspective) associated with the perceived outcome of the change. The results show that respondents have mixed beliefs about the personal benefits associated with the new performance management system. Thus, 63% of the respondents believe that it will benefit them, 75% believe that it will increase their self fulfilment, 68% believe that it will increase their feeling of accomplishment, 45% believe that their fringe benefit will remain the same after it is implemented, and 58% believe that they could earn high performance return. However, the result also shows that there are still many respondents who neither agree nor disagree to the presence of personal valance with regard to the new performance managements system. This study established that the use of various communication methods persuaded the majority of the City of Windhoek employees to view the new performance management system overall in a favourable light. However, interviewees criticised the communications process on the grounds that most of these methods were only used once and were not repeated, as they would have liked. Lessons on this item from other change studies are that when communicating new change initiatives, change agents should use different methods of communication and as many times as possible to influence the belief of employees. This report is structured as follows: Section 1 provides an abstract and introduction to the study. Section 1.2 presents a brief review of core theories and recent empirical studies relevant to the research problem. Section 1.3 provides a description of the research method followed in addressing aim and objectives of this study. Section 1.4 presents the results of the evaluation report and section 1.5 discusses the findings and make conclusive recommendations. Section 2 of this study complements subsection section 1.2 (in section 1) with a more detailed review of the literature on organizational change, organizational communication and change implementation. This more detailed review was conducted before the field work and write up of Section 1. The literature incorporated into Section 1 was distilled from this more detailed review. Similarly, Section 3 of this study complement subsection 1.3 (of section 1) with an extended description of the research design and research procedure followed in the study.

Masters in Public Administration - MPA / This study found that the performance appraisal process, in particular, is an instrument of the performance appraisal system that contributes tremendous value to planning and implementation. The research found that the performance appraisal process in the City of Cape Town can be improved through addressing the process and improving the approaches followed by the managers in daily appraisals, which can be done by following the complete process and including all the steps it involves, and by having a positive attitude towards the performance appraisal process. This mini–thesis also provides a set of conclusions and recommendations for the performance management and performance appraisal system, particularly regarding appraisal process enhancement. It is hoped that the lessons learned through this study will inform current practices and provide information for future research at local government level. / South Africa

Magister Legum - LLM / This study is a response to the dilemma of poor service delivery or the lack thereof. In this regard, this study posits the professionalisation of local government as part of the solution. The focus is on the administrative arm of local government, which is the major conduit for service delivery. Professionalisation of local government is a broader theme. For the present purposes, focus will be devoted to the competency component which entails attracting qualified personnel competent to discharge local government responsibilities. However, it is not limited to attracting already competent and professional staff. It also entails developing the skills of existing staff. This definition, in essence, relates to qualification through training, learning and specialisation.11 In essence, professionalisation of local government ensures that all employees act and behave in a professional way. In this regard, this study seeks to identify the competency standards set by the legislative framework and then explore the legal avenues for enforcing compliance, by the municipal administration, with such standards. This requires one to look at and answer the following questions: What constitutes municipal staff? ; What is the content of the competency framework in question? ; What are the enforcement mechanisms currently in place? ; Whose role is it to enforce compliance with the competency framework?; Broadly speaking, enforcement can take two forms: hard enforcement and soft enforcement. The hard form of enforcement relates to giving incentives for compliance with the competency framework and dismissal for non-compliance. Softer enforcement, in turn, relates to correction and monitoring. / South Africa

Masters in Public Administration - MPA / This study investigates the challenges facing the management of public enterprises in the Democratic of Congo. The study is based on the hypothesis that poor performance by state-owned enterprises in the Democratic Republic of the Congo has been influenced by inadequate management which has led to the dissatisfaction of citizens. Due to this problem increasing attention is being paid by the DRC Government to encouraging private sector involvement in services such as water and electricity. This study investigates the arguments for and against privatization, especially in the water and electricity sectors. In pursuing this debate the study explores the following main research question: To what extent will the DRC Reform Law aimed at transforming the commercialisation of public enterprises enhance the effectiveness of basic services such as water and electricity services in the DRC? A case study of the city of Bukavu in the eastern DRC is used to explore this question. / South Africa

After a review of the literature relevant to performance management systems both over time and across different types of organizations, this thesis confines its research to a case study of the development and implementation of a performance management system in a non-governmental organization (NGO), the Public Service Accountability Monitor (PSAM). Collation of the relevant data is followed by a discussion of the development and implementation of the performance management system at the PSAM over a 5-year period from 1 June 1999 to 31 May 2004. This is considered in terms of the PSAM’s achievement of commitments to Funders and the concurrent development of the performance management system. Next an analysis of major themes that emerged from the research, in terms of important items for consideration in the development and implementation of a performance management system in an NGO, and areas for possible future improvements to the system is presented. After analyzing the relevant information, it became apparent that the performance management system has no direct bearing on the ability of the PSAM to achieve its stated commitment to Funders. However, these short-term focused expectations of the performance management system are outweighed by the positive contributions that have been made by its introduction, specifically in the area of training and development. This important aspect of capacity building and staff empowerment speaks to the long term sustainability of the organization. Although the PSAM’s performance management system undergoes continual improvement, significant inroads have been made into providing a sensible, clear and dynamic solution to the problem of rewarding efficient and effective performance. The PSAM has indeed benefited from the introduction of the performance management system in a number of ways. It is evident that these benefits could be applicable to other NGOs.

The South African Constitution ensures the right to adequate water, health services and a healthy environment (Government Gazette, 1996: 1251, 1255). Associated legislation and national strategies require that water resources and water services are rendered in a sustainable manner to ensure the rights enshrined in the Constitution are realised (Trialogue, 2010: 34; Government Gazette, 1998(c):11; Department of Water Affairs (DWA), 2003:9). South Africa is the 30th driest country in the world (DWA, 2013: iii), has impounded most of its surface water resources and has utilised 40 percent of the groundwater resource with remaining groundwater water quality uncertain (DWA, 2013:6-7). Despite achieving the basic water and sanitation Millennium Development Goals approximately 2.2 million and 4.5 million households still require access to basic services respectively (DWA, 2013:28-29). South Africa’s ten water boards play a key role in bulk water service provision. They supply 57 percent of the countries domestic water supply (DWA, 2013:19) and are required to provide sustainable water services (Government Gazette, 1997:35). The Department of Water Affairs (Department of Water and Sanitation (DWS) as of June 2014) is mandated to monitor the performance of water boards and monitors of water boards through a Share Holders Compact (SHC) (Government Gazette, 1997: 50). The aim of the research is to achieve the following objectives: 1. Describe best practice for sustainability monitoring frameworks for water services. 2. Outline South Africa’s water services mandatory and voluntary requirements with regards to sustainability monitoring. 3. Evaluate South Africa’s current water services performance monitoring framework for water boards against findings from goals one and two. 4. Develop an improved framework for assessing South African water board’s performance in their compliance and sustainability journey. 5. Demonstrate the improved framework functionality with a sample of water boards audited performance data from the 2012/13 financial year. A qualitative normative theory evaluation research method was utilised to achieve first three objectives of the research. The objectives to first understand current best practice for sustainability monitoring frameworks and the mandatory requirements for water services sustainability monitoring frameworks in South Africa was achieved via an extensive literature review. The evaluation research method was used where South Africa’s current water services monitoring framework for water boards was evaluated for its adequacy to monitor sustainability compared with legislated requirements, national and international best practices (Hall et al, 2004: 55). Data collection for the evaluation research was sourced via documentation analysis. The 2012/13 water board audited annual reports, relevant national water services legislation, national and global water service sustainability monitoring best practices were critically reviewed. The review yielded a “thick description” of sustainability performance monitoring framework requirements that was used to evaluate the SHC (Holliday, 2002: 79). Despite the mandatory participation requirement for successful sustainability performance monitoring being met, the SHC was found to have inadequate aspects. It lacks an outcomes based approach, does not include environmental indicators, does not allow water board sustainability performance comparison and cannot indicate where individual water boards are on their sustainability journey. An improved sustainability performance monitoring framework for South African water boards is therefore proposed. This was developed with the assistance of an expert focus group drawn from multiple disciplines and organizations relevant to water board sustainability (Litosseliti, 2003: 8 and Hall et al, 2004: 51). The functionality of the framework is then demonstrated using 2012/13 audited performance data sourced from the annual reports of Overberg, Amatola, Rand and Umgeni Water boards. The proposed Water Services Sustainability Monitoring Framework (WSSMF) quantifies performance in terms of 10 water utility outcomes that are measured with 88 performance indicators that encompass the financial, social, environmental and governance dimensions of sustainability. The WSSMF demonstrated that the legislated intentions and best practice recommendations can be incorporated into an improved version of the SHC. Further refinement of the WSSMF is required. In depth engagement with DWS and all the water boards would promote the refinement of the indicator set, performance thresholds for indicators and indicator weighting. Further testing through a pilot project initiative would allow the WSSMF to be developed into a robust and adequate sustainability performance monitoring framework for South African water boards (Rametsteiner et al, 2011: 64; Muga and Mihelcic, 2008: 438; McAlphine and Birnie, 2005: 247 and van Leeuwen et al, 2012: 2192).

Fundamental properties of cloud computing such as resource sharing and on-demand self-servicing is driving a growing adoption of the cloud for hosting both legacy and new application services. A consequence of this growth is that the increasing scale and complexity of the underlying cloud infrastructure as well as the fluctuating service workloads is inducing performance incidents at a higher frequency than ever before with far-reaching impact on revenue, reliability, and reputation. Hence, effectively managing performance incidents with emphasis on timely detection, diagnosis and resolution has thus become a necessity rather than luxury. While other aspects of cloud management such as monitoring and resource management are experiencing greater automation, automated management of performance incidents remains a major concern. Given the volume of operational data produced by cloud datacenters and services, this thesis focus on how data analytics techniques can be used in the aspect of cloud performance management. In particular, this work investigates techniques and models for automated performance anomaly detection and prevention in cloud environments. To familiarize with developments in the research area, we present the outcome of an extensive survey of existing research contributions addressing various aspects of performance problem management in diverse systems domains. We discuss the design and evaluation of analytics models and algorithms for detecting performance anomalies in real-time behaviour of cloud datacenter resources and hosted services at different resolutions. We also discuss the design of a semi-supervised machine learning approach for mitigating performance degradation by actively driving quality of service from undesirable states to a desired target state via incremental capacity optimization. The research methods used in this thesis include experiments on real virtualized testbeds to evaluate aspects of proposed techniques while other aspects are evaluated using performance traces from real-world datacenters. Insights and outcomes from this thesis can be used by both cloud and service operators to enhance the automation of performance problem detection, diagnosis and resolution. They also have the potential to spur further research in the area while being applicable in related domains such as Internet of Things (IoT), industrial sensors as well as in edge and mobile clouds. / Grundläggande egenskaper för datormoln såsom resursdelning och självbetjäning driver ett växande nyttjande av molnet för internettjänster. En följd av denna tillväxt är att den underliggande molninfrastrukturens ökande storlek och komplexitet samt fluktuerade arbetsbelastning orsakar prestandaincidenter med högre frekvens än någonsin tidigare. En konsekvens av detta blir omfattande inverkan på intäkter, tillförlitlighet och rykte för de som äger tjänsterna. Det har därför blivit viktigt att snabbt och effektivt hantera prestandaincidenter med avseende på upptäckt, diagnos och korrigering. Även om andra aspekter av resurshantering för datormoln, som övervakning och resursallokering, på senare tid automatiserats i allt högre grad så är automatiserad hantering av prestandaincidenter fortfarande ett stort problem. Denna avhandling fokuserar på hur prestandahanteringen i molndatacenter kan förbättras genom användning av dataanalystekniker på de stora datamängder som produceras i de system som monitorerar prestanda hos datorresurser och tjänster. I synnerhet undersöks tekniker och modeller för automatisk upptäckt och förebyggande av prestandaanomalier i datormoln. För att kartlägga utvecklingen inom forskningsområdet presenterar vi resultatet av en omfattande undersökning av befintliga forskningsbidrag som behandlar olika aspekter av hantering av prestandaproblem inom i relevanta tillämpningsområden. Vi diskuterar design och utvärdering av analysmodeller och algoritmer för att upptäcka prestandaanomalier i realtid hos resurser och tjänster. Vi diskuterar också utformningen av ett maskininlärningsbaserat tillvägagångssätt för att mildra prestandaförluster genom att aktivt driva tjänsternas kvalitet från oönskade tillstånd till ett önskat målläge genom inkrementell kapacitetoptimering. Forskningsmetoderna som används i denna avhandling innefattar experiment på verkliga virtualiserade testmiljöer för att utvärdera aspekter av föreslagna tekniker medan andra aspekter utvärderas med hjälp av belastningsmönster från verkliga datacenter. Insikter och resultat från denna avhandling kan användas av både moln- och tjänsteoperatörer för att bättre automatisera detekteringen av prestandaproblem, inklusive dess diagnos och korrigering. Resultaten har också potential att uppmuntra vidare forskning inom området samtidigt som de är användbara inom relaterade områden som internet-av-saker, industriella sensorer, och storskaligt distribuerade moln eller telekomnätverk. / Cloud Control / eSSENCE