As terascale supercomputers become more common, and as the high-performance computing community turns its attention to petascale machines, the challenge of providing effective resource management for high-end machines grows in both importance and difficulty. These computing resources are by definition expensive, so the cost of underutilization is also high, e.g., wasting 5% of the compute nodes on a 10,000 node cluster is a much more serious problem than on a 100 node cluster. Moreover, the high energy and cooling costs incurred in maintaining these high end machines (often millions of dollars per year) can be justified only when these machines are used to their full capacity. On large clusters, conventional jobs schedulers are hard-pressed to achieve over 90% utilization with typical job-mixes. A fundamental problem is that most conventional parallel job schedulers only support static scheduling, so that the number of processors allocated to an application cannot be changed at runtime. As a result, it is common to see jobs stuck in the queue because they require just a few more processors than are currently available, resulting in long queue wait times for applications and low overall system utilization. A more flexible and effective approach is to support dynamic resource management and scheduling, where the number of processors allocated to jobs can be expanded or contracted at runtime. This is the focus of this dissertation --- dynamic resizing of parallel applications. Dynamic resizing significantly improves individual application turn-around time and helps the scheduler to achieve higher machine utilization and job throughput. This dissertation focuses on the potential benefits and challenges of dynamic resizing using ReSHAPE, a new framework for dynamic Resizing and Scheduling of Homogeneous Applications in a Parallel Environment. It also details several interesting and effective scheduling policies implemented in ReSHAPE and demonstrates their effectiveness to improve overall cluster utilization and individual application turn-around time. / Ph. D.

With increasing world population, demand for underground construction is expected to accelerate in the future. Design of tunnels in rock is still largely empirical, while rock failure in underground mines and tunnel construction continues to claim lives. A seismic method to aid in increasing safety during excavation is tomography. Seismic tomography is a non-invasive technique to map the stress changes induced by mining ahead of the active face. Seismic tomography maps the velocity distributions of elastic waves traveling through a rock mass. The velocity distributions mapped in the tomograms can relate to anomalies in the rock such as fracture zones and highly concentrated stresses. In order to develop a relationship between stress and elastic wave velocity, laboratory tests in a controlled environment are required. In the current study tomographic tests were conducted on Berea sandstone and Five Oaks limestone samples. The stress redistribution in the sandstone samples could be imaged by mapping velocity distributions. On an unconfined test the sandstone sample acted much like a coal mine pillar where the stress redistributes to the least confined area. On a sandstone test where the sample was indented by a steel platen the velocity contrast was seen directly under the load and the velocity remained almost unchanged over the rest of the sample. For the limestone tests, the stress redistribution could not be mapped in the tomograms. The ability to map the stress distribution in the tomograms were attributed to the elastic and non-elastic characteristics of the stress-strain curve. For sandstone, a porous rock, the stress redistribution could be mapped and for limestone, a stiff rock, the stress redistribution could not be mapped. A field data acquisition system to apply tomography to ground control problems in a mine was designed and calibrated. Data acquisition hardware were assembled and programmed in LabVIEW to collect seismic data in a mine. The design of a geophone array that will fit into a miniature 5.08 cm (2 in) diameter borehole is presented. / Master of Science

Despite considerable research efforts, the relationship between inequality and demand for redistribution remains a highly contested topic within comparative political economy. This paper argues that a central yet widely overlooked mechanism linking macro-level income inequality to preferences for redistribution has to with the micro-level implications of certain externalities of inequality. Focusing on fear of crime, as one such externality, I argue that because (i) in- equality and crime are positively related, and (ii) because crime and fear of crime have a negative effect in individual utility, it follows that increasing in- equality should have a positive effect on support for redistribution. Importantly, however, the argument of this paper also recognises that redistribution is but one of several means through which a concern about crime can be addressed, with the most relevant alternatives being increased policing and harsher punitive measures. Drawing on literatures in criminology and political sociology, I theorise that a key determinant of this choice | between redistribution and policing/punishment as alternative approaches to dealing with crime | is the level of ethnic heterogeneity in the population. Taken together, therefore, this paper's argument implies that inequality will have differential effects on support for redistribution in different contexts: in cases where the population is homogenous, fear of crime - and by consequence inequality - will boost demand for redistribution, whilst no such effects will follow in contexts of high heterogeneity. Using a two-step statistical methodology, I analyse Eurobarometer and ESS data from 21 OECD countries and find persuasive empirical support for my theoretical expectations. Fear of crime is more strongly associated to support for redistribution when the level of population heterogeneity remains low, whilst the opposite holds true for the relationship between fear of crime and support for policing and punishment.

La présente étude se concentre sur le travail de Nancy Fraser sur la justice sociale, lequel a suscité beaucoup d’intérêt dans la littérature au cours des dernières années. La reconnaissance et la redistribution sont les deux piliers originaux de son approche: les désavantages dont souffrent les gens dus au dénigrement culturel ou à la privation économique. Ces deux concepts servent à diagnostiquer et fournir le soutien moral aux multiples luttes que les victimes d’injustice entreprennent avec l’objectif d’établir une participation plus égalitaire à la société. Cependant, que peut-elle dire cette approche des groupes qui sont marginalisés et cherchent l’autogouvernance (ou la séparation même) plutôt que l’intégration dans la société? Le travail de Fraser manifeste une résistance envers les droits du groupe, et un silence quant à l’autodétermination. Mon intervention prend comme objectif d’inclure ces formes d’injustice dans son approche, la rendant plus sensible aux dynamiques des groupes et capable de répondre à leurs revendications trop souvent négligées sous prétexte de l’égalité. La question est, l’égalité de qui? / The focus of this study is Nancy Fraser’s work on social justice, which has gained prominence in the literature over the past few years. The two original pillars of her approach are redistribution and recognition – the injustices that individuals face as a result of economic hardship or cultural denigration. These two concepts serve to diagnose and provide moral backing to the multiple struggles that individuals undertake with the aim of a more equitable way of participating in society. But what does this approach have to say about marginalized groups who seek greater autonomy, or perhaps even separation, rather than further participation in society? Fraser’s work has manifested resistance to sanctioning group difference, and silence on the issue of self-determination. I aim to build these claims into her approach, ultimately to render it more sensible to group dynamics and more capable to respond to their demands all too often neglected under the pretext of equality. The question is, equality of whom?

L'équilibrage de charge est une étape importante conditionnant les performances des applications parallèles. Dans le cas où la charge varie au cours de la simulation, il est important de redistribuer régulièrement la charge entre les différents processeurs. Dans ce contexte, il peut s'avérer pertinent d'adapter le nombre de processeurs au cours d'une simulation afin d'obtenir une meilleure efficacité, ou de continuer l'exécution quand toute la mémoire des ressources courantes est utilisée. Contrairement au cas où le nombre de processeurs ne varie pas, le rééquilibrage dynamique avec un nombre variable de processeurs est un problème peu étudié que nous abordons ici.Cette thèse propose différentes méthodes basées sur le repartitionnement de graphe pour rééquilibrer la charge tout en changeant le nombre de processeurs. Nous appelons ce problème « repartitionnement M x N ». Ces méthodes se décomposent en deux grandes étapes. Dans un premier temps, nous étudions la phase de migration et nous construisons une « bonne » matrice de migration minimisant plusieurs critères objectifs comme le volume total de migration et le nombre total de messages échangés. Puis, dans un second temps, nous utilisons des heuristiques de partitionnement de graphe pour calculer une nouvelle distribution optimisant la migration en s'appuyant sur les résultats de l'étape précédente. En outre, nous proposons un algorithme de partitionnement k-aire direct permettant d'améliorer le partitionnement biaisé. Finalement, nous validons cette thèse par une étude expérimentale en comparant nos méthodes aux partitionneursactuels. / Load balancing is an important step conditioning the performance of parallel programs. If the workload varies drastically during the simulation, the load must be redistributed regularly among the processors. Dynamic load balancing is a well studied subject but most studies are limited to an initially fixed number of processors. Adjusting the number of processors at runtime allows to preserve the parallel code efficiency or to keep running the simulation when the memory of the current resources is exceeded.In this thesis, we propose some methods based on graph repartitioning in order to rebalance the load while changing the number of processors. We call this problem \M x N repartitioning". These methods are split in two main steps. Firstly, we study the migration phase and we build a \good" migration matrix minimizing several metrics like the migration volume or the number of exchanged messages. Secondly, we use graph partitioning heuristics to compute a new distribution optimizing the migration according to the previous step results. Besides, we propose a direct k-way partitioning algorithm that allows us to improve our biased partitioning. Finally, an experimental study validates our algorithms against state-of-the-art partitioning tools.

La présente étude se concentre sur le travail de Nancy Fraser sur la justice sociale, lequel a suscité beaucoup d’intérêt dans la littérature au cours des dernières années. La reconnaissance et la redistribution sont les deux piliers originaux de son approche: les désavantages dont souffrent les gens dus au dénigrement culturel ou à la privation économique. Ces deux concepts servent à diagnostiquer et fournir le soutien moral aux multiples luttes que les victimes d’injustice entreprennent avec l’objectif d’établir une participation plus égalitaire à la société. Cependant, que peut-elle dire cette approche des groupes qui sont marginalisés et cherchent l’autogouvernance (ou la séparation même) plutôt que l’intégration dans la société? Le travail de Fraser manifeste une résistance envers les droits du groupe, et un silence quant à l’autodétermination. Mon intervention prend comme objectif d’inclure ces formes d’injustice dans son approche, la rendant plus sensible aux dynamiques des groupes et capable de répondre à leurs revendications trop souvent négligées sous prétexte de l’égalité. La question est, l’égalité de qui? / The focus of this study is Nancy Fraser’s work on social justice, which has gained prominence in the literature over the past few years. The two original pillars of her approach are redistribution and recognition – the injustices that individuals face as a result of economic hardship or cultural denigration. These two concepts serve to diagnose and provide moral backing to the multiple struggles that individuals undertake with the aim of a more equitable way of participating in society. But what does this approach have to say about marginalized groups who seek greater autonomy, or perhaps even separation, rather than further participation in society? Fraser’s work has manifested resistance to sanctioning group difference, and silence on the issue of self-determination. I aim to build these claims into her approach, ultimately to render it more sensible to group dynamics and more capable to respond to their demands all too often neglected under the pretext of equality. The question is, equality of whom?

La thématique de la gouvernance et du développement en vue de la concorde en République du Bénin (1960-2010) nous permet de comprendre que les échecs à l’union ou à l’unité en République du Bénin sont déjà des entorses à l’unification. Les Béninois n’ont jamais réussi à être sincères les uns vis à vis des autres. Les jeux de trahison qui ont existé entre eux depuis les temps royaux où les affres de guerres ont marqué négativement toutes les parties belligérantes ont perduré jusqu’aux temps modernes. Et, c’est l’absence de ce pardon réciproque pour une sincère réconciliation qui occasionne les crocs en jambes à la gouvernance concertée et concordante dans notre pays. Cette situation est source d’instabilité politique et sociale. Elle a développé des comportements tribalistes et ethniques qui se mutent en individualisme outré chez les concitoyens quoiqu’en soient les types de régimes politiques que nous engrangeons. Or, aucune économie ne peut être prospère si elle ne tient contre du vrai sens de l’équité. C’est ce qui occasionne le sous- développement de la République du Bénin. Les théories classiques de l’économie (« croissance d’abord, redistribution ensuite » de Lewis) échouent dans notre cas. Celles dites extrémistes (« redistribution d’abord, croissance ensuite ») échouent de même que celles relevant des modèles réformistes (« redistribution et distribution ou satisfaction des besoins humains essentiels »). Il faut donc décrire des perspectives d’amélioration de la qualité des ressources humaines, de l’économie, de la forme de gouvernance … en vue de la concorde pour l’échéance présidentielle de 2011-2016 et des suites possibles de gouvernance en République du Bénin. / The theme of governance and development for harmony in the Republic of Benin (1960-2010) allows us to understand that failures to the union or unit in Benin are already sprains to the unification. The Beninese has never managed to be sincere towards each other. The games of betrayal that existed between them since the days when the horrors of royal wars marked negatively all the warring parties have survived to modern times. And is the lack of mutual forgiveness for a sincere reconciliation which causes the tripped governance concerted and consistent in our country. This is a source of political and social instability. She developed behaviors that are tribal and ethnic mutate into excessive individualism among citizens albeit are the types of political regimes that we have achieved. However, no economy can prosper if it is against the true sense of fairness. This causes the underdevelopment of the Republic of Benin. The classical theories of economics ("growth first, then redistribution" of Lewis) fail. These so-Called extremists ("Redistribution first, then growth") fail as well as those covered by the models reformist ("distribution and redistribution or satisfaction of basic human needs ").We must therefore describe the prospects for improving the quality of human resources, economics, form of governance ... for harmony for the presidential term 2011-2016 and possible consequences of governance in the Republic of Benin.

L'Afrique de l'Ouest se caractérise par de forts taux de pauvreté et par une protection sociale et des marchés financiers formels défaillants. Dans un tel contexte, les familles remplissent des rôles importants qui façonnent la vie économique et sociale de leurs membres. Cependant, l'efficacité de cette institution fait débat, notamment lorsqu'elle est analysée du point de vue individuel. Ainsi, un nombre croissant d'études en économie du développement se penchent sur les coûts individuels induits par les fonctions familiales. Ma thèse de doctorat s'appuie sur cette littérature et y contribue en analysant la façon dont les décisions économiques individuelles sont prises au sein de la famille au Sénégal, et illustre des cas où la famille ne parvient pas à assurer le bien-être individuel de ses membres.Le premier chapitre s'attache à mesurer les coûts individuels de la redistribution informelle qui a lieu au sein des réseaux sociaux et entre ces derniers, et en particulier au sein de la famille élargie. En effet, en tant que mécanisme d'assurance, de fortes pressions redistributives modifient les choix économiques individuels. En s'appuyant sur des données expérimentales, nous estimons une taxe sociale d'environ 9% et nos résultats indiquent l'existence de fortes distorsions dans les choix d'allocation individuels. Par ailleurs nos résultats montrent que la pression à la redistribution est du fait, principalement, de famille élargie et non du ménage, des cercles amicaux ou du voisinage. Le deuxième chapitre examine comment l'environnement sanitaire contraint les investissements parentaux en termes de santé des enfants. Son objectif est de déterminer si les risques concurrents en mortalité réduisent les investissements des parents dans la santé de leurs enfants. Pour cela nous faisons, dans cet article, l'hypothèse qu'il existe des complémentarités entre les investissements en santé spécifiques à certaines maladies ; et testons celle-ci en nous appuyant sur une intervention récente visant à éradiquer le paludisme au Sénégal. Nos résultats soutiennent ce modèle de risques concurrents et impliquent que les comportements privés d'investissement en santé des parents et les politiques publiques de santé sont complémentaires. Enfin, le troisième chapitre explore comment une stratégie de gestion des risques ex-post, le mariage d'enfants, est lié aux chocs de mortalité au sein des familles. En particulier, j'examine si les décès paternels ont des conséquences néfastes en termes de transition sur le marché du mariage des jeunes orphelines. Mes résultats mettent en évidence la grande vulnérabilité de ce groupe d'enfants en termes de mariages précoces. / In the West Africa subregion, poverty is pervasive and social protection at the state-level, as well as formal financial markets poorly function. In such a context, families fulfill important roles, which shape the economic and social life of its members and are key drivers of economic development. How this private institution performs is, however, quite a debate and a growing body of the literature in development economics has focused on the question of the economic inefficiencies of the family institution. My PhD thesis builds upon this literature and contributes to it by shedding light on how individual economic decisions are taken within the family in Senegal, and in which cases the family fails to ensure individual welfare of its members.The first chapter focuses on the individual costs of the informal redistribution that take place within and between social networks, and in particular within the extended family. Using a lab-in the field experiment, we aim at identifying the hidden costs of social obligations for redistribution on individual resource allocation choices. Our results are threefold: (i) we estimate a social tax of about 9\%; (ii) we provide evidence on strong distortions in individual allocation choices; (iii) our results point at people fearing redistributive pressure from the extended family members, but not from within the household or from friends and neighbors. We expand on prior literature by both identifying the individual cost of informal redistribution and then relating it to postexperiment resource-allocation decisions, and by disentangling intra- and interhousehold redistributive pressure. The second chapter investigates how the health environment prevents parents from investing in child health. Its main objective is to investigate whether the health risks faced by children, and in particular their competing nature as mortality risks, depress parental investment in child health. We argue that there are complementarities between disease-specific investments and we test this hypothesis by exploiting recent interventions that made anti-malaria products suddenly affordable to most households in 2009 in Senegal. Our first contribution is to be the first to use data on private health expenditures to validate a model with complementarities between disease-specific investments. Our second contribution is to show that parental behavioral responses clearly complement anti-malaria campaigns, whereby they magnify their impact on all-cause mortality for children. Finally, the third chapter explores how a quite harmful ex-post risk management strategy, child marriage, relates to changes in family structures (mortality shocks). In particular, I investigate whether paternal death induces adverse marriage outcomes for young orphans. I also discuss the channel that could induce such effects. My results underpinned the high vulnerability of this group of children: when the father dies, the young girls are more likely to marry as child brides and to be child mothers than their non-orphan counterparts. Those girls have more frequently broken marital trajectories, in particular due to divorce. This paper builds upon the existing demographic literature and provides at least two contributions. First, it is, to my knowledge, the first to study jointly the timing of the father's death and other dimensions of well-being such as fertility, marital dissolution and consumption. Second, it discusses the extent to which selection on unobservables might bias the analysis, an issue that is discarded in most studies.

The topic of this thesis is institutional equilibrium, its formation and disruption in human systems. The thesis analyses what is common to the most human systems and concludes that it is some kind of redistribution. That is reason for the thesis title "Institutional equilibrium in redistribution systems". As the term institutional equilibrium comes from institutional economics, the thesis begins with characteristic institutional economy and its differences form mainstream economy, defines the term institution and methodological institutionalism and explains the reasons of institution evolution. Then, the problematic of redistribution is analyzed. The thesis brings a general model of a redistribution system and describes a coalition forming in it. The thesis concludes that such a process must be facilitated and regulated by institutions and that institutions define the negotiation strength of the players and stabilize their position. In a fourth chapter, the term of institutional equilibrium is defined and its preconditions and consequences are described. Factors preventing the redistribution system from reaching the equilibrium are shown. The analysis of institutional equilibrium formation follows; the open and closed systems are addressed separately, with the emphasis being put on open systems. In open systems, the creative destruction process in ever-present, disrupting the institutional equilibrium. Therefore, a question is posed: may there be such a structure of institutions which will stay unchanged despite the ongoing exogenous changes? Or, more precisely, the change of which will not be in the interest of the players? The thesis finds that the preconditions for institutions system facilitating the institutional equilibrium are investments into skills of the players and the open access into the system. The thesis also shows that such preconditions help to reduce undesirable redistribution. The fifth chapter analyses parallel redistribution games as one of the specific way of disturbing institutional equilibrium. The parallel redistribution games have secret character; their players do not want the other players to know about their game, because their game breaches the norms of the majority of players. The parallel redistribution games may be characterized as a form of opportunistic behavior. The thesis investigates the causes of these games, the reasons why they remain uncovered and the strategies of their players. Eventually, the conditions under which the parallel games may be reduced are formulated. The institutions must fulfill these conditions as well. In the end, the institutions structure capable of introducing the redistribution system into the institutional equilibrium state is defined.

This short dissertation briefly examines the English law statutory provisions and the English judicial approach relating to the clean break principle. A comparison between English law and South African law (in particular s 7 of the Divorce Act 70 of 1979) shows the latter to allow the application of the clean break principle, but rigidity in the legislative wording and further rigidity in the wording by the courts result in the clean break principle not being a dominant feature of South African divorce law. A detailed analysis of South African case law follows in which the fundamental issues relative to the clean break principle are examined, namely the one third rule, universal partnership, misconduct and its effect on court orders, the nature of a contribution and what property is distributed on divorce. Trends in maintenance orders are looked at and how these affect the clean break principle, in particular rehabilitative maintenance and token maintenance wards. / Law / LL.M.