Data-Aware Scheduling in Datacenters

Datacenters have emerged as the dominant form of computing infrastructure over the last two decades. The tremendous increase in the requirements of data analysis has led to a proportional increase in power consumption and datacenters are now one of the fastest growing electricity consumers in the United States. Another rising concern is the loss of throughput due to network congestion. Scheduling models that do not explicitly account for data placement may lead to a transfer of large amounts of data over the network causing unacceptable delays. In this dissertation, we study different scheduling models that are inspired by the dual objectives of minimizing energy costs and network congestion in a datacenter. As datacenters are equipped to handle peak workloads, the average server utilization in most datacenters is very low. As a result, one can achieve huge energy savings by selectively shutting down machines when demand is low. In this dissertation, we introduce the network-aware machine activation problem to find a schedule that simultaneously minimizes the number of machines necessary and the congestion incurred in the network. Our model significantly generalizes well-studied combinatorial optimization problems such as hard-capacitated hypergraph covering and is thus strongly NP-hard. As a result, we focus on finding good approximation algorithms. Data-parallel computation frameworks such as MapReduce have popularized the design of applications that require a large amount of communication between different machines. Efficient scheduling of these communication demands is essential to guarantee efficient execution of the different applications. In the second part of the thesis, we study the approximability of the co-flow scheduling problem that has been recently introduced to capture these application-level demands. Finally, we also study the question, "In what order should one process jobs?'' Often, precedence constraints specify a partial order over the set of jobs and the objective is to find suitable schedules that satisfy the partial order. However, in the presence of hard deadline constraints, it may be impossible to find a schedule that satisfies all precedence constraints. In this thesis we formalize different variants of job scheduling with soft precedence constraints and conduct the first systematic study of these problems.

TAXONOMY, MORPHOLOGY, AND RNA-SEQ TRANSCRIPTOMICS OF THE CUBOZOAN ALATINA ALATA, AN EMERGING MODEL CNIDARIAN

Cnidarians are often considered simple animals, but the more than 13,000 estimated species (e.g., corals, hydroids and jellyfish) of the early diverging phylum exhibit a broad diversity of forms, functions and behaviors, some of which are demonstrably complex. In particular, cubozoans (box jellyfish) are cnidarians that have evolved a number of distinguishing features. Some cubozoan species possess complex mating behaviors or particularly potent stings, and all possess well-developed light sensation involving image-forming eyes. Like all cnidarians, cubozoans have specialized subcellular structures called nematocysts that are used in prey capture and defense. The objective of this study is to contribute to the development of the box jellyfish Alatina alata as a model cnidarian. This cubozoan species offers numerous advantages for investigating morphological and molecular traits underlying complex processes and coordinated behavior in free-living medusozoans (i.e., jellyfish), and more broadly throughout Metazoa. First, I provide an overview of Cnidaria with an emphasis on the current understanding of genes and proteins implicated in complex biological processes in a few select cnidarians. Second, to further develop resources for A. alata, I provide a formal redescription of this cubozoan and establish a neotype specimen voucher, which serve to stabilize the taxonomy of the species. Third, I generate the first functionally annotated transcriptome of adult and larval A. alata tissue and apply preliminary differential expression analyses to identify candidate genes implicated broadly in biological processes related to prey capture and defense, vision and the phototransduction pathway and sexual reproduction and gametogenesis. Fourth, to better understand venom diversity and mechanisms controlling venom synthesis in A. alata, I use bioinformatics to investigate gene candidates with dual roles in venom and digestion, and review the biology of prey capture and digestion in cubozoans. The morphological and molecular resources presented herein contribute to understanding the evolution of cubozoan characteristics and serve to facilitate further research on this emerging cubozoan model.

BLACK, MALE, AND HIGH-ACHIEVING: AN EXAMINATION OF A RISK FACTOR AND CULTURAL RESOURCES FOR BLACK MALE ACADEMIC ACHIEVEMENT

Black boys are confronted with unique educational circumstances. They are often misdiagnosed and misplaced into special education programs (Bush-Daniels, 2008; Patton, 1998; Terman et al., 1996). Additionally, they are less likely to be enrolled in gifted and talented programs, even if their former achievements reflect their aptitude to succeed (Black Alliance for Educational Options, n.d.; Moore & Flowers, 2012). Given these statistics, a considerable emphasis has been placed on the causes and the consequences of low/under achievement for this population. As a result, the experiences of Black males who are achieving have been greatly neglected. Moreover, little is known about the factors that facilitate academic achievement among high-achieving Black boys. In an effort to bring the heterogenic nature of schooling experiences for Black boys to light, the present study examined the influence risk and protective factors had on the academic experiences of high-achieving Black boys. Grounded in the risk and resilience framework and the Integrative Model for the Study of Minority Youth Development, this study explored whether the high-achieving Black high school boys in this sample (n =88) reported experiencing discrimination (i.e. academic-based) and how this academic-based discrimination related to their 1) academic performance (i.e. GPA), 2) perceptions of math ability, and 3) race-based academic self-concept. In addition to exploring how academic-based discrimination was linked to academic achievement, this study examined how cultural resources such as racial socialization messages and racial identity related to academic achievement. Specifically, cultural socialization, preparation for bias, egalitarianism, private regard and public regard were evaluated alongside the three academic outcomes under study. Finally, the study explored whether aspects of racial socialization or racial identity buffered the effects of discrimination on any of the outcomes. Interestingly, the race/ethnicity of the student mattered for how students perceived their math ability. The risk factor academic-based discrimination was linked to academic performance. Cultural resources cultural socialization, preparation for bias, and private regard were linked to various academic outcomes of interest. There was only one significant moderating effect: a high private regard buffered the relationship between academic-based discrimination and race-based academic self-concept. Limitations and implications of this study are discussed.

TRAFFIC IMPACT ANALYSIS OF SEVERAL DYNAMIC LANE MANAGEMENT STRATEGIES FOR CONGESTION MITIGATION BASED ON DTA MODEL

Persistent daily congestion has been increasing in recent years, particularly along major corridors during selected periods in the mornings and evenings. On certain segments, these roadways are often at or near capacity. However, a conventional Predefined control strategy did not fit the demands that changed over time, making it necessary to implement the various dynamical lane management strategies discussed in this thesis. Those strategies include hard shoulder running, reversible HOV lanes, dynamic tolls and variable speed limit. A mesoscopic agent-based DTA model is used to simulate different strategies and scenarios. From the analyses, all strategies aim to mitigate congestion in terms of the average speed and average density. The largest improvement can be found in hard shoulder running and reversible HOV lanes while the other two provide more stable traffic. In terms of average speed and travel time, hard shoulder running is the most congested strategy for I-270 to help relieve the traffic pressure.