Recent enthusiasm in grid computing has resulted in a tremendous amount of research in resource scheduling techniques for tasks in a workflow. Most of the work on resource scheduling is aimed at minimizing the total response time for the entire workflow and treats the estimated response time of a task running on a local resource as a constant. However in a dynamic environment such grid computing, the behavior of resources simply cannot be ensured. In this thesis, thus, we propose a probabilistic framework for resource scheduling in a grid environment that views the task response time as a probability distribution to take into consideration the uncertain factors. The goal is to dynamically assign resources to tasks so as to maximize the probability of completing the entire workflow within a desired total response time. We propose three algorithms for the dynamic resource scheduling in grid environment, namely the integer programming, the max-max heuristic and the min-max heuristic. Experimental results using synthetic data derived from a real protein annotation workflow application demonstrate that the proposed probability-based scheduling strategies have similar performance in an environment with homogeneous resources and perform better in an environment with heterogeneous resources, when compared with the existing methods that consider the response time as a constant. Of the two proposed heuristics, min-max generally yields better performance.

Recent enthusiasm in grid computing has resulted in a tremendous amount of research in resource scheduling techniques for tasks in a (scientific) workflow. There are many factors that may affect the scheduling results, one of which is whether the application is computing-intensive or data-intensive. Most of the grid scheduling researches focus on a single aspect of the environments. In this thesis, we base on our previous work, a probability-based framework for dynamic resource scheduling, and consider data transmission overhead in our scheduling algorithms. The goal is to dynamically assign resources to tasks so as to maximize the probability of completing the entire workflow within a desired total response time. We propose two algorithms for the dynamic resource scheduling in grid environment, namely largest deadline completion probability (LDCP) and smallest deadline completion probability (SDCP). Furthermore, considering the data transmission overhead, we propose a suite of push-based scheduling algorithms, which schedule all the immediate descendant tasks when a task is completed. These are algorithms will be compared to the pull-demand scheduling algorithms in our previous work and workflow-based algorithms proposed by other researchers. We use GridSim toolkit to model the grid environment and evaluate the performance of the various scheduling algorithms.

published_or_final_version / Applied Statistics / Master / Master of Social Sciences

published_or_final_version / abstract / toc / Statistics and Actuarial Science / Master / Master of Philosophy

No description available.

No description available.

In this paper I have separated three periods of probability theory teaching: 1. Probability theory teaching in Lithuania during 1920-1940’s. 2. Probability theory teaching in Lithuania during 1960-1990’s. 3. Probability theory teaching in Lithuania during 1991-2000’s. In the prewar Lithuania probability theory was a separate science already, examined in university, applied in other disciplines. In secondary schools subject was not taught yet. At that moment first textbook of probability theory was published: V. Biržiška, Fundamentals of mathematical probability theory: lectures in Lithuanian university during 1928 – 1929’s in Kaunas, 1930, p. 588. This science was very interesting for the mathematicians. Probability theory gained importance in the everyday life, too. At the end of 1940’s elements of probability theory were included into mathematics programs in the further education schools. During 1960-1990’s probability theory was studied in the Lithuanian schools as an optional subject. Course was rather wide and complicated. At that period many additional training aids appeared, scholars carried out research and experiments, searched for optimal methods of teaching probability theory. Science as such was interesting not only for the scientists, methodologists, but for the psychologists, too, who examined children, trying to decide, what age is the best to start learning probability theory. At that moment Vytautas Liutikas published his book „How probabilities of events... [to full text]

No description available.

No description available.

No description available.