Catalogue Search | MBRL
Are you sure you want to remove the book from the shelf?
{{itemTitle}}
4,299
result(s) for
"Computational grids"
Sort by:
Grid computing : techniques and applications
Based on the author's own course, this textbook is designed for a first course on grid computing at the senior undergraduate and first- year graduate levels. Unlike other texts on the subject, this book includes a lecture structure and student programming assignments.
Book
Efficient allocation of independent gridlet on small, medium, and large grid
Gridlet allocation in a computational grid environment is a major research issue to obtain not only the efficient gridlet allocation technique but also the time needed to obtain the efficient allocation technique. Grid computing networks have various nodes to process user jobs. To achieve the high performance of computational grid, task scheduling is an important issue. The users who are using services of grid systems are more cautious about time to complete their job. Hence, this work concentrates on gridlet allocation method used for time reduction by a genetic algorithm with the MapReduce programming model for independent tasks in computational grid. In computational grid environment, multi-objective problem formulation minimization of makespan and flowtime is considered. In this proposed technique, fitness function formulation for makespan and flowtime has been formulated mathematically. The genetic algorithm with the MapReduce programming model is implemented using MapReduce written in Java and then combined with GridSim. The experimental outcome with regard to time needed, flowtime, and makespan clearly reveals that the genetic algorithm with the MapReduce model effectively optimizes time, makespan, and flowtime in computational grid environment. A comparative study of performance efficiency among genetic algorithm with the MapReduce and sequential genetic algorithm (SGA) and parallel genetic algorithm (PGA) depicts the usefulness of the model. The execution time achieved by GA with the MapReduce model in small grid is 10.48 s, medium grid is 18.76 s, and large grid is 33.73 s.
Journal Article
A Load Distribution Based Resource Allocation Strategy for Bag of Tasks (BoT) in Computational Grid Environment
In the ever-evolving landscape of computational grid systems, the meticulous selection of resources tailored to specific tasks is a formidable challenge. This paper introduces an efficient load distribution strategy known as Load Distribution Based Resource Allocation (LDRA), one of the foremost goals is to allocate resources to gain enhanced resource utilization and also try to achieve least possible execution time to fulfil the need for grid systems. A comprehensive performance evaluation unfolds to elevate grid efficiency, pitting LDRA against existing heuristics using the ETC Simulation Benchmark. The study expands further on the real-world dataset from the Gaia Cluster Configurations (
https://hpc.uni.lu/systems/gaia/
) to verify its significance in the real environment. The LDRA algorithm emerges with superior performance when compared to state-of-the-art such as Max–Min, Opportunistic Load Balancing, AlgHybrid_LB, and Resource Aware Load Balancing for resource utilization, makespan, flowtime, and energy efficiency in the majority of the cases in experimental evaluation. In some cases, the experimental results show that LDRA’s usage of the grid resources is remarkable, reaching over 99% in four cases and approaching 98% in two cases of the ETC simulation benchmark. These accomplishments are further mirrored in the evaluation against real datasets, where LDRA’s performance among peers is nothing short of exemplary in the cases under study.
Journal Article
High-performance parallel database processing and grid databases
The latest techniques and principles of parallel and grid database processing The growth in grid databases, coupled with the utility of parallel query processing, presents an important opportunity to understand and utilize high-performance parallel database processing within a major database management system (DBMS). This important new book provides readers with a fundamental understanding of parallelism in data-intensive applications, and demonstrates how to develop faster capabilities to support them. It presents a balanced treatment of the theoretical and practical aspects of high-performance databases to demonstrate how parallel query is executed in a DBMS, including concepts, algorithms, analytical models, and grid transactions. High-Performance Parallel Database Processing and Grid Databases serves as a valuable resource for researchers working in parallel databases and for practitioners interested in building a high-performance database. It is also a much-needed, self-contained textbook for database courses at the advanced undergraduate and graduate levels.
eBook
Numerical Solution of the One-Dimensional Forward Magnetotelluric Sounding Problem Using a Computational Grid Adaptation Approach
The paper considers an implementation of an adaptive computational grid constructing algorithm in a numerical solution of the one-dimensional forward magnetotelluric sounding problem (the Tikhonov–Cagniard problem). The numerical solution of the problem is realized by a method of local integral equations which was proposed by the authors previously. The adaptive computational grid construction is based on geometrical principles of optimizing a piecewise constant interpolant of the electrical conductivity function to be approximated. Numerical experiments are carried out to study and illustrate the effectiveness of the combined method. The algorithm is tested on the Kato–Kikuchi model with a known exact solution.
Journal Article
A reliable, TOPSIS-based multi-criteria, and hierarchical load balancing method for computational grid
Load balancing is a very important and complex problem in computational grids. In load balancing, jobs should be effectively distributed among resources in order to minimize the average completion time and maximize the utilization of all resources even those with low reliabilities and capacities. However, using the less reliable and slow resources implies worse completion time, whereas always selecting the powerful and reliable resources undermines the utilization of other resources. So, it is essential to develop an efficient load balancing method which makes a good tradeoff between these criteria in a way that satisfies the quality of service of jobs and fairly distributes jobs between resources based on their reliabilities and capacities. This paper proposes an efficient multicriteria load balancing method using technique for order preference by similarity to ideal solution which treats load balancing as a multi criteria decision making problem. Also, an effective weighting mechanism is proposed, which adaptively adjusts the weights of the considered criteria according to the system’s current state and jobs’ characteristics. This mechanism can make an efficient tradeoff between the considered criteria and accurately reflect the importance of each one. By simulation, the proposed method was evaluated and compared with other approaches from the literature. In the range of examined parameters’ values, the simulation results show that proposed method minimizes the average completion time by 8.7–15.7%, increases the throughput ratio up to 15.8–19.4%, and maximizes the load balancing level by 7.68–20.1%.
Journal Article