Abstrait

A SCHEDULING IN HETEROGENEOUS COMPUTING ENVIRONMENT USING ALEA SIMULATOR BASED ON GRIDSIM

Satheesh Kumar N, Ilayaraja AP and PNV Pavan

Grid computing is well established and is used for large scale simulations. The resources available on the grid vary from supercomputers to clusters and single PC’s. In this project we are using simulator of grid (GridSim) that simulates computers and installed ALEA scheduler on the top of the GridSim. This work will concentrate on the representation of parallel jobs and their execution. Our goal lies in the comparison of various scheduling approaches and evaluating the best performance algorithm for the grid environment. Grid Simulators are often used for evaluation of new scheduling methods. The job is composed of large number of independent tasks executed on separate processing units. The objective of the project is to evaluate the performance of simplest scheduling algorithms based on various parameters. Taking into consideration the resource and the job specifications, this project also calculates the total usage of the resources. Grids build on the concept of distributed computing, where there is no longer a centralized computer establishing a hierarchy. Instead, software provides a way to divide up tasks so they are processed in parallel. That makes it possible to use a range of inexpensive computers from single or different manufacturers to address a complex computational task, recreating the power of a supercomputer with off-the-shelf PCs.