Development of Problem Solving Environment For Handling Middleware Faults in Grid

Abstract

Grid enable the sharing, selection and aggregation of a wide variety of resources including supercomputers, storage systems, data sources and specialized devices that are geographi- cally distributed and owned by di erent organizations for solving large-scale computational and data intensive problems in science, engineering and commerce.However, the major challenge in such highly heterogeneous and complex computing environment is to design an e cient problem solving environment.The large number of components and their distribu- tion in grid as well as the dynamic structure of the grid increases the risk of failures.Thus providing a scalable and generic problem solving environment as a basic service for grids, is of primary importance.This thesis aims at furnishing fault tolerant grid system by iden- tifying and handling middleware faults.Grid middleware is core component of grid system responsible for overall progress of grid application.In the research work carried out during this thesis, middleware undertaken for experimentation is alchemi.Alchemi is a .NET-based grid computing framework that provides the runtime machinery and programming envi- ronment required to construct computational grid.Alchemi supports scheduling algorithm which selects highest priority thread among the pool of threads and dedicate to it to the available executor keeping scope for enhancement of application performance.This thesis proposes customized scheduler algorithm for handling faults so that scheduling of jobs will be based on available CPU power.Also a new manager proposed, recovers the jobs from its failure, providing problem solving environment for grid application resolving faults and its impacts.