Publication

Abstract : Fault tolerance is one of the major challenges during fault occurrence in any computational environment. Since, computational grids provide access to huge pool of shared processing power, the chances of fault occurrence are high. The proposed system is focused on performing job scheduling using Two Choices Scheduling Algorithm (TCSA) with tolerating the faults that occur during the process. Job allocation is done with the help of grid system consisting of resource broker and GIS. Once, the faulty resources are identified, scheduler analyses the load of remaining idle resources and reschedules the faulty jobs to the new set of resources which currently has the least load. A comparison is done between Round Robin algorithm and TCSA before and after incorporating fault tolerance. It is observed that the time taken for the completion of allocated tasks has been minimized after the implementation of improvised TCSA incorporating fault tolerance. The work focuses on tolerating the faults that are occurred during task allocation and performs rescheduling by reallocating the tasks to different resources which helped the system to obtain significant results while the system is exposed to high loads. © 2005 – ongoing JATIT & LLS.