Comparative Analysis of Some Grid Scheduling Algorithms on Real Computational Grid Environment

The development of computers as well as network technologies en compass us towards the world of high performance computing termed as Grid computing, w hich is the evolution of parallel processing. Grid scheduling is a vital component of compu tational grid environment. As regard to that, grid scheduling tend to be one of the major ch allenges of grid computing, yet an NP-complete problem. Moreover, some researchers have considered slack ti me using different techniques, but not yet considered operational research slack time c oncept by combining hard real time system and soft real time system techniques which will m ax mize the number of deadlines met, minimize tardiness of tasks, maximize the number of hig h priority tasks in mee ting their deadlines and creating fairness. However, we propose two hybrid scheduling alg orithms (Least Slack Time Rate First Round Robin Based Scheduling Algorithm (LSTRRGCD), Earliest Deadline First Round Robin Based Scheduling Algorithm (EDFRRGCD)), ba sed on deadline, slack time using a dynamic time quantum, computed a s the greatest common divisor (GCD) of all jobs burst time rate and also evaluate with First Come First Serve d Scheduling Algorithm (FCFS) and Round Robin Scheduling Algorithm (RR)) as base line approaches in a real grid environment using real workload traces, taken from leading compu tational centers.