Article Preview
TopThere are many challenges associated with computational grid, like, scheduling of jobs in appropriate resources for subsequent execution, increasing the number of deadline meets for submitted jobs, balancing of workload in grid and fault tolerance. Important related works in this arena of research are summarized in this section.
The load balancing process must take into consideration the dynamic loads of participating resources of the grid (Abo Rizka & Rekaby, 2012). Job scheduling has to ensure higher number of deadlines meets, thereby improving the performance of the grid. In (Goswami & Das, 2016), an adaptive execution scheme has been proposed which ensures guaranteed performance with respect to service quality agreements.
The properties of general distributed system is quite different from that of computational grid environment. Moreover, the client-server framework of grid proposed by Michael Stal (Stal, 1995) does not solve the problem of load balancing among participating resources in grid. Various job scheduling schemes in grid have been described in (Kant Soni, Sharma & Kumar Mishra, 2010). X-Dimension binary tree data model (Abo Rizka & Rekaby, 2012) have limited success in workload balancing among grid resources.