A Hybrid Batch Mode Fault Tolerance Strategy in Desktop Grids

2. Review Of Literature

Desktop Grid resources are volatile, dynamic and diverse in nature. Faults are expected to occur due to the unpredictable environment of grid (Zheng & Veeravalli 2009; Lee et al., 2009; Sathya & Babu, 2010). Various fault handling strategies are designed to handle failures without affecting the quality of service (QoS). Fault handling strategies can be pro-active or post-active based upon whether faults are handled before or after job execution

Proactive fault tolerance strategies handle the failures before the execution of the task and scheduled with the expectation that task will never fail. Proactive strategies need information concerning i.e. grid tasks, resources, fault types, fault occurrence etc. for effective scheduling. Based on the available information, the decisions to handle the probable failures are taken.