Fault Tolerant Techniques in Mobile Grid Computing: A Survey

Computational grid systems are the systems, where a task is distributed in various geographically distributed nodes. An important property of the grid system is Fault tolerance; it is a property of the system which enables a system to operate properly in case of failures. Checkpointing and replication are the two commonly used techniques for fault tolerance. Furthermore, these techniques mitigate the lost work, they commonly migrates a snapshot of the latest correct state of the failed node or share information between the resources to ensure consistency between them. However, these techniques are not full proof as they may introduce some kind of runtime overheads. Increasing fault tolerance means making your system more reliable and available to the users. It is also essential to satisfy the QoS requirement of the computational grids. Reliability and availability are the two greatest challenges in grid due to the unreliable nature of its resources.