A TRANSPARENT FAULT TOLERANT STORAGE MIGRATION ON CLOUDS

Cloud computing is fast catching up as a computing paradigm of the coming decade. Many major players has entered the arena and commercial cloud will soon become a household name. Due to the high scalability of the cloud the chances for failures is very high in the cloud data centres. My work is to design the cloud storage which guarantees data availability in spite of network failure. The binomial graph network topology and the self-healing procedure chooses whether to migrate, replicate or remove the data based on storage availability. The performance evaluation proves that the approach dynamically finds an optimal resource allocation for the data availability. The experiments shows that the network is reconfigured and the data is made available in linear time and experiments are conducted to measure the fault tolerant capability of the system.