DefinePK

Abstract

Computational Data Grid provides massive resource sharing and aggregated computing resources in a dynamic manner. Due to the limitation of available heterogeneous resources distributed through several networks in computational data grid, occurrence of failure poses severe problem. Providing efficient fault tolerance mechanism is a key optimization technique for improving scalability and attain QoS based fault tolerant dynamic replication in a wise manner. In this paper, Merkle Damgard Clock Synchronized based Fault Tolerant (MDCS-FT) mechanism is developed to overcome the fault occurrence in computational data grid. MDCS-FT in cloud services allows for concurrent transaction without relying on a centralized grid component, which amounts for better scalability. Clock Synchronized with Berkeley algorithm uses the optimized different data grid sequences to attain QoS based fault tolerant dynamic replication. Berkeley algorithm in MDCS-FT mechanism is more suitable for easily identifying the fault with time server. Time server in MDCS-FT Mechanism periodically fetches the time from all the clients and averages the results on cloud zone to secure data objects replication in cloud data grid by removing the fault. Experimental results demonstrate that the proposed mechanism achieves better performance by improving the scalability and QoS (error rate and transmission delay) and minimizes the fault tolerance compared to the state-of-the-art works.