A Survey of Distributed Fault Tolerance Strategies

Abstract: Grid computing is defined as geographically distributed, heterogeneity (different hardware, software and networks), resource sharing, multiple administrators, dependable access, and Pervasive access within dynamic organizations. In grid computing, the rate of failure is much greater than in traditional parallel computing. Therefore, the fault tolerance is an important property in order to achieve reliability, availability and QOS. In this paper, we give a survey on various fault tolerance techniques and fault management in different situations with related issues. The fault tolerance service deals with various types of resource failures. This survey provides the related research results about fault tolerance in grid infrastructure and also the future directions about fault tolerance techniques, and this survey attempts to provide guide for researcher.

Keywords: Fault Masking, Failure, Replication, Reliability, Availability.