Recovery of Data in Cluster Computing By Using Fault Tolerant Mechanisms
Iosr Journals, Sridevi Balne · Figshare · 2015
Many applications executing on cluster architectures consisting of several number of computers create several problems with reliability. The problems are system crashes, code failures and network problems etc. This paper using Fault tolerant mechanisms to recover the data in cluster computing. The two Rollback Recovery protocols(5) are Theft Induced check pointing protocol (TIC) and systematic event logging protocol TIC protocol is used to recover the information during crash, the information is send to client (receiver) by using SEL protocol. It is proven that the complexity of protocol is very small and the information lost by crashed process is recovered here. Section 1: describes the main objectives of TIC&SEL protocols Section 2: describes the background for grid or cluster computing Section 3: describes the definition of checkpoints & different types of checkpoints Section 4: gives the details about work stealing or task stealing of different tasks Section 5: describes an algorithm, which is mainly used for retrieving information from database server and creates backup of db Section 6: gives the experimental results Section 7: gives concluded the different techniques