A Dynamic Approach For Admission Control Policy And Replica Placement and Replacement Algorithem In DRTDBS

Abstract

Many real time applications need data services in distributed environment. newlineRecent distributed real time system (DRTS) are intended to manage large volume of newlinedispersed data. However providing such data services are a challenging task due to newlinelong remote data accessing delay and stringent time requirement of real time newlinetransaction. To make distributed real time data processing an efficient data newlinemanagement system algorithm and protocol must be required for accessing and newlinemanipulating the data. newlineThe presence of multiple sites in the distributed environment raises issue that is newlinenot present in centralized system. Therefore a new research in distributed real time newlinedatabase system (DRTDBS) is needed to investigate the possible way of applying newlinedistributed database system technology to real time system. newlineThe performance of DRTDBS depends on the several design factors. The newlinedesign issues of DRTDBS are predictability, consistency, data access mechanism and newlineinvariance, admission control policy, scheduling of transaction with deadline, time newlinecognizant, buffer and lock management etc. The resources that can be schedule are newlineprocessor, main memory, disk and data items stored in database. Other performance newlinegradients are new matrices of data correctness and performance maintaining global newlinesystem information, security, fault tolerance, failure and recovery, memory newlineoptimization and possibility of distributed system dead lock etc. newlineIn order to resolve the contention for these resources DRTDBS have to establish newlinedata replication policy which is widely used for availability, scalability, maximum data newlineread capacity and timing constraints operation etc. A large volume of data file is newlineproduced so we need a dynamic replication approach which we can reduce the data newlineaccess time. In order to replicate it at the nearest site replica placement and newlinereplacement algorithm are required to be evaluated. The failure of one database replica newlinedoes not stop the system from working as available other database replicas can take newlineover the tasks of the failed replica. He