Design and implementation of a synchronous and asynchronous-based data replication technique in cloud computing

Abstract EN

Failures are usual rather exceptional in cloud computing environment.

To access from the nearby site, the often used data should get replicated to multiple locations to compose the users to advance the system accessibility.

A challenging task in cloud computing is to decide a sensible number and right location of replicas.

At this point, we propose an adapted dynamic data replication approach to decide a sensible number and right location of replicas and we compare both the adapted dynamic data replication approach and normal dynamic data replication approach.

The normal dynamic data replication approach has three dissimilar stages which are the recognition of data file to replicate, number of replicas to be created and placing new replicas.

We adapt the popularity degree in the initial stage of normal dynamic data replication approach and also we consider the failure probability for replica factor calculation and the other two stages are related to the normal dynamic data replication approach.

When we update the major data center we moreover integrate the synchronous and asynchronous updation of replice data file.