In a distributed data processing system, a database can be centralized or distributed. What are the options? Explain.

What will be an ideal response?

In a distributed data processing system, a database can be centralized or distributed. When the database is centralized, the entire database is stored at a central site which processes requests from users at remote locations. Certain concerns arise when data processing is distributed. Questions arise with regard to data currency when multiple users have access to the database. Database lockout prevents more than one user from making changes at the same time.

Distributed databases can be partitioned with parts stored at different sites, or replicated, with the entire database stored in multiple locations. When the database is partitioned, users have more control over data stored at local sites, transaction processing time is improved, and the potential of data loss is reduced. When the database is replicated, the entire database is stored at multiple locations. This works well when the primary use of the database is for querying. When transactions are processed at many sites, problems of database concurrency arise.

PTS: 1