JAWAHARLAL NEHRU TECHNOLOGICAL UNIVERSITY HYDERABAD B.Tech IV Year DISTRIBUTED DATABASES Previous Question Paper (July/August - 2022 - R18)

A Distributed Database System (DDBS) is a collection of multiple, logically interrelated databases distributed over a computer network. The system is managed by a Distributed Database Management System (DDBMS), which makes the distribution transparent to the user, ensuring that the system appears as a single, centralized database.

The three dimensions that define data delivery alternatives in a distributed database are:

• Distribution Dimension: This refers to whether the data is centralized at a single site or partitioned/replicated across multiple sites. It determines the physical location of data entities.
• Fragmentation Dimension: This dimension defines whether the data is stored in its entirety at a site or divided into smaller logical units (fragments). Fragmentation can be horizontal (rows), vertical (columns), or mixed, which helps in optimizing local access.
• Replication Dimension: This dimension specifies whether a fragment is stored at only one site (non-replicated) or copied across multiple sites (replicated). Replication improves availability and query performance but introduces significant complexity in maintaining consistency.

The architecture of a Distributed DBMS consists of several functional components that work together to manage distributed data. The primary components are:

• User Processor (Front-End): This component interacts with the user, parses the query, and performs semantic analysis. It includes the Query Processor, which decomposes the global query into sub-queries for different sites.
• Data Processor (Back-End): This component resides at each site and is responsible for local data management. It handles Local Query Optimization, Transaction Management, and Concurrency Control.
• Communication Network: The underlying infrastructure that facilitates data exchange and message passing between the user processor and data processors.
• Global Catalog: A metadata repository that stores information about the distribution of data, fragmentation schemas, and replication details, allowing the DDBMS to locate data across the network.

The architectural flow can be visualized as:

The User Processor ensures location transparency, while the Data Processor ensures that local integrity and transaction atomicity are maintained.