Distributed databases and data distribution: What is the term for a strategy in which duplicate copies of the entire database are stored at all (or multiple) locations so that each site has a full, synchronized copy?

Introduction / Context:
Organizations with multiple offices often need the same data close to each user to improve availability and performance. Distributed database design offers several data placement strategies. This question asks you to identify the strategy in which a complete copy of the database exists at each site, rather than splitting it into pieces.

Given Data / Assumptions:

• Multiple locations (sites or nodes) must access the same data.
• The goal is for each location to have a full copy of the database.
• Copies are intended to be synchronized so users see consistent information.

Concept / Approach:
Replication in distributed databases means maintaining one or more duplicate copies of data across different nodes. In full replication, every site keeps the entire database; in partial replication, only selected tables or fragments are copied. Replication improves read performance and availability, but it introduces complexity for write coordination and conflict resolution. This differs from partitioning (also called fragmentation), where data is split (horizontally or vertically) and each site stores only a portion.

Step-by-Step Solution:

Verification / Alternative check:
Distributed systems texts classify data placement strategies as replication vs partitioning (or hybrids). Only replication yields full duplicates at all sites.

Why Other Options Are Wrong:

• Partitioning: Splits data rather than duplicates it fully.
• Polling: A technique to query or check devices, not a data placement strategy.
• Data sharing: Vague; does not imply full duplicates.
• None: Incorrect because “replication” fits exactly.

Common Pitfalls:
Confusing replication with caching or partitioning; ignoring write-conflict and latency issues that replication introduces.

Final Answer:
replication