🧬Data Replication

NEW

Copying and maintaining data across systems or regions to improve availability, fault tolerance, and read scalability.

DESCRIPTION

Data replication maintains one or more copies of data across nodes, zones, or regions. It improves availability and read performance, and supports failover for disaster recovery. Modes include synchronous (strong consistency) and asynchronous (lower latency with potential lag). Replication can be storage-level, database-native, or application-driven. Conflict detection and resolution are crucial for multi-master setups. Proper topology, monitoring, and testing ensure resilience and correctness.

ALTERNATIVE DEFINITIONS

Wikipedia

In computing, data replication is the process of storing data in more than one site or node to improve availability and reliability.

Links replication directly to availability and reliability benefits across nodes.

Source

MySQL

Replication enables data from one MySQL database server (the source) to be copied to one or more MySQL database servers (replicas).

Concrete DBMS-centric definition illustrating source–replica architecture.

Source

PostgreSQL

Streaming replication allows a standby server to stay up to date by reading a stream of WAL records from the primary.

Describes WAL-based mechanism for near-real-time state on a standby server.

Source

RELATED TERMS

Synchronous Replication Asynchronous Replication High Availability Disaster Recovery Primary-Replica Change Data Capture Log Shipping Consensus Conflict Resolution Multi-Region Replication

COMPETING TERMS

Data Virtualization Sharding

RELATIONS

• Broader: Replication

KNOWLEDGE AREAS

Data Storage & Operations Data Engineering Data Architecture Cloud

TARGET AUDIENCE

Data Engineer Cloud Architect Data Architect DevOps Engineer Chief Information Officer

EXTERNAL SOURCES

• Wikipedia - Replication (computing)
• MySQL - MySQL replication architecture and setup
• PostgreSQL - Streaming replication and standby documentation
• Google Cloud - Globally distributed replication model
• Microsoft Architecture Center - Patterns including replication trade-offs

TOP QUESTIONS

Why implement Data Replication?

To improve availability, enable failover, reduce read latency, and support geo-proximity access for global users.

Category: Relevancy

What is the difference between synchronous and asynchronous Data Replication?

Synchronous commits to replicas before acknowledging writes, ensuring strong consistency; asynchronous replicates after, lowering latency but allowing lag.

Category: Comparison

What are the risks of Data Replication?

Lag, split-brain, and conflict resolution challenges can cause inconsistency; network failures amplify risk without careful design.

Category: Risks

How does Data Replication impact performance?

Replication adds write latency (especially synchronous) but can scale reads. Proper topology and capacity planning are crucial.

Category: Performance

What is multi-master Data Replication?

Multiple nodes accept writes, requiring conflict detection and resolution strategies like CRDTs or last-write-wins.

Category: Overview

How is Data Replication monitored?

Track replication lag, apply rates, WAL positions, and health of primary and replicas; alert on thresholds.

Category: Operations

What role does CDC play in Data Replication?

Change Data Capture streams transaction logs to propagate changes efficiently without full table copies.

Category: Example

How does Data Replication support disaster recovery?

Standby replicas in separate zones or regions enable failover and reduce RTO/RPO during outages.

Category: Governance

When is Data Virtualization preferable to Data Replication?

When data freshness requirements are modest and duplication costs are high, virtualization can simplify access without copying.

Category: Alternatives