Understanding Database Read‑Write Separation and Master‑Slave Replication

Hello, I am mikechen. In high‑concurrency and large‑data scenarios, database performance becomes critical, especially regarding read‑write separation mechanisms.

What is read‑write separation

Read‑write separation divides the database into a primary (master) for write operations and multiple replicas (slaves) for read operations, allowing reads to be load‑balanced across the replicas.

The master and replicas synchronize data through a communication mechanism, a common pattern in database architectures.

Why use read‑write separation?

Writing large volumes of data can be time‑consuming (e.g., inserting 10,000 rows into Oracle may take three minutes), whereas reading the same amount can be much faster (e.g., retrieving 10,000 rows may take only five seconds).

Separating reads from writes dramatically improves query efficiency by offloading read traffic to replicas.

When should read‑write separation be applied?

It is not mandatory for every database, but it becomes advantageous when an application performs many reads and few writes, reducing database load and enhancing performance. Other optimization techniques include sharding, partitioning, or using search engines.

Principle of read‑write separation

The master handles transactional write operations (INSERT, UPDATE, DELETE), while replicas handle SELECT queries.

This distributes request traffic across different database nodes, allowing one master and multiple slaves.

The overall architecture is shown below:

A key technology is master‑slave replication, which copies data from the master to the slaves to ensure consistency.

MySQL master‑slave replication example

Master‑slave replication steps

The slave connects to the master and creates an I/O thread to pull binlog updates from the master.

The fetched binlog entries are written to the slave’s relay log file.

The slave creates an SQL thread that replays the relay log, applying the changes to its own data, making it identical to the master.

When the master updates data, it records changes in the binary log (binlog) and starts a binlog dump thread that streams these updates to the slave’s I/O thread. This process is asynchronous, avoiding performance bottlenecks.

Read‑write separation summary

In production environments, handling both reads and writes on a single database server cannot meet requirements for security, high availability, and high concurrency.

Using master‑slave replication to synchronize data and applying read‑write separation enhances database concurrency, solves availability issues, and improves performance.

For job seekers preparing for interviews, a comprehensive Java interview question set covering Java, multithreading, JVM, Spring, MySQL, Redis, Dubbo, and middleware is available via the link below.

Java interview questions and answers