Master MyBatis Second‑Level Cache to Supercharge Your Queries

MyBatis Second-Level Cache

Cache is data stored in memory, often derived from database query results; using cache avoids frequent database interactions and improves query response speed.

MyBatis provides cache support, divided into first-level and second-level cache as shown:

First-level cache: SqlSession‑level cache, valid only within a SqlSession.

Second-level cache: mapper‑level cache, shared across the same namespace; it must be manually enabled.

Why Use MyBatis Second-Level Cache?

First-level cache’s sharing scope is limited to a single SqlSession; enabling second-level cache allows multiple SqlSessions to share cached data.

How to Configure MyBatis Second-Level Cache

1. Enable globally in the MyBatis configuration file:

<settings>
    <!-- Enable logging -->
    <setting name="logImpl" value="STDOUT_LOGGING"/>
    <!-- Enable second-level cache (global) -->
    <setting name="cacheEnabled" value="true"/>
</settings>

2. Configure cache for a specific mapper namespace:

<!-- cache node: configure cache for current namespace mapper -->
<cache eviction="LRU" flushInterval="100000" readOnly="true" size="1024"/>

3. Ensure cached entity classes implement the Serializable interface so they can be serialized and deserialized.

Implementation Principle

When second-level cache is enabled, MyBatis decorates the Executor with CachingExecutor; before the first-level cache query process, CachingExecutor checks the second-level cache. The workflow is:

All statements under the same namespace share the same Cache, making the second-level cache a global variable shared by multiple SqlSessions.

The query execution flow becomes: second-level cache → first-level cache → database.

Application Scenarios

Second-level cache is recommended for data with high query frequency and low update frequency.

For read‑heavy requests where real‑time freshness is not critical, using MyBatis second-level cache can reduce database access and improve response speed.

Typical cases include time‑consuming statistical SQLs, telephone bill queries, etc.; by setting an appropriate refresh interval, MyBatis can automatically clear the cache based on data change frequency.