Handling Large Data Sets in MySQL: Regular, Streaming, and Cursor Queries with MyBatis

The article discusses common scenarios for massive data operations such as data migration, export, and batch processing, and explains why loading millions of rows into JVM memory can cause OOM and slow performance.

It presents three query strategies for handling large result sets in MySQL: regular query (full or paginated load), streaming query using a server‑side cursor that returns an iterator, and cursor query that controls fetch size.

For each method, the author provides MyBatis‑Plus code examples, showing how to define a mapper interface, use @Select , @Options with ResultSetType.FORWARD_ONLY and fetchSize , and optionally @ResultType with a ResultHandler for row‑by‑row processing.

The article also notes practical considerations: the connection must stay open for streaming, the cursor must be closed after use, and fetch size determines memory consumption; it compares non‑streaming versus streaming memory behavior.

Additional tips include using MySQL’s ResultSet.next() behavior, differences with Oracle, and the importance of clearing temporary containers after each batch.