5.2.6 Lakehouse Incremental Load

As of MySQL 9.0.0, Lakehouse includes Lakehouse Incremental Load. After an initial data load, the data can become stale, and the external tables require a refresh. Subsequent calls to Auto Parallel Load will refresh the data.

Lakehouse Incremental Load uses the existing ENGINE_ATTRIBUTE, see: Section 5.2.2, “Lakehouse External Table Syntax”. It is not possible to change file parameters, file names, file patterns or file prefixes. However, it is possible to add or remove individual files if they still match the defined file pattern or file prefix.

After the initial data load, a subsequent call to Auto Parallel Load with the refresh_external_tables option set to true will refresh the data, see: Section 2.2.3.2, “Auto Parallel Load Syntax”.

If a subsequent call to Auto Parallel Load includes tables that have not yet been loaded to Lakehouse, then Auto Parallel Load will load them for the first time.

A call to Auto Parallel Load might contain both loaded and unloaded tables. Those that are unloaded will be loaded, and those that are already loaded will be refreshed.