5.1.2. Choosing a Master MySQL Standalone Replication Type

Depending on the downstream database target to which data is being replicted, the exact configuration of the deployment is different. There are two different types of configuraiton, which are configured by using one of two options to tpm:

  • Batch-based Applier Targets

    Where replication to the target database is being handled by a batch applier, that is, using the JavaScript batch loading environment as described in Section 7.2, “Batch Loading for Data Warehouses”, the batch configuration method should be used. When configuring for batch, the THL generated adds primary key data to inserts and full row data to deletes. This is requred because during batch loading the data is written to CSV files and must contain enough information for the batchloading process to operate correctly.

    A batch-loading configuration should be used when creating a master to be used with Vertica, Hadoop and Redshift.

    For more information on this type of standalone MySQL master configuration, see Section 5.1.2.1, “Deploying a Heterogeneous MySQL Master for Batch Appliers”.

  • Direct Applier Targets

    A standalone heterogeneous configuration should be used when the master is going to be used with appliers that write directly to the target database, either through a native connection or JDBC. For these targets, the primary key information can be used directly within the environment to perform updates. The additional insert and delete information will not assist the operation.

    A heterogeneous configuration should be used when creating a master for use with Oracle or MongoDB.

    For more information on this type of standalone MySQL master configuration, see Section 5.1.2.2, “Deploying a Heterogeneous MySQL Master for Direct Appliers”.