4.1.3. Creating an Oracle to MySQL Deployment

Replication Operation Support
Statements Replicated No
Rows Replicated Yes
Schema Replicated No
ddlscan Supported Yes

The Oracle extractor enables information to be extracted from an Oracle database, generating row-based information that can be replicated to other replication services, including MySQL. The transactions are extracted by Oracle by capturing the change events and writing them to change tables; Tungsten Replicator extracts the information from the change tables and uses this to generate the row-changed data that is then written to the THL and applied to the destination.

Replication from Oracle has the following parameters:

  • Data is replicated using row-based replication; data is extracted by row from the source Oracle database and applied by row to the target MySQL database.

  • DDL is not replicated; schemas and tables must be created on the target database before replication starts.

  • Tungsten Replicator relies on two different users within Oracle configuration the configuration, both are created automatically during the CDC configuration:

    1. Publisher — the user designated to issue the CDC commands and generates and is responsible for the CDC table data.

    2. Subscriber — the user that reads the CDC change table data for translation into THL.

  • The slave replicator (applier), writes information into the target MySQL database using a standard JDBC connection.

The basic process for creating an Oracle to MySQL replication is as follows:

  1. Configure the Oracle database, including configuring users and CDC configuration.

  2. Configure the MySQL database, including creating tables and schemas.

  3. Extract the schema from MySQL and translate it to Oracle DDL.

  4. Install the Master replicator to extract information from the Oracle database.

  5. Install the Slave replicator to read data from the master database and apply it to MySQL.