Oracle to OceanBase MySQL Migration in Stages | NineData
An Oracle to OceanBase MySQL migration must handle differences in object structure, data types, constraints, and ongoing business changes. NineData supports schema, full, incremental, and data comparison stages to move through the heterogeneous migration with observable checkpoints.
When to use this workflow
- Replace an Oracle core business system with OceanBase MySQL as part of a database localization project.
- Prepare an OceanBase MySQL environment for application integration and validation.
- Keep the source and target aligned during a migration rehearsal, parallel run, or target acceptance period.
Migration workflow
- Schema replication: Define the database, table, and column scope, then copy the selected structure to OceanBase MySQL.
- Full replication: Initialize existing Oracle data on the target. Evaluate resources and schedule the operation window before starting.
- Incremental replication: After full replication finishes, continue synchronizing source inserts, updates, and deletes to create a buffer for application validation and cutover.
- Data comparison: After the target catches up, compare critical objects and locate source-to-target differences.
Before you begin
- The Oracle source is
23ai,21c,19c,18c,12c, or11g, and the target is OceanBase MySQL V4.0. - Add the Oracle source and OceanBase MySQL target to NineData, and confirm network connectivity.
- Grant the permissions required for the selected replication type. See Oracle to OceanBase MySQL.
- For incremental replication, set the Oracle source to
ARCHIVELOGmode and enable supplemental logging. - Make sure replicated tables have a primary key or unique constraint and unique column names. Use the precheck to confirm target object, data, type, and constraint compatibility.
- Prepare business validation, backup, rollback, and test-data cleanup plans before cutover.
Create and validate the replication task
- Add the Oracle and OceanBase MySQL data sources to NineData.
- Create a Data Replication task, select the source and target, and choose schema, full, and incremental replication for the migration stages.
- Select the migration scope, configure object and field mappings, and confirm target object names.
- Save the task and run the precheck. Resolve connection, permission, log, or object compatibility issues.
- Start the task and monitor its status. After the target catches up, run data comparison and complete business validation before cutover.
caution
For a heterogeneous migration, use the precheck and business validation to confirm type, object-name, and constraint compatibility. Do not execute generated repair or cutover actions in bulk without review.