Migration from MySQL to Oracle using SQL Developer Migration Workbench

For migrating MySQL to Oracle we need to have:

• Oracle Database XE (small footprint, free version of Oracle DB)
• SQL Developer

Database installation should be very easy for this particular version, being the only one with a non-Java installer.
SQL Developer installation consists in unzipping the archive.

Now the steps:

1. Open SQL Developer and add MySQL Plugin. This is actually the ODBC driver for MySQL, the well-known MySQL connector. Without this you cannot connect to MySQL from SQL Developer.

2. Create a connection to Oracle with a powerful user (preferable SYS or SYSTEM)

3. Create another Oracle user/schema to hold the migration repository (things needed for the migration process)

4. Grant this user with CONNECT and RESOURCE roles

5. Grant the user with CREATE VIEW system privilege and click Apply

6. Create a connection to the recently created user/schema

7. Transform this empty schema into a migration repository.

8. This will create in the empty schema some internal objects to hold all the migration data. It’s not necessary to deal with this objects directly

9. Create a new connection to MySQL

10. Expand the connection, choose the desired database to migrate and click on “Capture Schema”

11. A window with captured objects will appear. You can also see what was captured by going to the “Captured Objects” pane on the left.

12. Go to Captured Objects pane on the left and choose “Convert to Oracle Model”

13. Tune the default data mappings if needed.

14. A window with converted objects will appear. You can also see what was captured by going to the “Converted Objects” pane on the left. It’s very important not to have any errors at this stage. Having errors at one object means that the object will not appear in the DDL and will not be created in the Oracle Database.

15. Go to Converted Objects pane on the left and choose “Generate”. This will generate a SQL script for creating all the objects in Oracle. The script will not be run yet, but just diplayed.

16. Choose the connection under you want to run the script. This should be a user that have the permission to create other user, so this is recommended to run in under the first connection (SYS or SYSTEM)

17. Run the script by pressing F5 or by using the appropriate icon

18. Create a new connection with the recently created user. This user has now all the objects migrated from MySQL, but no data.

19. Go to Converted Objects pane on the left and choose “Move Data”.

20. Select source and target connections. Make sure that the source is your MySQL connection and target is your recently created Oracle user (step 18).

21. Wait for Data Move to finish. This will work on separate threads. If the database is too big, you can tune the COMMIT interval.

22. Have a look at the Oracle migrated objects.

That was all!
Enjoy!

Read also:

Migration from Firebird to MySQL

In addition to what we installed during Part 1, we need to install MySQL 5.0 Community Server, the free version, which comes by default (not the Essential Edition) with MySQL Migration Toolkit.

1. Start MySQL Migration Toolkit

2. Go for Direct Migration

3. Set the Source connection as:

• Database System: Generic JDBC
• Class Name: org.firebirdsql.jdbc.FBDriver
• Connection string: jdbc:firebirdsql:localhost/3050:employee
• Username: sysdba
• Password: masterkey

4. Set the Target connection as:

• Database System: MySQL Server
• Hostname: localhost
• Port: 3306
• Username: root
• Password:

5. Get Schema information

6. Pick your desired schema to migrate

7. Reverse engineer the schema

8. Select objects to migrate (default all)

9. Set object mapping (go for Generic first)

10. Migrate the Metadata and have the SQL that is generating tables in MySQL

11. Manual edit the Generated code (no changes here for my example)

12. Create objects directly from the migration tool or save the SQL for running it later (do it online)

13. Execution of the DDL against the target database

14. Execution report

15. Data Mapping options, transfer data online or generate a script for running it later (do it online)

16. Data movement from Firebird to MySQL

17. Execution report

That was it!
The next step is migration from MySQL to Oracle

Good luck!

Read also:

Firebird connection through JDBC on Windows

I downloaded Firebird 1.5 which comes with a demo database Employee.fbd.
After this I downloaded SquirrelSQL, a cross-platform open-source JavaSQL Client which uses JDBC connection to a large number of databases. Firebird is one of them and as long as Firebird comes just with a console application, I found SquirrelSQL very useful.

In order to connect to Firebird through JDBC you will need

• Java JDK (I used 1.5)
• JayBird JDBC Driver for Firebird.

1. Install JDK into a convenient location, preferred C:\Java in order not to use spaces in the path.
2. Copy JayBird jar file to C:\Java\lib or add it to the CLASSPATH environment variable
3. Add jaybird21.dll to system path (I copied it to the system32 folder).
4. Start SquirrelSQL and try to connect to the Firebird JayBird driver in the drivers pane until it will appear with a blue check icon.
5. The class name should be org.firebirdsql.jdbc.FBDriver and make sure that jaybird jar file is present into the Java Class path. If it’s blue then means that the driver is active and we can test it.
6. We should first create a new alias for employee database in the Firebird aliases.conf file. After creating such an alias, the connection string should be “jdbc:firebirdsql:localhost/3050:employee”, username/password are sysdba/masterkey (default ones).
7. You are connected. you will see the objects list, data, DDL, whatever you need.

The next step is migration from Firebird to MySQL (the temporary host)

Good luck!

Read also: