To convert an Oracle database to Unicode

  1. As is recommended before any activity involving dbupgrade or dbreorg, perform a backup of the database that will be affected.
  2. Verify that the current database is in a good state. At the Landmark command line, type

    dbverify -q dataArea

    Any inconsistencies or issues should be resolved before you continue.

  3. Create a database instance that has a database character set of AL32UTF8. To check if your database already has this setting, see To determine if the Oracle database has the required settings for Unicode conversion.
  4. Open the db.cfg file at $LARUNDIR/dataArea in a text editor and add a new section representing the new database. For example, if you current db.cfg file has the following: 
    [common]
ARRAYBUFSIZE=10
INSERTBUFSIZE=10

[oradb]
DBSERVER=dbhost.acmecorp.com
ORACLE_HOME=c:\ora\ora11
PORT=50000
LOGINNAME=lawson
PASSWORD=landmark

    You could add a new section as follows:

    [oradbu]
DBSERVER=dbhost_u.acmecorp.com
ORACLE_HOME=c:\ora\ora11
PORT=50000
LOGINNAME=lawson
PASSWORD=landmark
  5. Run changeda -s to change the database name and character set on each database space. At the Landmark command line, type

    changeda –s dbname=newDBName charset=UTF-16 dataArea databaseSpace

    where newDBName is the new database section name (such as oradbu) you added to the db.cfg file previously.

  6. Run changeda -d in order to change the default validation character set in Landmark to UTF-16. At the Landmark command line, type

    changeda -d charset=UTF-16 productLine

  7. Rebuild the data area. At the Landmark command line, type

    stagelandmark dataarea

  8. Perform the database reorganization. At the Landmark command line, type

    activatelandmark dataarea

  9. Verify that the database reorganization was successful. At the Landmark command line, type

    dbverify -q dataArea

  10. When the process is complete, you can delete the old database section from the db.cfg file.