7.2 Upgrade Oracle Machine Learning for R

You can upgrade OML4R to the current release from any previous release by reinstalling the product.

Note:

Upgrading from OML4R 1.1 is not supported on IBM AIX. To upgrade Oracle R Enterprise 1.1 on IBM AIX, first uninstall Oracle R Enterprise 1.1 (including R) and then download and install the later version.

To upgrade OML4R and migrate your data:

1. Ensure that you have the version of R that is required for the release of OML4R that you are installing.

   See the table of configuration requirements and server support in Oracle Machine Learning for R System Requirements for On-Premises Database for the R requirement.

   If you are don’t need to upgrade R, proceed to Step 2.

   If you do need to upgrade R, do the following:

   1. Back up your OML4R user schema, data store objects, R scripts, and the RQSYS schema.

   2. Remove the Oracle R Distribution RPMs or open source R components.

   3. Install the required R version, then proceed to Step 2.

2. To upgrade OML4R Server for Oracle Database Release 12c or earlier, run the server.sh or server.bat script to perform an installation.

   When the script detects an earlier version of OML4R Server, it asks if you want to upgrade. Type Yes to start the upgrade. (Type No to abort the process.)

   See About the Server Script for details.

3. To upgrade OML4R Server for Oracle Database 18c or later, use the rquncfg.sql script corresponding to the source database to uninstall the OML4R server, and rqcfg.sql to reinstall the OML4R server to the target database.

   See Install Oracle Machine Learning for R Server for Oracle Database 19c or later for details.

4. To upgrade OML4R Client, re-install the OML4R packages and supporting packages. You do not need to uninstall the current packages before installing the new packages.

   See Install Oracle Machine Learning for R Client for instructions.