SAS Marketing Optimization functionality writes to various SAS^® data sets that are stored in a data store. As a result, there are times in which you might want to clean up orphaned or rogue entries in the data store, particularly if you see a decrease in performance.

To clean up the data store:

1. Run an interactive SAS session on the SAS compute-tier machine.
2. Run the SAS Marketing Optimization autoexec file. When you are prompted, use an external user account (for example, sasdemo) to authenticate.
3. Submit the following LIBNAME statement:
   libname mometa "&mometa" server=&moshare authdomain="&modatausr_authdomain";
4. Then submit the following statement:
   %mo_datastore_cleanup();

These steps do the following:

• automatically make a backup copy of the data store in the .../datastore/backups directory.
• keep the data store locked down
• rewrite the tables to eliminate deleted observations from the tables

Note: The cleanup tool is safe to run. It only removes orphaned rows to reduce the size of data-store tables.

The more scenarios and constraints you have, the longer the upload process takes. Part of the upload process is to copy the associated scenarios and update them based upon new incoming information from the new refreshed input data.

As a best practice, always remove scenarios that are no longer used. This action makes the upload process is faster.

Operating System and Release Information