Remove snapshot management from metadata import - Use datastore only

If datasets could utilize the metadata from SKP new datastores rather than from snapshot management datasources that would enable a proper separation of refresh execution and metadata importing utilized for object management within migrate.

• At CHS we have had to create special snapshot datasources specifically for metadata importing because Maximo sources do not have primary keys and because of db2 source data that we cannot import metadata from.

• It has created difficulties involving hours of labor to switch tables back and forth in snapshot management between a datasource that looks at the ADMM SQL environment for dataset metadata and the Maximo sources utilized for data refreshes.

• This would also help with activation of tables for building source and target tables vs. activating the tables for refreshing. For our JDE sources we have our tables set to inactive because accidental kicking off of those tables would truncate the table (Replicate should do that not ADMM). In order to build the source tables for them we have to temporarily activate them and twice a CHS resource has kicked one of those not deactivated and caused us to have to do a db restore to get the data back.

Now knowing the overlap between metadata and refreshing within Snapshot management and the issues we have had with that I would recommend that people have separate snapshot datasources. One snapshot for metadata importing and one for refreshing. But this would only be a workaround, and would be avoided by using these datastores that were added with replicate preview just released for metadata.

Appreciate your consideration!