Beside handling ROWID correctly, the simplest way is to copy the content of the HIST, TCRV and DDL_COLMAP tables. For TCRV and DDL_COLMAP this is not supported by Oracle. You need TVD_FBA_HELPER (or something similar) to enable/disable DML on these tables. I suggest to discuss this further with Oracle support, because there is currently no supported way to do what you want.

I have disaccociate and reassociated the table. But when we qury the fda data using as of timestamp, they are returning wrong results. Oracle claims that since rowed is changed with table move, it does not support, as FDA data uses row id to track.

I have updated row id in FDA tables but even still I got wrong results.

is there a way t fix the issues?

In my attempt, I did things in this order:
– Create the target table without FBDA
– Populate the target table with ‘current’ data
– Enable FBDA for the target table
– Populate the HIST table

We are looking for the ability to clone a schema, such as from production to test, with the FBDA history in place. In one of our sample schemas, there is 1 FBDA-enabled table with over 100 million rows, and 5 others with over 10 million. We have already found that it is not practical to have FBDA enabled when we do these large initial loads (the FBDA background process backs up for days), and we really only need it to track changes after the initial population.

So, following this technique, we would have to:
– Have FBDA enabled during our initial ‘import’ into the target environment. (And I doubt that we would be able to use DataPump for this stage, either – I’ll have to try that.) This is already not workable.
– Add the entire contents of the current table to the HIST table. This would practically double our storage requirements.

Conclusion: We need to get at that TCRV table.