All,
I am well versed with normal usage of expdp and impdp, however, I have a specific situation for which I do not know a simple answer.
I have a 12c (12.1) database (normal non CDB), and I use REFERENCE PARTITIONING for certain tables because it makes the most sense. The parent table for these reference partitions is pretty small, only few tens to few hundreds of rows. But the tables that reference this may have millions.
The problems comes when we import, while importing sometimes we use QUERY clauses to import only certain data set based on requirements. So we use QUERY clause and parent table and use similar clauses on child tables as well. BUT (and this is a rather big but) IMPDP insists on importing largest segments first, so it tries to import child tables (where query clause points to a restriction via IN clause on parent table), and imports only 0 rows. This is because parent table is small, so it gets imported towards the end.
Is there any way I can run a single import and control order of import? So, I can import my parent (small) table first before child tables (large)? Currently I run two imports, first one just to import parent table, then second import to import everything else and no data from parent table (restricted using a clause like "where 1=2" to simplify parfile).
Thanks in advance
Raj