There is an oracle 11g schema with amount of data about 700 Gb. There are tables/indexes only in that schema, no procedures/packages.
Doing expdp on that schema
expdp schemas=TheSchema directory=dir dumpfile=dump logfile=log content=metadata_only
, I expected quite a litte dumpfile because of
metadata_only. Doing tools/export-user-objects in pl/sql developer gives an sql file with text size a half of megabyte.
But instead there appeared a 20 Gb dumpfile!
Is there something large in dumpfile created by this expdp, like index contents?
What could be done to minimize the dumpfile done with expdp, if I want just to create tables like in TheSchema in another schema?
UP: There is quite a few number of tables/indexes in the text-sql-dump of schema:
> grep "create table" TheSchema.sql | wc -l
> grep "index" TheSchema.sql | wc -l
> grep "partition" TheSchema.sql | wc -l