1. It would be wonderful if I could write one query for all tables. Something like: query = 'schema_2.<all_tables>: "where >rownum=0"'. I >apparently do not understand what kind of query can be written. Is there any good documents about expdp >QUERY parameter?For explanation, you would want to look in the utility guide. I can't think of a single query that would work. The query is described as:
2. If I write query, what you have suggested, I would like to include only meta data for indexes to. I also can not write query like: >query='schema_2.index_name_for_table_1:"where rownum=0"'. I do not know is there solution for this? It would be great if I do >not need to rebuild indexes.I'm not sure what you mean for metadata only for indexes. Data Pump doesn't move anything but metadata for indexes. When we import an index, the index gets created on the fly. (unless you are using transportable tablespace) In that case, you can't filter out the data.
Allow me to describe my goal.The data can be eliminated for the one schema as show above. The data for the index will not be moved. It will only be the metadata. The index will be rebuilt during the import command.
I would like to export database, more precisely two schemas. But I have problem with amount of data in tables and large size of the >indexes from one schema, hence I would like to export only meta data for these schema objects.