nshilling
asked on
Oracle 9.2 - using IMP with rows=n...how come the USERS.dbf datafile grew to 7GB and there is no data in any tables?
I recevied an export DMP file that was created with rows=n. I am importing that DMP file via IMP and my USERS datafile starts at .5 GB and after the IMP is done with ROWS=N (no data), the datafile grew to 7GB.
When you look at the stats, it shows Rows for most tables as >0, indicating there are rows in the tables, but when you do SELECT COUNT(1) there are none. A COMPUTE STATISTICS then calcs all those Rows as 0, proving there is no data in the tables, but yet the datafile still sits at 7GB. Why does it grow during IMP when there is no data being imported?
When you look at the stats, it shows Rows for most tables as >0, indicating there are rows in the tables, but when you do SELECT COUNT(1) there are none. A COMPUTE STATISTICS then calcs all those Rows as 0, proving there is no data in the tables, but yet the datafile still sits at 7GB. Why does it grow during IMP when there is no data being imported?
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Just wanted to add:
When you export you have the option 'compress extents'. If you selected 'Y' then Oracle looks at the total table size and makes the 'initial' extent that size for import. This guarantees the table will be contiguous on import.
So, it doesn't have to be 100's of tables in the export. It could be just 1 big table.
Add 'show=y' to the import and look at the output. you'll see the 'create table' statements but it won't do anything.