Re: Strange Datapump behavior.
From: Andrew Kerber <andrew.kerber_at_gmail.com>
Date: Mon, 20 Jun 2011 12:55:30 -0500
Message-ID: <BANLkTikSM9HN7HFu6PV9dsbKw7x_gu5yOg_at_mail.gmail.com>
On the import, the newuser tables were new and created by datapump, so no FK conditions could be causing it. It is possible that flashback could have caused a row count difference, but the magnitude of the difference is much more than I would have expected to be possible. There are no errors in the export or import. Can't find anything in dw* files.
Date: Mon, 20 Jun 2011 12:55:30 -0500
Message-ID: <BANLkTikSM9HN7HFu6PV9dsbKw7x_gu5yOg_at_mail.gmail.com>
On the import, the newuser tables were new and created by datapump, so no FK conditions could be causing it. It is possible that flashback could have caused a row count difference, but the magnitude of the difference is much more than I would have expected to be possible. There are no errors in the export or import. Can't find anything in dw* files.
The only thing I could think of was the possibility it is getting the row count from the space used instead of directly, and there was a huge amount of allocated unused space on the table.
On Mon, Jun 20, 2011 at 12:44 PM, rjamya <rjamya_at_gmail.com> wrote:
> Were those tables by newuser pre-existing or created via datapump? are any
> existing FK involved? is impdp complaining with import? anything in dwnn.rc
> files (usually in bdump) ? is it possible that due to flashback, the table
> actually had less rows than current?
>
> Raj
>
>
-- Andrew W. Kerber 'If at first you dont succeed, dont take up skydiving.' -- http://www.freelists.org/webpage/oracle-lReceived on Mon Jun 20 2011 - 12:55:30 CDT