I have a strange import problem. I am compressing and splitting an export
to keep file sizes under 2GB. When I run an import with a show=y or
indexfile=y (rows=n) switch, I get the expected results. When I run the
import (indexes=n, grants=y), I consistently terminate unsuccessfully
about halfway through the import.
Since the entire .dmp file has to be successfully reconstructed in
all three cases, whyever mightn't it be working in the last case?
Any ideas, I'd appreciate hearing about them.
If it is convenient, e-mail will reach me faster.
Regards,
jh
--
Jared Hecker | HWA Inc. - Oracle architecture and Administration
ja...@hwai.com | ** serving NYC and New Jersey **
In article <SZW43.498$m4.2288@nntp1>,
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
What are you getting to indicate the termination? Broken pipe?
Are things being created where they were previously (in particular,
things like constraints mysteriously winding up in system tablespace,
running it out of extents)?
Are you specifying BUFFER, and is it bigger than your biggest row?
And excuse the silly ones, they have to be asked:
Does the indexfile created work?
IGNORE=Y?
PCTFREE larger than it was when data created? (So it is trying to load
data less compacted than when exported... :)
>
>Any ideas, I'd appreciate hearing about them.
>
>If it is convenient, e-mail will reach me faster.
>
>Regards,
>jh
>--
>Jared Hecker | HWA Inc. - Oracle architecture and Administration
>ja...@hwai.com | ** serving NYC and New Jersey **
--
These opinions are my own.
http://ourworld.compuserve.com/homepages/joel_garry Remove nospam to reply.
mailto:joel-...@nospam.home.com Oracle and unix guy.