Re: IMP: still more problems

Just my tuppence-worth.

Oracle is on record as saying that cumulative/incremental exports will be de-supported in the future, and (from memory only) the reason they gave was that RMAN is the cumulative/incremental guru in these parts.

This is just my wild speculation (usual disclaimers apply) but I don't honestly see much of a future for export and import. You've got dbms_metadata in 9i for the structural components. You've got logminer for transactional replay. You've got other things besides... strikes me that the days of export/import are distinctly numbered.

And it wouldn't surprise me in the least if one of the factors in that turned out to be exactly what you're saying: so much is happening in the database that it gets harder and harder to update export/import to keep up.

Regards
HJR "Chuck" <chuckh_at_softhome.net> wrote in message news:ah9nug$rb5fg$1_at_ID-85580.news.dfncis.de...
>
> "Sybrand Bakker" <postbus_at_sybrandb.demon.nl> wrote in message
> news:tskgjusrf6nusjhskojhoq2057mj0em6c8_at_4ax.com...
> > On Fri, 19 Jul 2002 11:34:57 -0400, "Chuck" <chuckh_at_softhome.net>
> > wrote:
> >
> > >I am encountering even *more* problems with the 8.1.7 import utility on
> AIX.
> > >Am I doing something wrong or is this utility just a piece of cr*p?
> > >
> > >Latest issue. I am trying to use a norows full export to create
indexes,
> > >primary keys, constraints, grants etc. on a database that I copied from
> > >another platform using CTAS on every table. (I used CTAS to load the
> data
> > >because import was too slow. It took 70+ hours whereas CTAS took about
> 3.)
> > >The create index and alter statements are in the export file to create
> the
> > >unique indexes for the PKs, then alter them to PK constraints. I can
grep
> > >them out of the file with no problem. The import however never even
tries
> to
> > >apply them. There are no errors in the log file indicating that it
failed
> to
> > >apply them for any reason. It just ignores them. It also ignores all of
> my
> > >FK constraints.
> > >
> > >Before I open a tar with oracle, is this expected behavior for EXP/IMP?
> Am I
> > >forgetting something?
> > >
> > >Here is the command I am using to run the export and the import. The
last
> > >line of the export log file says it finished with no warnings.
> > >
> > >exp user/pass file=norows full=y rows=n log=norows_export.log
> > >imp user/pass file=norows full=y log=norows_import.log
> > >
> > >
> >
> > >I am encountering even *more* problems with the 8.1.7 import utility on
> AIX.
> > >Am I doing something wrong or is this utility just a piece of cr*p?
> >
> > Millions of people can use it.
>
>
> I used to be one of them - from version 7.0 to 8.1. But I keep finding
more
> and more problems with a tool that used to work well. As features were
added
> to the RDBMS, imp/exp just hasn't kept up. Wrong sizes exported for
> datafiles, grants that are not exported, etc. I would warn anyone who is
> depending on it as a backup/restore method to reconsider. The most basic
> function of a backup/restore solution is to backup data and restore it to
> the same logical state as when it was backed up. EXP/IMP does not do
that -
> period. Is it any wonder Oracle keeps losing market share?
>
>
Received on Sat Jul 20 2002 - 04:38:25 CDT