Re: Maybe stupid question about exp/imp

"Noons" <wizofoz2k_at_yahoo.com.au.nospam> wrote in message news:Xns935ECC66CBB1BTokenthis_at_210.49.20.254...
> Following up on Jan Gelbrich, 15 Apr 2003:
<snip>
> > Recently I heard some roumors saying that, after re-importing,
> > there might be a slightly better overall performance, "for some reason".
>
>
> Those are either greatly exagerated or related to
> some very old version of Oracle. Don't listen to them.

I know where you are coming from, and certainly would wish to reinforce the 'exp/imp is not a tuning tool' message but (theres always a but)

Resetting HWM, coalescing free space, reducing or eliminating chaining, reducing or eliminating migration, separating objects across disks, changing block size can all have a performance impact on a database and can all happen when a database is EXPorted/IMPorted. It is important not to confuse the reason for the performance increase with the process one has just performed. Otherwise it is relatively easy to 'prove' , for example, that storing objects in one extent improves performance - just do EXP COMPRESS=Y and reimport. It must be the changed number of extents that accounts for the improvement right?

:(

--
Niall Litchfield
Oracle DBA
Audit Commission UK