Re: shrinking rollback segments

I did once a close look at optsize-column which is for all segments really NULL.
The hwmsize for all rollback segments got in the meantime much more than the inital rollback segment was.
But ORACLE shrinks down - sometimes fully down to the intial size The only thing I did was an alter rollback segment shrink last saturday. But the values I used there are only valid for the execution of the statement, aren't they?
I was quite sure until now *g
So I still become a little bit confused about this behaviour.

Sybrand,
the term 'bug' I used was not so serious in this context. but if i create a datafile with multiple size of 4GB and loose in consequence my database...
and ORACLE lists this under term 'bug' too - all a point of view ;-)

"Peter van Rijn" <p.vanrijn_at_zhew.nl> schrieb im Newsbeitrag news:uja0a5er2dm68e_at_corp.supernews.com...
> Sybrand,
>
> Any reason for being so pissed off?
> Johannes never mentioned he's using OPTIMAL. On the contrary he says he is
> NOT using it, and still seeing shrinks in v$rollstat. He is seeking an
> explanation for that.
>
> Joahannes,
> As for the problem itself: I don't see this behavior on any of my
databases
> (of course I don't use optimal ;-) ), and perhaps you should double check
> your rbs-settings: optimal is really not being set?
>
> regards,
> Peter
>
>
> "Sybrand Bakker" <postbus_at_sybrandb.demon.nl> schreef in bericht
> news:n1n8ju0vqn85ugrhhfncn8m2r5d2o6o7ef_at_4ax.com...
> > On Tue, 16 Jul 2002 19:00:57 +0200, "Johannes" <johannes_at_merckle.de>
> > wrote:
> >
> > >Hello all,
> > >
> > >I have 60 rollback segments set with storage option 'optimal NULL'.
> > >Now I wonder myself why getting in v$rollstat much entries for the
column
> > >'shrinks' for each rollback segment.
> > >Shoulnd't they stay extented until i explicitly say: alter rollback
> segment
> > >shrink ?
> > >I restarted the DB, so there can't be any data from before. Or is it a
> bug
> > >like $GB? :-)
> > >
> > >Thanx for your help
> > >Johannes
> > >
> > >
> > Yeah sure, everything that doesn't work is automatically a bug.
> > Why on earth BTW do you have 60 rollback segments.
> > And implementing optimal was not one of Oracle's brightest moments:
> > stay away from it, you are increasing the possibility of hitting
> > ora-1555 snapshot too old.
> >
> > Regards
> >
> > Sybrand Bakker, Senior Oracle DBA
> >
> > To reply remove -verwijderdit from my e-mail address
>
>
Received on Wed Jul 17 2002 - 05:35:37 CDT