Re: Maxextents unlimited on LMT ?

Whoops !

I can't believe I fell into that trap.

The number of times I've explained how
initial, next, and minextents get used (abused ?) when you have LMTs.

>
>And 10 updates to tsq$ shouldn't
>have taken that long anyway.
>

Yes, I'll believe that 990MB/32K = 31,680 inserts might take that long.

Do you want to ask Pete how big his
log buffer and redo log file were, and
how many times he checkpointed in the
period ?

On the other hand, is it possibly that one of you has a version of Oracle which updates tsq$ on every extent release, and the other has a version that updates only when the segment is released. It might be worth
checking your v$sql after the test for the number of executions of updates to tsq$. (It might also be worth checking the value in tsq$, or the user-friendly view, to see if your quota has actually been given back).

--
Jonathan Lewis
http://www.jlcomp.demon.co.uk

Next Seminars
        UK            July / Sept
        Australia      July / August
        Malaysia        September
        USA (MI)        November
http://www.jlcomp.demon.co.uk/seminar.html

Jonathan Lewis wrote in message
<1026249931.1873.0.nnrp-10.9e984b29_at_news.demon.co.uk>...


>

>Here's an oddity -

>

>How come the create reported time

>correctly, but the drop fell back into

>the mod(60) bug on the seconds ?

>

>And 10 updates to tsq$ shouldn't

>have taken that long anyway.

>

>

>--

>Jonathan Lewis

>http://www.jlcomp.demon.co.uk

>

>Next Seminars

>        UK            July / Sept

>        Australia      July / August

>        Malaysia        September

>        USA (MI)        November

>http://www.jlcomp.demon.co.uk/seminar.html

>

>Connor McDonald wrote in message <3D2B34C3.1E50_at_yahoo.com>...

>>

>>Table created.

>>

>>Elapsed: 00:02:04.57

>>SQL> drop table x.p;

>>

>>Table dropped.

>>

>>Elapsed: 00:01:92.14

>

>

>