Re: Global Temporary Table performance with inserts.

Jonathan Lewis wrote:
> People have been known to use Oracle
> for multi-user systems - the fact that you
> cannot TIME a difference on your simple,
> single-user test case does not mean the
> test case will perform well at high concurrency.
>
> If (for the sake of argument) one option produces
> one redo entry, and the other option produces
> 100 redo entries, with corresponding difference in
> the number of redo allocation latch gets, then there
> is a potential concurrency issue that may be significant
> in a highly concurrent system.

I absolutely agree.

But in all the experience I have ... not once has a tuning problem come down to 100 rows in a GTT. Doesn't mean it couldn't be a probem. But usually the problems are far more basic issues.

-- 
Daniel Morgan
http://www.outreach.washington.edu/ext/certificates/oad/oad_crs.asp
http://www.outreach.washington.edu/ext/certificates/aoa/aoa_crs.asp
damorgan_at_x.washington.edu
(replace 'x' with a 'u' to reply)