Re: Query performance

"set timing on " should be more simple to use for SQLPLUS. This would outputs the execution time of all subsequent statements until you "set timing off"

alanpol_at_my-deja.com schrieb:

> We're running 8.0.5 on an NT server.
>
> I've a couple of quick questions on doing performance investigation.
>
> Firstly, are there any recomended ways to definitively record response
> times for a query ? The method I've been using in SQL*Plus is:
>
> set termout off
> timing start a
> <execute query>
> set termout on
> timing stop a (outputs timing - millisecs on NT)
>
> Will this produce accurate results ? Does anyone have a better way ?
>
> My second question is rather more general - any pointers to relevant
> references (especially URLs) would be great.
> If a query returns 200 rows, why can it be markedly quicker if the
> number of rows returned is limited to, say, 10 ? The execution plan
> should be the same, so:
> - what mechanisms are involved in returning a rowset to an application
> such as SQL*Plus
> - how do these mechanisms impact timing and vary with the number of
> rows returned
>
> Any help would be much appreciated.
>
> Alan
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
Received on Tue Apr 11 2000 - 00:00:00 CDT