Re: SET TIMING ON

Only puzzled hopefully, not sleepless :) It looks to me this is a SQL worksheet feature which allows you to buffer multiple statements. Probably the buffer is scanned sequentially and your command inserted everytime you execute it. Try repeating your statement in a single command environment like sql*plus.
What you describe is simply impossible. A repeat should perform faster, because the data is already in buffer cache.

Hth,

Sybrand Bakker, Oracle DBA

On Wed, 15 Sep 1999 13:27:02 GMT, Martin Douglas <Martin.Douglas_at_Boeing.com> wrote:

>Puzzled in Seattle:
>
>In SQL*Worksheet, I turned on my timing to clock a simple query and got
>some unexpected results. I clocked it at 0.39 (Elapsed) 0.22 (CPU) the
>first time. Then I kept rerunning the thing by pressing F5, always
>waiting for it to finish first. I expected to have the time be stable
>around 0.40 or so. But guess what! With each rerun, the damn thing
>kept rising by about 0.03 seconds. At around 0.8 seconds I thought it
>would not be possible to go any higher,... but it kept going up with
>each rerun. After another 30 reruns, I was up around 2.23 seconds!!!
>The table is not changing, I'm the only one on the database, and I'm the
>only one on the machine hosting the database. So what the #^$% is going
>on here??? Does this mean that if some application processed a query a
>couple of thousand times, that it would start taking hours for it to
>come back with 20 out of 60 records from a table? Drats!
>
>Help?
Received on Sat Sep 18 1999 - 04:03:12 CDT