From: "Jim Kennedy" <kennedy-family@attbi.com>
Newsgroups: comp.databases.oracle.server
References: <bV3f9.10197$_J3.409715@news0.telusplanet.net> <u04f9.259894$aA.44922@sccrnsc02> <tm4f9.10209$_J3.414519@news0.telusplanet.net> <3d7cddb3.21695256@ausnews.austin.ibm.com> <tI5f9.10234$_J3.426166@news0.telusplanet.net>
Subject: Re: Performance
Lines: 102
X-Priority: 3
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 6.00.2600.0000
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2600.0000
Message-ID: <D5df9.370622$me6.41828@sccrnsc01>
NNTP-Posting-Host: 12.241.212.232
X-Complaints-To: abuse@attbi.com
X-Trace: sccrnsc01 1031626211 12.241.212.232 (Tue, 10 Sep 2002 02:50:11 GMT)
NNTP-Posting-Date: Tue, 10 Sep 2002 02:50:11 GMT
Organization: AT&T Broadband
Date: Tue, 10 Sep 2002 02:50:11 GMT


Not a good test.  You are testing how fast your computer will format the
output visually.  You can't display 3,000 on the screen at one time anyway.
Your application should use bind variables, and reuse cursors, and just
retrieve what it can display and get more as the user requests it, also it
should use the array interface.(so you fetch rows in batches, not singly)
Your test does not measure what you are attempting to measure.
Jim

"Lucas" <Lucas_44@hotmail.com> wrote in message
news:tI5f9.10234$_J3.426166@news0.telusplanet.net...
> Yes with the Plus test scrolling across the screen, that's what it was.
 I'm
> not complaining that it is slow, I'm just checking if
> there are ways to optimize or if this is the normal response time.
> Retreiving that many records is
> not the norm but it can happen. Even with 3000 or so, it is slow...again,
> not the norm but if the user wants to see all
> 3000 on the screen, up to him/her. No, I haven't tried anything yet, just
> started looking at it. What about storage parameters? Just looking for
 some
> suggestions on where to start to possibly improve speed.
>
>
> "Ed Stevens" <spamdump@nospam.noway.nohow> wrote in message
> news:3d7cddb3.21695256@ausnews.austin.ibm.com...
> > So, is it taking a minute for Oracle to return a result set, or is it
 taking a
> > minute for Plus to scroll 17,000 rows across you screen?  What are you
 really
> > measureing, and how are you measuring it? Have you tried an explain
 plan?
 A
> > trace?  A TKPROF?
> >
> > On Mon, 09 Sep 2002 16:53:45 GMT, "Lucas" <Lucas_44@hotmail.com> wrote:
> >
> > >Ok, I was a little vague.
> > >Win2000
> > >P3 900 mhz
> > >384 RAM
> > >Oracle 8 , testing on local machine only, not even on server with
 concurent
> > >users.
> > >
> > >I'm trying to optimize our db, I'm testing our SQL calls and seeing if
> > >indexing would help speed up the program. We can retreive records from
> > >different
> > >levels in the program, the higher the level the more records. I tried a
> > >straight call (select * from table), it has 60 fields, 17,000 recs and
 it
> > >took close to 1 min
> > >to get all the rows. This is from SQL*Plus, so obviously in a program
 with
> > >user interface, getting all the record plus all the lookups makes it
 ever
> > >slower. The fields
> > >are basic varchars, integers, couple of varchar2(255), 1 longraw with
 hardly
> > >any images in it.
> > >
> > >Lucas
> > >
> > >
> > >
> > >
> > >"Michael J. Moore" <hicamel_x_the_spam@attbi.com> wrote in message
> > >news:u04f9.259894$aA.44922@sccrnsc02...
> > >> With the small amount of information you have given, I'd say that
 generally
> > >> you are right, that's pretty darn slow.
> > >> But if you are running on Win95, with 500 concurrent users, and your
 columns
> > >> are imbedded tables or clobs,
> > >> maybe that is totally reasonable. You really need to give a lot more
> > >> information for anybody to make the kind of assessment you are
 looking
 for.
> > >>
> > >> "Lucas" <Lucas_44@hotmail.com> wrote in message
> > >> news:bV3f9.10197$_J3.409715@news0.telusplanet.net...
> > >> > With just a basic (select * from ATable ) that has 60 fields and
 17,000
> > >> > records it takes close to 1 min to retrieve all the row, seems slow
 to
 me?
> > >> > Or am I expecting too much?
> > >> >
> > >> > Lucas
> > >> >
> > >> >
> > >>
> > >>
> > >
> > >
> >
> > --
> > Ed Stevens
> > (Opinions expressed do not necessarily represent those of my employer.)
>
>



