The other thing you want to make sure you are doing is using bind variables.
Jim
"Ed Stevens" <Ed_Stevens_at_nospam.noway.nohow> wrote in message
news:3c39d1ab.14432322_at_ausnews.austin.ibm.com...
> Our tuning efforts have lead us to looking at SQL*net issues. I ran a
series of
> tests from SQL*Plus, varyinig both the SDU/TDU values and the array size.
>
> My method of testing was to run a fixed set of queries (taken from a
running
> production instance) with client trace set to 16, then counting the number
of
> packets received by the client. I would consider a reduction in the
number of
> packets received to indicate less network traffic, less re-assembling of
> packets, and thus indicative of some degree of performance improvement.
>
> The results of these tests showed that playing with SDU/TDU resulted in a
small
> decrease in packets received, but increasing array size made a significant
> reduction in packets received.
>
> Now, the real problem. For these tests, setting array size was a function
of
> the SQL*Plus session and is set by a Plus SET command. So, how to
translate
> that to something the application can use? The app is written with MS
tools and
> uses ADO as the microsoft layer accessing the Oracle DB. We have looked
all
> through the ADO API reference and the closest we come is a discussion of a
> CacheSize property. Within that discussion was this statement:
>
> "CacheSize is based on the MaximumOpenRows provider-specific property (in
the
> Properties collection of the Recordset object).
>
> This looks like MaximumOpenRows is what I'm after, but I can find no other
> reference anywhere in the reference manual or anywhere on Technet. Does
anyone
> here have any experience with this, to point me in the right direction>
> --
> Ed Stevens
> (Opinions expressed do not necessarily represent those of my employer.)
> --
> Ed Stevens
> (Opinions expressed do not necessarily represent those of my employer.)
Received on Mon Jan 07 2002 - 20:32:34 CST
