what is a 'softer' soft parse?
----- Original Message -----
From: "Bobak, Mark" <Mark.Bobak_at_il.proquest.com>
To: <oracle-l_at_freelists.org>
Sent: Tuesday, March 02, 2004 5:41 PM
Subject: RE: concerning soft parses
> Ryan,
>
> In the ideal, perfect world, the first session would connect, the query
would be parsed (that would be a hard parse), then the variables would be
> bound, and the statement executed. That session would never have to parse
again. It simply needs to re-bind and re-execute as many times as
necessary.
> The second session would come along and parse (this would be a soft
parse), then bind variables and execute. Same here, it never needs to parse
again.
> So, you have one hard parse per unique sql statement and one soft parse
per session per unique sql statement. That's the ideal. Approximately zero
> applications work this way! ;-)
>
> To answer your question, utilizing the session cursor cache does not
eliminate soft parsing. It does, however, make for a 'softer' soft parse,
which
> provides for greater scalability. This can be demonstrated w/ some simple
testing and observing the amount of library cache latching. If you look
> at V$STATNAME, you'll see stats such as:
>
> STATISTIC# NAME
> ---------- ---------------------------------------------------------------
-
> 179 parse count (total)
> 180 parse count (hard)
> 191 session cursor cache hits
> 193 cursor authentications
>
> I did some experimentation here, and I thought it was this list that the
results were posted to.....yeah, I just checked my archive, look around
1/8/04
> for a thread entitled "Re: Suggestions needed: Latch free - library
cache".
>
> Hope that helps,
>
> -Mark
>
>
> -----Original Message-----
> From: ryan.gaffuri_at_cox.net [mailto:ryan.gaffuri_at_cox.net]
> Sent: Tuesday, March 02, 2004 5:15 PM
> To: oracle-l_at_freelists.org
> Subject: concerning soft parses
>
>
> the only way to eliminate a soft parse on a query with bind variables is
to set session_cached_cursors? Now is it possible for two different sessions
to share the same cursor or will this always result in atleast a soft parse?
>
> ----------------------------------------------------------------
> Please see the official ORACLE-L FAQ: http://www.orafaq.com
> ----------------------------------------------------------------
> To unsubscribe send email to: oracle-l-request_at_freelists.org
> put 'unsubscribe' in the subject line.
> --
> Archives are at http://www.freelists.org/archives/oracle-l/
> FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html
> -----------------------------------------------------------------
> ----------------------------------------------------------------
> Please see the official ORACLE-L FAQ: http://www.orafaq.com
> ----------------------------------------------------------------
> To unsubscribe send email to: oracle-l-request_at_freelists.org
> put 'unsubscribe' in the subject line.
> --
> Archives are at http://www.freelists.org/archives/oracle-l/
> FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html
> -----------------------------------------------------------------
Please see the official ORACLE-L FAQ:
http://www.orafaq.com
To unsubscribe send email to: oracle-l-request_at_freelists.org
put 'unsubscribe' in the subject line.
--
Archives are at http://www.freelists.org/archives/oracle-l/
FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html
-----------------------------------------------------------------
Received on Tue Mar 02 2004 - 19:44:18 CST
