Re: V$SQLAREA
From: Jonathan Lewis <jonathan_at_jlcomp.demon.co.uk>
Date: Thu, 15 Dec 2011 17:03:39 -0000
Message-ID: <MZSdnTrvqO_ftXfTnZ2dnUVZ7qWdnZ2d_at_bt.com>
Date: Thu, 15 Dec 2011 17:03:39 -0000
Message-ID: <MZSdnTrvqO_ftXfTnZ2dnUVZ7qWdnZ2d_at_bt.com>
However, just as they make something viable, Oracle often makes it redundant.
v$sqlstats came out as a "latch-free" equivalent of v$sqlarea in 10g, and then 11g gave us v$sqlstats_plan_hash.
The latter sums child cursors by plan hash value - so you can get a better idea from the summary of the different amount of work that different plans may give for the same query.
-- Regards Jonathan Lewis http://jonathanlewis.wordpress.com Oracle Core (Apress 2011) http://www.apress.com/9781430239543 "Mladen Gogala" <gogala.mladen_at_gmail.com> wrote in message news:pan.2011.12.14.20.47.02_at_gmail.com...Received on Thu Dec 15 2011 - 11:03:39 CST
| In older versions of Oracle, there was a rule saying that V$SQLAREA is
| evil because queries against it will hit library cache latch once for
| every row. I was unable to confirm this in 11.2.0.3. The only latch it
| seems to be hitting is SQL memory manager workarea list latch.
| Did anyone else notice that? If my findings are right, that would make
| V$SQLAREA view much more useful for monitoring and querying purposes.
|
|
| --
| http://mgogala.byethost5.com