Re: Historical Plan & Time

Thanks, Allen, Niall.

At least in this land of pleny, I have happened not to see that as a problem, mostly. With multi year multi billion orders flowing in company's way, the mood is to add CPUs and meomry rather than imporve code:(

 I happened to agree with the car tag I saw this morning 'Life is Good'. :)

We have got that covered, Niall. Thanks.

On 6/15/07, Niall Litchfield <niall.litchfield_at_gmail.com> wrote:
>
> It's good stuff indeed, though you need to have paid the extra money to
> Oracle in order to run this query, except in se/se1 where it impossible to
> legaly run this query. :(
>
>
>
> On 6/14/07, Shivaswamy Raghunath <shivaswamykr_at_gmail.com> wrote:
> >
> > Jeremy,
> >
> > Thanks. This is a good one.
> >
> > In order to find out which plan for the given SQL_ID was in effect at
> > what time, I added plan_hash_value to your query.
> > I think this is exactly what I was looking for:
> >
> > 16:01:55 Shiva>select distinct plan_hash_value, min(begin_interval_time)
> > first,
> > 16:02:02 2 max(end_interval_time) last
> > 16:02:02 3 from dba_hist_sqlstat natural join dba_hist_snapshot
> > 16:02:02 4 where sql_id='druz74ks91v35'
> > 16:02:02 5 group by sql_id, plan_hash_value
> > 16:02:02 6 order by 3;
> > Plan
> > Hash
> > Value FIRST LAST
> > ------------- -------------------------- --------------------------
> > 1366996245 22-MAY-07 02.00.52.480 PM 22-MAY-07 03.00.08.009 PM
> > 4248165369 25-MAY-07 03.00.46.849 PM 25-MAY-07 04.01.01.358 PM
> > 3113788595 30-MAY-07 02.00.42.408 PM 30-MAY-07 03.00.57.097 PM
> > 3244881609 30-MAY-07 04.00.11.473 PM 30-MAY-07 05.00.25.653 PM
> > 3700884553 31-MAY-07 12.00.05.902 PM 31-MAY-07 01.00.19.467 PM
> > 795688398 31-MAY-07 09.00.13.286 PM 31-MAY-07 10.00.27.115 PM
> > 2960769051 31-MAY-07 07.00.45.399 PM 01-JUN-07 12.00.11.623 PM
> > 853937962 01-JUN-07 01.00.27.646 PM 03-JUN-07 07.00.41.343 PM
> > 2354157810 22-MAY-07 09.00.36.037 AM 05-JUN-07 07.00.42.765 AM
> > 2424588842 10-JUN-07 05.00.44.530 PM 11-JUN-07 09.00.44.032 AM
> >
> > I have 10 different plan here and I know exactly which is in effect at
> > datte and time! Amazing, is it not?
> >
> > Thanks, again.
> >
> > On 6/14/07, Jeremy Paul Schneider < jeremy.schneider_at_ardentperf.com>
> > wrote:
> > >
> > > You can also query the AWR:
> > >
> > >
> > > SQL> select min(snap_id), max(snap_id) from dba_hist_sqlstat where
> > > sql_id='1nh4y7yurm73g';
> > > MIN(SNAP_ID) MAX(SNAP_ID)
> > > ------------ ------------
> > > 8301 8414
> > >
> > > SQL> col first format a35
> > > SQL> col last format a35
> > > SQL> select min(begin_interval_time) first, max(end_interval_time)
> > > last
> > > 2 from dba_hist_sqlstat natural join dba_hist_snapshot
> > > 3 where sql_id='1nh4y7yurm73g';
> > > FIRST LAST
> > > -----------------------------------
> > > -----------------------------------
> > > 04-JUN-07 05.00.26.802 PM 09-JUN-07 11.00.08.921 AM
> > >
> > >
> > >
> > > On 6/11/07, Hallas, John (EXP N-ARM) <john.hallas_at_lmco.com > wrote:
> > > >
> > > > Can you not write the dbms_xplan.display to an interim table
> > > > (create table xx as select * from ...) and then join that with the
> > > > last_load_time from v$sql based on the sql_id in both tables
> > > >
> > > > John
> > > >
> > > > ------------------------------
> > > > *From:* oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org]
> > > > *On Behalf Of *Shivaswamy Raghunath
> > > > *Sent:* 11 June 2007 14:05
> > > > *To:* Oracle-L
> > > > *Subject:* Historical Plan & Time
> > > >
> > > >
> > > > Hello.
> > > >
> > > > I have found that, the following sql is very helpful in finding out
> > > > the historical (What is in AWR) plan of a given SQL_ID.
> > > >
> > > > select * from table (sys.dbms_xplan.display_awr('&Your_SQL_ID')) ;
> > > >
> > > > But so far I could not figure out, how I can figure out WHEN this
> > > > plan was in effect. Is there a way, I can corelate this plan to Sysdate or
> > > > in other words, find out when this plan was in effect?
> > > >
> > > > Thanks for your insight,
> > > > Shiva
> > > >
> > >
> > >
> > >
> > > --
> > > Jeremy Schneider
> > > Chicago, IL
> > > http://www.ardentperf.com/category/technical
> >
> >
> >
>
>
> --
> Niall Litchfield
> Oracle DBA
> http://www.orawin.info

--
http://www.freelists.org/webpage/oracle-l