Oracle FAQ Your Portal to the Oracle Knowledge Grid
HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US
 

Home -> Community -> Mailing Lists -> Oracle-L -> RE: AUTOSYS performance issues

RE: AUTOSYS performance issues

From: Thomas Jeff <jeff.thomas_at_thomson.net>
Date: Fri, 20 Feb 2004 08:31:39 -0500
Message-ID: <358728A276824E419580403633AABFD0021F60A6@INDYSMAIL03.am.thmulti.com>


Niall,

I agree with avoiding FORCE, it's a solution of the last resort, too many side effects.

I've traced much of the issues to middle-tier anonymous PL/SQL code that is calling functions
on the backend.

Below is the kind of thing I see. Why would the last two calls, both using all bind=20
variables, be parsing so much? Is it because all of these bind variables are defined as=20
OUT parameters in the function?

BEGIN :retVal :=3D cred_get('NT','IS80001','TCEECOMP',:A0,:A1); END;

call count cpu elapsed disk query current rows
------- ------ -------- ---------- ---------- ---------- ----------



Parse 24 0.10 0.07 0 0 0 0
Execute 24 0.25 0.23 0 0 0 24
Fetch 0 0.00 0.00 0 0 0 0
------- ------ -------- ---------- ---------- ---------- ----------
total       48      0.35       0.31          0          0          0
24     =20


BEGIN :retVal :=3D get_event(:time0, :myID, :evt_num, :eoid, :joid, = :prio,
:event, :status, :alarm, :gmt, :excd, :mach, :pid, :jc_pid, :rnum,
:ntry,
:txt, :snq, :jname, :bname, :qprio, :AUTOSERV); END;

call count cpu elapsed disk query current rows
------- ------ -------- ---------- ---------- ---------- ----------



Parse 103 5.90 5.73 0 0 0 0
Execute 103 7.13 6.98 18 10729 2000 103
Fetch 0 0.00 0.00 0 0 0 0
------- ------ -------- ---------- ---------- ---------- ----------
total      206     13.03      12.71         18      10729       2000
103      =20


BEGIN :retVal :=3D get_jobrow(:ijoid, :joid, :job_name, :job_type, = :owner,
:permission, :box_joid, :machine, :n_retrys, :date_conditions,
:days_of_week, :run_calendar, :exclude_calendar, :start_times,

:start_mins,

:run_window, :command, :condition, :description, :term_run_time,
:box_terminator, :job_terminator, :std_in_file, :std_out_file,
:std_err_file, :watch_file, :watch_file_min_size, :watch_interval,
:min_run_alarm, :max_run_alarm, :alarm_if_fail, :chk_files,
:free_procs,
:profile, :heartbeat_interval, :auto_hold, :job_load, :priority,
:auto_delete, :numero, :max_exit_success, :box_success, :box_failure,
:status, :status_time, :last_start, :last_end, :next_start,
:run_window_end,

   :exit_code, :run_machine, :que_name, :jc_pid, :pid, :run_num, :ntry,
:appl_ntry, :last_heartbeat, :run_priority, :next_priority, :evt_num,
:over_num, :box_name, :command2, :condition2, :external_app,
:timezone);
  END; call count cpu elapsed disk query current rows
------- ------ -------- ---------- ---------- ---------- ----------



Parse 135 14.03 23.41 0 0 0 0
Execute 135 16.42 34.87 19 1493 0 135
Fetch 0 0.00 0.00 0 0 0 0
------- ------ -------- ---------- ---------- ---------- ----------

total 270 30.45 58.28 19 1493 0 135
=20

-----Original Message-----
From: oracle-l-bounce_at_freelists.org
[mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Niall Litchfield Sent: Friday, February 20, 2004 6:30 AM
To: oracle-l_at_freelists.org
Subject: RE: AUTOSYS performance issues

Hi Mladen

You don't need to enable query rewrite for CURSOR_SHARING, which is just as=3D well as it isn't available in std edition anyway much to my frustration.

I don't think I'd ever recommend FORCE as the setting except if I wished to=3D pull out my hair, 9.2 has SIMILAR which works quite well, but

  1. moves parsing from hard to soft which helps bad design, but doesn't smac=3D k the developer around the head and
  2. is still somehwat buggy - though being improved all the time. We have a =3D bug in that CURSOR_SHARING causes ORA-00600 error with intermedia = on 9.2.0.=3D 2 which I am informed by those nice folk at support is fixed = in 9.2.0.4.=3D20

What are the chances of anything called AUTO..... being A Good Thing(tm)

Niall Litchfield
Oracle DBA
Audit Commission
+44 117 975 7805

> -----Original Message-----
> From: mladen_at_wangtrading.com
> Sent: 19 February 2004 16:58
> To: mladen_at_wangtrading.com; oracle-l_at_freelists.org
> Subject: Re: AUTOSYS performance issues
>
>
> Well, you can improve parsing by enabling query rewrite and setting=20
> cursor_sharing to FORCE.
>
> On 02/19/2004 11:31:22 AM, Thomas Jeff wrote:
> > > We are experiencing sluggish performance with the AUTOSYS
> scheduler
> > that
> > > has it's repository in our 9.2.0.2/Standard Edition db on AIX=3D20
> > 4.3.3.
> > >
> > > Monitoring via statspack on 15-minute intervals, below are typical
> > > numbers.    The parsing numbers are horrendous.
> > >
> > > Anyone else using AUTOSYS having similar issues or had similar
> > issues?
> > > What did you do?
> > >
> > >
> > >
> > > Load Profile
> > > ~~~~~~~~~~~~                            Per Second       Per
> > Transaction
> > >                                    ---------------
> > ---------------
> > >                   Redo size:              8,201.40
> > 2,518.54
> > >               Logical reads:              1,431.28
> > 439.53
> > >               Block changes:                 57.73
> > 17.73
> > >              Physical reads:                287.78
> > 88.37
> > >             Physical writes:                  2.93
> > 0.90
> > >                  User calls:                 79.09
> > 24.29
> > >                      Parses:                 42.48
> > 13.05
> > >                 Hard parses:                 14.98
> > 4.60
> > >                       Sorts:                 18.16
> > 5.58
> > >                      Logons:                  0.27
> > 0.08
> > >                    Executes:                 69.45
> > 21.33
> > >                Transactions:                  3.26
> > >
> > >   % Blocks changed per Read:    4.03    Recursive Call %:    90.49
> > >  Rollback per transaction %:    0.00       Rows per Sort:     5.66
> > >
> > > Instance Efficiency Percentages (Target 100%)=20
> > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> > >             Buffer Nowait %:  100.00       Redo NoWait %:  100.00
> > >             Buffer  Hit   %:   79.90    In-memory Sort %:  100.00
> > >             Library Hit   %:   98.72        Soft Parse %:   64.74
> > >          Execute to Parse %:   38.83         Latch Hit %:   97.30
> > > Parse CPU to Parse Elapsd %:   49.44     % Non-Parse CPU:   52.86
> > >
> > >  Shared Pool Statistics        Begin   End
> > >                                ------  ------
> > >              Memory Usage %:   91.52   92.24
> > >     % SQL with executions>1:   68.78   65.79
> > >   % Memory for SQL w/exec>1:   41.37   20.20
> > >
> > > Top 5 Timed Events
> > > ~~~~~~~~~~~~~~~~~~
> >   %
> > > Total
> > > Event                                               Waits    Time
> > (s) Ela
> > > Time
> > > -------------------------------------------- ------------
> > -----------
> > > --------
> > > CPU time
> > 535
> > > 56.86
> > > library cache pin                                     141
> > 345
> > > 36.71
> > > latch free                                        153,263
> > 30
> > > 3.17
> > > db file scattered read                             25,593
> > 10
> > > 1.06
> > > db file sequential read                            32,554
> > 7
> > > .78
> > >
> > >
> > >
> > > --------------------------------------------
> > > Jeffery D Thomas
> > > DBA
> > > Thomson Information Services
> > > Thomson, Inc.
> > >
> > > Email: jeff.thomas_at_thomson.net
> > >
> > > Indy DBA Master Documentation available at:=20
> > > http://gkmqp.tce.com/tis_dba <http://gkmqp.tce.com/tis_dba>
> > > --------------------------------------------
> > >
> > >
> >
> > ----------------------------------------------------------------
> > Please see the official ORACLE-L FAQ: http://www.orafaq.com
> > ----------------------------------------------------------------
> > To unsubscribe send email to:  oracle-l-request_at_freelists.org put=20
> > '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=20
> '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
> -----------------------------------------------------------------
>
>=3D20




**********************************************************************
This email contains information intended for the addressee only. It may be confidential and may be the subject of legal and/or
professional privilege. Any dissemination, distribution, copyright or use of this
communication without prior permission of the sender is strictly prohibited.


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 Fri Feb 20 2004 - 07:31:39 CST

Original text of this message

HOME | ASK QUESTION | ADD INFO | SEARCH | E-MAIL US