RE: should I analyze system statistics on ORACLE 10GR2?
Date: Fri, 4 Mar 2011 09:10:32 -0600
Message-ID: <C5533BD628A9524496D63801704AE56D68C9D6C317_at_SPOBMEXC14.adprod.directory>
This is one of the areas that is really frustrating as a DBA...having to treat the SYS objects differently and THEN having to treat them differently in different databases. Some apps work well with SYS statistics, some fail miserably.
Chris Taylor
Sr. Oracle DBA
Ingram Barge Company
Nashville, TN 37205
Office: 615-517-3355
Cell: 615-663-1673
Email: chris.taylor_at_ingrambarge.com<mailto:chris.taylor_at_ingrambarge.com>
CONFIDENTIALITY NOTICE: This e-mail and any attachments are confidential and may also be privileged. If you are not the named recipient, please notify the sender immediately and delete the contents of this message without disclosing the contents to anyone, using them for any purpose, or storing or copying the information on any medium.
From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org] On Behalf Of David Fitzjarrell
Sent: Friday, March 04, 2011 8:17 AM
To: deshpande.subodh_at_gmail.com; mccdba1_at_yahoo.com
Cc: oracle-l_at_freelists.org; oracle-db-l_at_groups.ittoolbox.com
Subject: Re: should I analyze system statistics on ORACLE 10GR2?
Collecting system statistics is really an 'it depends' situation since if you're running Siebel or SAP applications in 10g and later releases creating a job to collect system statistics can cause problems with performance ( the SAP team where I worked my last contract did NOT collect system statistics because when they tried it caused some critical query plans to change for the worse ). In other cases it may be acceptable and improve the performance of some queries so it's best to test this tactic before you implement in a production environment so the users aren't caught by surprise with dismal execution plans for queries that once ran acceptably.
David Fitzjarrell
From: Subodh Deshpande <deshpande.subodh_at_gmail.com> To: mccdba1_at_yahoo.com
Cc: oracle-l_at_freelists.org; oracle-db-l_at_groups.ittoolbox.com Sent: Thu, March 3, 2011 11:03:19 PM
Subject: Re: should I analyze system statistics on ORACLE 10GR2? if you using cost based optimizer then you should collect stats.. I think...over a period earlier stats can be purged thanks....subodh
On 17 February 2011 01:26, dba1 mcc <mccdba1_at_yahoo.com<mailto:mccdba1_at_yahoo.com>> wrote: I saw a ORACLE document "Performance Tuning Guide for Siebel CRM Application on Oracle Part I - Cost-Based Optimization Guide" On this article it say "don't analyze system statistics":
System Optimizer Statistics
In 9i, Oracle introduced cpu_costing, a mechanism that allows the CPU cost of an operation to be included as part of the overall estimate. This feature is enabled in 9i only if you collect system statistics. This feature in 10g is enabled by default. Oracle Database 10g gathers two types of system statistics - statistics captured without a workload (noworkload) and statistics captured with a workload. In addition to CPU cost, disk I/O and memory utilization are also captured, to ensure that CBO picks the plan with the lowest estimated cost relative to other plans. Many Oracle DBAs attempt to leverage system optimizer statistics for further tuning of Siebel CRM application. Such actions may cause unpredictable degradation in performance of certain Siebel CRM application views and queries, and are strongly discouraged. We recommend that DBAs refrain from collecting system optimizer statistics, or altering any Oracle default settings pertaining to systems statistics.
I also see another ORACLE document say you need "analyze system statistics" to avoid stange thing happen.
Can Anyone tell me should I or Should NOT 'analyze system statistics"? Thanks.
--
http://www.freelists.org/webpage/oracle-l
--
DO NOT FORGET TO SMILE TODAY
--
http://www.freelists.org/webpage/oracle-l
Received on Fri Mar 04 2011 - 09:10:32 CST