<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	>
<channel>
	<title>Comments on: Statistics Gathering</title>
	<atom:link href="http://kerryosborne.oracle-guy.com/2009/02/statistics-gathering/feed/" rel="self" type="application/rss+xml" />
	<link>http://kerryosborne.oracle-guy.com/2009/02/statistics-gathering/</link>
	<description>Just another Oracle blog</description>
	<pubDate>Fri, 12 Mar 2010 10:33:37 +0000</pubDate>
	<generator>http://wordpress.org/?v=2.6.3</generator>
		<item>
		<title>By: osborne</title>
		<link>http://kerryosborne.oracle-guy.com/2009/02/statistics-gathering/#comment-6341</link>
		<dc:creator>osborne</dc:creator>
		<pubDate>Fri, 08 Jan 2010 17:23:46 +0000</pubDate>
		<guid isPermaLink="false">http://kerryosborne.oracle-guy.com/?p=964#comment-6341</guid>
		<description>Josip,

  I can't find that patch any more either (possibly due to it being rolled into 10.2.0.5). Check out this document on Metalink: 7116357.8.

Kerry</description>
		<content:encoded><![CDATA[<p>Josip,</p>
<p>  I can&#8217;t find that patch any more either (possibly due to it being rolled into 10.2.0.5). Check out this document on Metalink: 7116357.8.</p>
<p>Kerry</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Josip Voyager</title>
		<link>http://kerryosborne.oracle-guy.com/2009/02/statistics-gathering/#comment-6320</link>
		<dc:creator>Josip Voyager</dc:creator>
		<pubDate>Thu, 07 Jan 2010 15:35:21 +0000</pubDate>
		<guid isPermaLink="false">http://kerryosborne.oracle-guy.com/?p=964#comment-6320</guid>
		<description>Hi Kerry,

I’ve found the article on Optimizer Development Group Blog, but there are no patch 8413548 to enable APPROX_GLOBAL AND PARTITION option for statistics gathering on large partitioned tables in Oracle 10g release 2. 

I’ve search the Metalink without success. I assume that you mistype the patch number (and the guys behind Optimizer Development Group Blog) because patch number 8413548 doesn’t exist. Could you please help me to speed up my global statistics gathering.

Josip</description>
		<content:encoded><![CDATA[<p>Hi Kerry,</p>
<p>I’ve found the article on Optimizer Development Group Blog, but there are no patch 8413548 to enable APPROX_GLOBAL AND PARTITION option for statistics gathering on large partitioned tables in Oracle 10g release 2. </p>
<p>I’ve search the Metalink without success. I assume that you mistype the patch number (and the guys behind Optimizer Development Group Blog) because patch number 8413548 doesn’t exist. Could you please help me to speed up my global statistics gathering.</p>
<p>Josip</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: osborne</title>
		<link>http://kerryosborne.oracle-guy.com/2009/02/statistics-gathering/#comment-6038</link>
		<dc:creator>osborne</dc:creator>
		<pubDate>Fri, 18 Dec 2009 04:29:54 +0000</pubDate>
		<guid isPermaLink="false">http://kerryosborne.oracle-guy.com/?p=964#comment-6038</guid>
		<description>Hi Josip,

10g has a couple of options that can be used to simulate the 11g functionality. Namely the 'APPROX_GLOBAL AND PARTITION' ganularity setting on the gather_table_stats procedure and the dbms_stats.copy_table_stats procedure. Both are available with 10.2.0.4 but require a one off patch (8413548). It's not exactly the same, but close.

You could do something like this:

EXEC DBMS_STATS.GATHER_TABLE_STATS ('&#038;owner', '&#038;table_name', '&#038;partition_name', GRANULARITY =&gt; 'APPROX_GLOBAL AND PARTITION');

Do a google search and you should find a couple of hits on this. For example, the Optimizer Development Group Blog has this post:

http://optimizermagic.blogspot.com/2009/02/maintaining-statistics-on-large.html

Kerry</description>
		<content:encoded><![CDATA[<p>Hi Josip,</p>
<p>10g has a couple of options that can be used to simulate the 11g functionality. Namely the &#8216;APPROX_GLOBAL AND PARTITION&#8217; ganularity setting on the gather_table_stats procedure and the dbms_stats.copy_table_stats procedure. Both are available with 10.2.0.4 but require a one off patch (8413548). It&#8217;s not exactly the same, but close.</p>
<p>You could do something like this:</p>
<p>EXEC DBMS_STATS.GATHER_TABLE_STATS (&#8217;&#038;owner&#8217;, &#8216;&#038;table_name&#8217;, &#8216;&#038;partition_name&#8217;, GRANULARITY => &#8216;APPROX_GLOBAL AND PARTITION&#8217;);</p>
<p>Do a google search and you should find a couple of hits on this. For example, the Optimizer Development Group Blog has this post:</p>
<p><a href="http://optimizermagic.blogspot.com/2009/02/maintaining-statistics-on-large.html" onclick="javascript:pageTracker._trackPageview('/outbound/comment/optimizermagic.blogspot.com');" rel="nofollow">http://optimizermagic.blogspot.com/2009/02/maintaining-statistics-on-large.html</a></p>
<p>Kerry</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Josip Voyager</title>
		<link>http://kerryosborne.oracle-guy.com/2009/02/statistics-gathering/#comment-5997</link>
		<dc:creator>Josip Voyager</dc:creator>
		<pubDate>Tue, 15 Dec 2009 09:25:35 +0000</pubDate>
		<guid isPermaLink="false">http://kerryosborne.oracle-guy.com/?p=964#comment-5997</guid>
		<description>Hi Kerry, 

Let say I have one large table in my data warehouse that is partitioned by range (on date column).

My ETL procedure is executing on a daily basis and only last three or four partition have been changed between two ETL batch procedures.

I can maintain partition level statistics for changed partitions and local indexes easily (it’s very fast on my system), but the problem is global statistics maintenance.

Question is how can I incremental update/maintain global statistics for large, range partitioned table on 10g release 2 database running on IBM AIX server?

Please don’t tell me to upgrade to 11g because feature of incremental updating global statistics is implemented there, while my database is about 20 Tb, and Oracle 11g release 2 has not been released for IBM AIX platform yet.

I can send you the way I collect statistics right now if you like.

I’d like If you could provide an example.


Regards

Josip</description>
		<content:encoded><![CDATA[<p>Hi Kerry, </p>
<p>Let say I have one large table in my data warehouse that is partitioned by range (on date column).</p>
<p>My ETL procedure is executing on a daily basis and only last three or four partition have been changed between two ETL batch procedures.</p>
<p>I can maintain partition level statistics for changed partitions and local indexes easily (it’s very fast on my system), but the problem is global statistics maintenance.</p>
<p>Question is how can I incremental update/maintain global statistics for large, range partitioned table on 10g release 2 database running on IBM AIX server?</p>
<p>Please don’t tell me to upgrade to 11g because feature of incremental updating global statistics is implemented there, while my database is about 20 Tb, and Oracle 11g release 2 has not been released for IBM AIX platform yet.</p>
<p>I can send you the way I collect statistics right now if you like.</p>
<p>I’d like If you could provide an example.</p>
<p>Regards</p>
<p>Josip</p>
]]></content:encoded>
	</item>
</channel>
</rss>
