<?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: The Three Pillars of Oracle Data Warehousing</title>
	<atom:link href="http://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/feed/" rel="self" type="application/rss+xml" />
	<link>http://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/</link>
	<description>Oracle Data Warehouse Design and Architecture</description>
	<pubDate>Fri, 04 Jul 2008 18:47:32 +0000</pubDate>
	<generator>http://wordpress.org/?v=MU</generator>
		<item>
		<title>By: A metablog&#8230; &#124; El Mundo Con Otros Ojos</title>
		<link>http://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-51071</link>
		<dc:creator>A metablog&#8230; &#124; El Mundo Con Otros Ojos</dc:creator>
		<pubDate>Wed, 25 Jun 2008 22:01:37 +0000</pubDate>
		<guid isPermaLink="false">https://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-51071</guid>
		<description>[...] data warehousing techniques and other things both Oracle and non-Oracle related. He needs to learn how to spell – but does all right most of the [...]</description>
		<content:encoded><![CDATA[<p>[...] data warehousing techniques and other things both Oracle and non-Oracle related. He needs to learn how to spell – but does all right most of the [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: infonitive</title>
		<link>http://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-51029</link>
		<dc:creator>infonitive</dc:creator>
		<pubDate>Mon, 26 May 2008 00:32:19 +0000</pubDate>
		<guid isPermaLink="false">https://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-51029</guid>
		<description>Interesting views on data warehousing. You might find www.infonitive.com an interesting forum.</description>
		<content:encoded><![CDATA[<p>Interesting views on data warehousing. You might find <a href="http://www.infonitive.com" rel="nofollow">http://www.infonitive.com</a> an interesting forum.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jonathan Lewis</title>
		<link>http://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1064</link>
		<dc:creator>Jonathan Lewis</dc:creator>
		<pubDate>Sun, 28 May 2006 10:29:00 +0000</pubDate>
		<guid isPermaLink="false">https://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1064</guid>
		<description>"why isn't query-rewrite automatic"

It's all in the way you read the text. 

As you pointed out, there's an optional clause that you can put in the mview definition.  What I wrote was "Then create a materialized view with query rewrite enabled" - i.e. create a materialized view with that optional clause in place.</description>
		<content:encoded><![CDATA[<p>&#8220;why isn&#8217;t query-rewrite automatic&#8221;</p>
<p>It&#8217;s all in the way you read the text. </p>
<p>As you pointed out, there&#8217;s an optional clause that you can put in the mview definition.  What I wrote was &#8220;Then create a materialized view with query rewrite enabled&#8221; - i.e. create a materialized view with that optional clause in place.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: David Weigel</title>
		<link>http://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1063</link>
		<dc:creator>David Weigel</dc:creator>
		<pubDate>Fri, 26 May 2006 22:16:00 +0000</pubDate>
		<guid isPermaLink="false">https://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1063</guid>
		<description>&lt;I&gt;with query rewrite enabled...&lt;/I&gt;

I'm sure I'm missing something obvious, but why isn't query-rewrite automatic?  There's a special grant the user needs, and there's a clause to put in the mview defintion.  When would you turn query-rewrite &lt;I&gt;off&lt;/I&gt;?</description>
		<content:encoded><![CDATA[<p><i>with query rewrite enabled&#8230;</i></p>
<p>I&#8217;m sure I&#8217;m missing something obvious, but why isn&#8217;t query-rewrite automatic?  There&#8217;s a special grant the user needs, and there&#8217;s a clause to put in the mview defintion.  When would you turn query-rewrite <i>off</i>?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: David Aldridge</title>
		<link>http://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1062</link>
		<dc:creator>David Aldridge</dc:creator>
		<pubDate>Sun, 21 May 2006 08:50:00 +0000</pubDate>
		<guid isPermaLink="false">https://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1062</guid>
		<description>Along a similar line, having two summary tables with exactly the same data but with different partitioning schemes is a useful technique. Query rewrite is very happy to pick the appropriate one to use.</description>
		<content:encoded><![CDATA[<p>Along a similar line, having two summary tables with exactly the same data but with different partitioning schemes is a useful technique. Query rewrite is very happy to pick the appropriate one to use.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jonathan Lewis</title>
		<link>http://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1061</link>
		<dc:creator>Jonathan Lewis</dc:creator>
		<pubDate>Sat, 20 May 2006 10:28:00 +0000</pubDate>
		<guid isPermaLink="false">https://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1061</guid>
		<description>Pete,

True: any attempt at clustering is likely to be a benefit along just one dimension - but it's always worth checking if there's one dimension that's worth the effort, so long as it doesn't cause too much disruption elsewhere.

There's also the option that you can cluster in a second dimension by copying the data in a different order, e.g.

create table t2 as 
select * from t1 
order by colx

Then create a materialized view with query rewrite enabled so that queries against t1 can visit t2.

It's not a strategy that I've used often - but it's just a special case of a summary table, and disks are just so big these days you need something to fill the space with to stop other people from interfering.

Regards
Jonathan Lewis</description>
		<content:encoded><![CDATA[<p>Pete,</p>
<p>True: any attempt at clustering is likely to be a benefit along just one dimension - but it&#8217;s always worth checking if there&#8217;s one dimension that&#8217;s worth the effort, so long as it doesn&#8217;t cause too much disruption elsewhere.</p>
<p>There&#8217;s also the option that you can cluster in a second dimension by copying the data in a different order, e.g.</p>
<p>create table t2 as<br />
select * from t1<br />
order by colx</p>
<p>Then create a materialized view with query rewrite enabled so that queries against t1 can visit t2.</p>
<p>It&#8217;s not a strategy that I&#8217;ve used often - but it&#8217;s just a special case of a summary table, and disks are just so big these days you need something to fill the space with to stop other people from interfering.</p>
<p>Regards<br />
Jonathan Lewis</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Pete_S</title>
		<link>http://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1060</link>
		<dc:creator>Pete_S</dc:creator>
		<pubDate>Sat, 20 May 2006 07:45:00 +0000</pubDate>
		<guid isPermaLink="false">https://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1060</guid>
		<description>Clustering (with or with out the P) can be problematic - we can only really cluster on one dimension (perhaps a couple if there is some correlation between them – say a brewer supplies products that are beers!) but when you try to cluster product, location and customer…</description>
		<content:encoded><![CDATA[<p>Clustering (with or with out the P) can be problematic - we can only really cluster on one dimension (perhaps a couple if there is some correlation between them – say a brewer supplies products that are beers!) but when you try to cluster product, location and customer…</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: David Aldridge</title>
		<link>http://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1059</link>
		<dc:creator>David Aldridge</dc:creator>
		<pubDate>Fri, 19 May 2006 09:58:00 +0000</pubDate>
		<guid isPermaLink="false">https://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1059</guid>
		<description>Note to self: don't go swimming with Pete :D

DB: yes, I think if I had to pick just one of those three techniques it would be summaries. There's nothing like making a scan of 100,000,000 fact table records return in 0.01 seconds for brightening your day.

JL: Maybe Pack, Parcel or Pigeonhole would also be psuitable.</description>
		<content:encoded><![CDATA[<p>Note to self: don&#8217;t go swimming with Pete :D</p>
<p>DB: yes, I think if I had to pick just one of those three techniques it would be summaries. There&#8217;s nothing like making a scan of 100,000,000 fact table records return in 0.01 seconds for brightening your day.</p>
<p>JL: Maybe Pack, Parcel or Pigeonhole would also be psuitable.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jonathan Lewis</title>
		<link>http://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1058</link>
		<dc:creator>Jonathan Lewis</dc:creator>
		<pubDate>Fri, 19 May 2006 09:30:00 +0000</pubDate>
		<guid isPermaLink="false">https://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1058</guid>
		<description>You could add Placement, Postition, or possibly just Psorting - as another strategy for optimising physical colocation of data.  (I don't think you can get away with claiming a silent P in Pclustering, otherwise that would do as the more generic description).


Jonathan Lewis</description>
		<content:encoded><![CDATA[<p>You could add Placement, Postition, or possibly just Psorting - as another strategy for optimising physical colocation of data.  (I don&#8217;t think you can get away with claiming a silent P in Pclustering, otherwise that would do as the more generic description).</p>
<p>Jonathan Lewis</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Anonymous</title>
		<link>http://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1057</link>
		<dc:creator>Anonymous</dc:creator>
		<pubDate>Fri, 19 May 2006 07:47:00 +0000</pubDate>
		<guid isPermaLink="false">https://oraclesponge.wordpress.com/2006/05/18/the-three-pillars-of-oracle-data-warehousing/#comment-1057</guid>
		<description>Don,

Why you think it is tough without MV's and rollups?.  Could you please give an example?</description>
		<content:encoded><![CDATA[<p>Don,</p>
<p>Why you think it is tough without MV&#8217;s and rollups?.  Could you please give an example?</p>
]]></content:encoded>
	</item>
</channel>
</rss>
