<?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"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
		>
<channel>
	<title>Comments on: So what&#8217;s Oracle&#8217;s MPP-aware optimizer and query execution plan story?</title>
	<atom:link href="http://www.dbms2.com/2008/09/25/oracls-mpp-optimizer-execution-plan/feed/" rel="self" type="application/rss+xml" />
	<link>http://www.dbms2.com/2008/09/25/oracls-mpp-optimizer-execution-plan/</link>
	<description>Choices in data management and analysis</description>
	<lastBuildDate>Sat, 30 Jan 2010 06:11:40 -0500</lastBuildDate>
	<generator>http://wordpress.org/?v=2.8.4</generator>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
		<item>
		<title>By: Oracle Database Machine and Exadata pricing: Part 2 &#124; DBMS2 -- DataBase Management System Services</title>
		<link>http://www.dbms2.com/2008/09/25/oracls-mpp-optimizer-execution-plan/comment-page-1/#comment-98190</link>
		<dc:creator>Oracle Database Machine and Exadata pricing: Part 2 &#124; DBMS2 -- DataBase Management System Services</dc:creator>
		<pubDate>Tue, 30 Sep 2008 07:06:01 +0000</pubDate>
		<guid isPermaLink="false">http://www.dbms2.com/?p=572#comment-98190</guid>
		<description>[...] look like good things to have no matter what. For example, as per a comment thread last week, what what Oracle packages as its no-added-charge optimizer is somewhat questionable, and many sites need a third-party product and/or Oracle&#8217;s extra-charge Tuning Pack to fill [...]</description>
		<content:encoded><![CDATA[<p>[...] look like good things to have no matter what. For example, as per a comment thread last week, what what Oracle packages as its no-added-charge optimizer is somewhat questionable, and many sites need a third-party product and/or Oracle&#8217;s extra-charge Tuning Pack to fill [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Curt Monash</title>
		<link>http://www.dbms2.com/2008/09/25/oracls-mpp-optimizer-execution-plan/comment-page-1/#comment-98167</link>
		<dc:creator>Curt Monash</dc:creator>
		<pubDate>Mon, 29 Sep 2008 21:12:41 +0000</pubDate>
		<guid isPermaLink="false">http://www.dbms2.com/?p=572#comment-98167</guid>
		<description>I was talking with an Oracle development manager on the DSS side, and he spoke favorably of Oracle&#039;s SQL Tuning chargeable option.

CAM</description>
		<content:encoded><![CDATA[<p>I was talking with an Oracle development manager on the DSS side, and he spoke favorably of Oracle&#8217;s SQL Tuning chargeable option.</p>
<p>CAM</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: David Aldridge</title>
		<link>http://www.dbms2.com/2008/09/25/oracls-mpp-optimizer-execution-plan/comment-page-1/#comment-98165</link>
		<dc:creator>David Aldridge</dc:creator>
		<pubDate>Mon, 29 Sep 2008 20:50:47 +0000</pubDate>
		<guid isPermaLink="false">http://www.dbms2.com/?p=572#comment-98165</guid>
		<description>&quot;there is at least one product you can buy that takes in an Oracle query, and puts out a new Oracle query with the same semantics, but which executes much faster&quot; ...

That would also be what the Oracle cost-based optimizer does in its query transformation step, wouldn&#039;t it? Star transformations, unnesting subqeries, transitive closure etc? 

With regard to getting queries right for Oracle, there may be some truth in Daniel&#039;s statement for complex OLTP queries, but BI/DW queries are generally not rocket science -- that is after all practically the raison d&#039;etre for the star schema. So I question the relevance here.</description>
		<content:encoded><![CDATA[<p>&#8220;there is at least one product you can buy that takes in an Oracle query, and puts out a new Oracle query with the same semantics, but which executes much faster&#8221; &#8230;</p>
<p>That would also be what the Oracle cost-based optimizer does in its query transformation step, wouldn&#8217;t it? Star transformations, unnesting subqeries, transitive closure etc? </p>
<p>With regard to getting queries right for Oracle, there may be some truth in Daniel&#8217;s statement for complex OLTP queries, but BI/DW queries are generally not rocket science &#8212; that is after all practically the raison d&#8217;etre for the star schema. So I question the relevance here.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Curt Monash</title>
		<link>http://www.dbms2.com/2008/09/25/oracls-mpp-optimizer-execution-plan/comment-page-1/#comment-98028</link>
		<dc:creator>Curt Monash</dc:creator>
		<pubDate>Fri, 26 Sep 2008 23:09:32 +0000</pubDate>
		<guid isPermaLink="false">http://www.dbms2.com/?p=572#comment-98028</guid>
		<description>Fair enough, Jay.  I was just curious as to what Dan thought was the one and only such product -- but upon rereading, he didn&#039;t assert there was only one.  Oops!

Thanks,

CAM</description>
		<content:encoded><![CDATA[<p>Fair enough, Jay.  I was just curious as to what Dan thought was the one and only such product &#8212; but upon rereading, he didn&#8217;t assert there was only one.  Oops!</p>
<p>Thanks,</p>
<p>CAM</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jay Jakosky</title>
		<link>http://www.dbms2.com/2008/09/25/oracls-mpp-optimizer-execution-plan/comment-page-1/#comment-98024</link>
		<dc:creator>Jay Jakosky</dc:creator>
		<pubDate>Fri, 26 Sep 2008 20:21:32 +0000</pubDate>
		<guid isPermaLink="false">http://www.dbms2.com/?p=572#comment-98024</guid>
		<description>I don&#039;t know which one Dan refers to but they&#039;re pretty common. There&#039;s one inside Toad that parses the query, creates candidates and asks the Oracle optimizer to rate those candidates. You can then make a selection and execute the queries to confirm Oracle&#039;s estimates.

I used the product briefly but regularly and it did an excellent job. And this is just an extra-value component to an already popular product. I can only imagine the effectiveness of other solutions.</description>
		<content:encoded><![CDATA[<p>I don&#8217;t know which one Dan refers to but they&#8217;re pretty common. There&#8217;s one inside Toad that parses the query, creates candidates and asks the Oracle optimizer to rate those candidates. You can then make a selection and execute the queries to confirm Oracle&#8217;s estimates.</p>
<p>I used the product briefly but regularly and it did an excellent job. And this is just an extra-value component to an already popular product. I can only imagine the effectiveness of other solutions.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Curt Monash</title>
		<link>http://www.dbms2.com/2008/09/25/oracls-mpp-optimizer-execution-plan/comment-page-1/#comment-97992</link>
		<dc:creator>Curt Monash</dc:creator>
		<pubDate>Fri, 26 Sep 2008 11:16:23 +0000</pubDate>
		<guid isPermaLink="false">http://www.dbms2.com/?p=572#comment-97992</guid>
		<description>Dan,

Which product are you referring to?

CAM</description>
		<content:encoded><![CDATA[<p>Dan,</p>
<p>Which product are you referring to?</p>
<p>CAM</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Daniel Weinreb</title>
		<link>http://www.dbms2.com/2008/09/25/oracls-mpp-optimizer-execution-plan/comment-page-1/#comment-97987</link>
		<dc:creator>Daniel Weinreb</dc:creator>
		<pubDate>Fri, 26 Sep 2008 10:13:38 +0000</pubDate>
		<guid isPermaLink="false">http://www.dbms2.com/?p=572#comment-97987</guid>
		<description>Pardon me, when you say &quot;Oracle&#039;s query optimizer&quot;, are you referring to the one in Oracle Database, the flagship RDBMS product?  If so, I beg to differ. It is very, very hard to write queries for Oracle that will produce good performance.  My own experience and that of my DB-savvy friends is that if you want your queries written right, you have to get a super-experienced Oracle wizard.  I only know one such person, although I&#039;m sure there are many more.  But they cost a lot (as consultants).

Also, there is at least one product you can buy that takes in an Oracle query, and puts out a new Oracle query with the same semantics, but which executes much faster: a third-party query optimizer!  (That&#039;s almost as stupid as those third-party uninstall utilities for Windows.)</description>
		<content:encoded><![CDATA[<p>Pardon me, when you say &#8220;Oracle&#8217;s query optimizer&#8221;, are you referring to the one in Oracle Database, the flagship RDBMS product?  If so, I beg to differ. It is very, very hard to write queries for Oracle that will produce good performance.  My own experience and that of my DB-savvy friends is that if you want your queries written right, you have to get a super-experienced Oracle wizard.  I only know one such person, although I&#8217;m sure there are many more.  But they cost a lot (as consultants).</p>
<p>Also, there is at least one product you can buy that takes in an Oracle query, and puts out a new Oracle query with the same semantics, but which executes much faster: a third-party query optimizer!  (That&#8217;s almost as stupid as those third-party uninstall utilities for Windows.)</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: gsalem</title>
		<link>http://www.dbms2.com/2008/09/25/oracls-mpp-optimizer-execution-plan/comment-page-1/#comment-97957</link>
		<dc:creator>gsalem</dc:creator>
		<pubDate>Thu, 25 Sep 2008 22:08:34 +0000</pubDate>
		<guid isPermaLink="false">http://www.dbms2.com/?p=572#comment-97957</guid>
		<description>Curt,
you&#039;re terribly mistaken: If you have RAC, you can make all your nodes work on the query, and all of them profit from exadata: they can scan/join/aggregate data and send it back to the node that your client is connected to.</description>
		<content:encoded><![CDATA[<p>Curt,<br />
you&#8217;re terribly mistaken: If you have RAC, you can make all your nodes work on the query, and all of them profit from exadata: they can scan/join/aggregate data and send it back to the node that your client is connected to.</p>
]]></content:encoded>
	</item>
</channel>
</rss>

<!-- Dynamic page generated in 0.227 seconds. -->
<!-- Cached page generated by WP-Super-Cache on 2010-01-31 05:16:25 -->
