<?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: SSIS &#8211; Slowly Changing Dimensions with Checksum</title>
	<atom:link href="http://blog.stevienova.com/2008/11/22/ssis-slowly-changing-dimensions-with-checksum/feed/" rel="self" type="application/rss+xml" />
	<link>http://blog.stevienova.com/2008/11/22/ssis-slowly-changing-dimensions-with-checksum/</link>
	<description>Homepage of Steve Novoselac</description>
	<lastBuildDate>Fri, 12 Mar 2010 15:59:49 -0800</lastBuildDate>
	<generator>http://wordpress.org/?v=2.9.2</generator>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
		<item>
		<title>By: Maintaining a Type 1 Slowly Changing Dimension (SCD) using T-SQL &#171; laqua dot us</title>
		<link>http://blog.stevienova.com/2008/11/22/ssis-slowly-changing-dimensions-with-checksum/comment-page-1/#comment-160522</link>
		<dc:creator>Maintaining a Type 1 Slowly Changing Dimension (SCD) using T-SQL &#171; laqua dot us</dc:creator>
		<pubDate>Tue, 24 Nov 2009 21:42:13 +0000</pubDate>
		<guid isPermaLink="false">http://blog.stevienova.com/?p=1010#comment-160522</guid>
		<description>[...] of the SCD wizard to speed things up, maybe even some fancy checksum voodoo for the updates (see http://blog.stevienova.com/2008/11/22/ssis-slowly-changing-dimensions-with-checksum/ for an example). Then after thinking about it a little more &#8211; why are we sending a million [...]</description>
		<content:encoded><![CDATA[<p>[...] of the SCD wizard to speed things up, maybe even some fancy checksum voodoo for the updates (see <a href="http://blog.stevienova.com/2008/11/22/ssis-slowly-changing-dimensions-with-checksum/" rel="nofollow">http://blog.stevienova.com/2008/11/22/ssis-slowly-changing-dimensions-with-checksum/</a> for an example). Then after thinking about it a little more &#8211; why are we sending a million [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Todd McDermid</title>
		<link>http://blog.stevienova.com/2008/11/22/ssis-slowly-changing-dimensions-with-checksum/comment-page-1/#comment-160485</link>
		<dc:creator>Todd McDermid</dc:creator>
		<pubDate>Mon, 28 Sep 2009 20:58:13 +0000</pubDate>
		<guid isPermaLink="false">http://blog.stevienova.com/?p=1010#comment-160485</guid>
		<description>You may also want to try out a single-component SCD: the Kimball Method SCD (&lt;a href=&quot;http://kimballscd.codeplex.com&quot; rel=&quot;nofollow&quot;&gt;http://kimballscd.codeplex.com&lt;/a&gt;).  I&#039;ve experienced a hundred-fold increase in performance, and I haven&#039;t added checksumming to it yet, although other users have, and report even higher performance gains...</description>
		<content:encoded><![CDATA[<p>You may also want to try out a single-component SCD: the Kimball Method SCD (<a href="http://kimballscd.codeplex.com" rel="nofollow">http://kimballscd.codeplex.com</a>).  I&#39;ve experienced a hundred-fold increase in performance, and I haven&#39;t added checksumming to it yet, although other users have, and report even higher performance gains&#8230;</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Maintaining a Type 1 Slowly Changing Dimension (SCD) using T-SQL &#124; tim laqua dot com</title>
		<link>http://blog.stevienova.com/2008/11/22/ssis-slowly-changing-dimensions-with-checksum/comment-page-1/#comment-156464</link>
		<dc:creator>Maintaining a Type 1 Slowly Changing Dimension (SCD) using T-SQL &#124; tim laqua dot com</dc:creator>
		<pubDate>Sat, 23 May 2009 18:16:08 +0000</pubDate>
		<guid isPermaLink="false">http://blog.stevienova.com/?p=1010#comment-156464</guid>
		<description>[...] of the SCD wizard to speed things up, maybe even some fancy checksum voodoo for the updates (see http://blog.stevienova.com/2008/11/22/ssis-slowly-changing-dimensions-with-checksum/ for an example). Then after thinking about it a little more - why are we sending a million rows [...]</description>
		<content:encoded><![CDATA[<p>[...] of the SCD wizard to speed things up, maybe even some fancy checksum voodoo for the updates (see <a href="http://blog.stevienova.com/2008/11/22/ssis-slowly-changing-dimensions-with-checksum/" rel="nofollow">http://blog.stevienova.com/2008/11/22/ssis-slowly-changing-dimensions-with-checksum/</a> for an example). Then after thinking about it a little more &#8211; why are we sending a million rows [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: John Shanks</title>
		<link>http://blog.stevienova.com/2008/11/22/ssis-slowly-changing-dimensions-with-checksum/comment-page-1/#comment-143104</link>
		<dc:creator>John Shanks</dc:creator>
		<pubDate>Thu, 26 Feb 2009 22:25:00 +0000</pubDate>
		<guid isPermaLink="false">http://blog.stevienova.com/?p=1010#comment-143104</guid>
		<description>I would have to agree with you Steve that the SCD functionality, in SSIS is slow.  I saw even worse performance than you described but switched to using the Checksum, Lookup, Conditional Split approach and the performance improved dramatically.  I saw 200 thousand rows go from 10 minutes to 2 minutes.</description>
		<content:encoded><![CDATA[<p>I would have to agree with you Steve that the SCD functionality, in SSIS is slow.  I saw even worse performance than you described but switched to using the Checksum, Lookup, Conditional Split approach and the performance improved dramatically.  I saw 200 thousand rows go from 10 minutes to 2 minutes.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Marco</title>
		<link>http://blog.stevienova.com/2008/11/22/ssis-slowly-changing-dimensions-with-checksum/comment-page-1/#comment-143094</link>
		<dc:creator>Marco</dc:creator>
		<pubDate>Thu, 26 Feb 2009 21:46:57 +0000</pubDate>
		<guid isPermaLink="false">http://blog.stevienova.com/?p=1010#comment-143094</guid>
		<description>Very nice write up Steve.  I just recently dove into SSIS and SCD ... I must say I&#039;m very impressed w/the technology as a whole.  A couple points I wanted to relay, please provide feedback when convenient:

1)  Regarding your performance figures:  I&#039;ve resorted to using the SCD Wizard.  We&#039;re getting completion times of around 10 - 20 minutes for target dimension tables that are 300+ million rows, now closely approaching 400 million rows.  Each write (the SCD packages run every night @ 1 am) handles updating about 8k - 12k rows.  This is on a 8 cpu/32 gig machine.  What hardware profile have you based your performance figures on?

2)  Regarding the lookup transformation, you&#039;ve explicitly created this in your solution...and I see you&#039;re explicitly handling new rows that don&#039;t exist in your destination table.  Is a similar transformation created &quot;behind the scenes&quot; when using the SCD wizard (I didn&#039;t see the Lookup Transformation operation box in any of my SCD packages).  One of my SCD packages has recently started failing, citing the error &quot;The Lookup transformation failed to retrieve any rows.&quot; It looks like existing records are updated fine ... so do I need to handle new records that don&#039;t exist in the destination table explicitly? ... do you have any experience w/treating such an error?</description>
		<content:encoded><![CDATA[<p>Very nice write up Steve.  I just recently dove into SSIS and SCD &#8230; I must say I&#8217;m very impressed w/the technology as a whole.  A couple points I wanted to relay, please provide feedback when convenient:</p>
<p>1)  Regarding your performance figures:  I&#8217;ve resorted to using the SCD Wizard.  We&#8217;re getting completion times of around 10 &#8211; 20 minutes for target dimension tables that are 300+ million rows, now closely approaching 400 million rows.  Each write (the SCD packages run every night @ 1 am) handles updating about 8k &#8211; 12k rows.  This is on a 8 cpu/32 gig machine.  What hardware profile have you based your performance figures on?</p>
<p>2)  Regarding the lookup transformation, you&#8217;ve explicitly created this in your solution&#8230;and I see you&#8217;re explicitly handling new rows that don&#8217;t exist in your destination table.  Is a similar transformation created &#8220;behind the scenes&#8221; when using the SCD wizard (I didn&#8217;t see the Lookup Transformation operation box in any of my SCD packages).  One of my SCD packages has recently started failing, citing the error &#8220;The Lookup transformation failed to retrieve any rows.&#8221; It looks like existing records are updated fine &#8230; so do I need to handle new records that don&#8217;t exist in the destination table explicitly? &#8230; do you have any experience w/treating such an error?</p>
]]></content:encoded>
	</item>
</channel>
</rss>
