<?xml version="1.0" encoding="UTF-8" ?>
<?xml-stylesheet type="text/xsl" href="http://sqlblog.com/utility/FeedStylesheets/rss.xsl" media="screen"?><rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:slash="http://purl.org/rss/1.0/modules/slash/" xmlns:wfw="http://wellformedweb.org/CommentAPI/"><channel><title>SSAS - Speed up dimensions using a NULL default cube measure</title><link>http://sqlblog.com/blogs/jorg_klein/archive/2009/02/06/ssas-speed-up-dimensions-using-a-null-default-cube-measure.aspx</link><description>Recently I faced some problems with the performance of SSAS dimensions. The cube users were using a large dimension with more than 100.000 members that didn't perform well. They tried to add a leaf dimension member on an Excel 2007 pivot table. When dragging</description><dc:language>en</dc:language><generator>CommunityServer 2.1 SP2 (Build: 61129.1)</generator><item><title>re: SSAS - Speed up dimensions using a NULL default cube measure</title><link>http://sqlblog.com/blogs/jorg_klein/archive/2009/02/06/ssas-speed-up-dimensions-using-a-null-default-cube-measure.aspx#22622</link><pubDate>Thu, 25 Feb 2010 18:23:20 GMT</pubDate><guid isPermaLink="false">21093a07-8b3d-42db-8cbf-3350fcbf5496:22622</guid><dc:creator>Reed Jacobson</dc:creator><description>&lt;p&gt;Hi Jorg,&lt;/p&gt;
&lt;p&gt;This was cool. I've been doing a bunch of ROLAP (against Teradata) lately, and the &amp;quot;default&amp;quot; query is even slower because they're SQL. &lt;/p&gt;
&lt;p&gt;I did find that using 0 for Null as the value of the default measure made Excel work better: With Null as the value, Excel would hide all the axis members (unless you explicitly convert the PT to show empty cells). But with 0 as the value, Excel shows the members (even though it doesn't show the measure itself if it's not explicitly selected). &lt;/p&gt;
&lt;p&gt;Cheers,&lt;/p&gt;
&lt;p&gt;Reed &lt;/p&gt;
</description></item><item><title>re: SSAS - Speed up dimensions using a NULL default cube measure</title><link>http://sqlblog.com/blogs/jorg_klein/archive/2009/02/06/ssas-speed-up-dimensions-using-a-null-default-cube-measure.aspx#22630</link><pubDate>Fri, 26 Feb 2010 09:45:06 GMT</pubDate><guid isPermaLink="false">21093a07-8b3d-42db-8cbf-3350fcbf5496:22630</guid><dc:creator>jorg</dc:creator><description>&lt;p&gt;Hi Reed,&lt;/p&gt;
&lt;p&gt;You are absolutely right, using NULL as the value makes all the dimension members hidden when no measures are selected (they are not queried!). When you use 0, SSAS will return all dimension members which is a operation that requires more resources.&lt;/p&gt;
&lt;p&gt;In some cases its OK to use 0 but sometimes with really large dimensions this takes to much resources. So it really depends on the situation.&lt;/p&gt;
&lt;p&gt;Thanks for the useful reply!&lt;/p&gt;
&lt;p&gt;-Jorg&lt;/p&gt;
</description></item><item><title>re: SSAS - Speed up dimensions using a NULL default cube measure</title><link>http://sqlblog.com/blogs/jorg_klein/archive/2009/02/06/ssas-speed-up-dimensions-using-a-null-default-cube-measure.aspx#26173</link><pubDate>Sun, 13 Jun 2010 09:29:48 GMT</pubDate><guid isPermaLink="false">21093a07-8b3d-42db-8cbf-3350fcbf5496:26173</guid><dc:creator>Richard Madders</dc:creator><description>&lt;p&gt;Hi Jorg, interesting idea.&lt;/p&gt;
&lt;p&gt;I don't quite understand why you would ever run a query without a measure?&lt;/p&gt;
&lt;p&gt;Presumably, as long as you have one measure, the defaultmember measure is never used?&lt;/p&gt;
&lt;p&gt;Thanks,&lt;/p&gt;
&lt;p&gt;R&lt;/p&gt;
</description></item><item><title>re: SSAS - Speed up dimensions using a NULL default cube measure</title><link>http://sqlblog.com/blogs/jorg_klein/archive/2009/02/06/ssas-speed-up-dimensions-using-a-null-default-cube-measure.aspx#26219</link><pubDate>Wed, 16 Jun 2010 08:19:05 GMT</pubDate><guid isPermaLink="false">21093a07-8b3d-42db-8cbf-3350fcbf5496:26219</guid><dc:creator>jorg</dc:creator><description>&lt;p&gt;Hi Richard,&lt;/p&gt;
&lt;p&gt;The default cube measure is used every time you drag and drop a dimension member in your browser/excel without using a measure. For example when somebody wants to see sales by customer, in this case it is likely the person drags and drops the customers on the rows before selecting the sales amount measure. When you have large customer dimension this can cause a serious delay.&lt;/p&gt;
&lt;p&gt;So what you say is right, as long as you have one measure, the default cube measure is never used!&lt;/p&gt;
&lt;p&gt;-Jorg&lt;/p&gt;
</description></item></channel></rss>