﻿<?xml version='1.0' encoding='UTF-8'?><rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/"><channel><title>SQLServerCentral / Data Warehousing / Analysis Services  / Multiple dimensions from 1 table or 1 dimension with multiple hierarchies / Latest Posts</title><generator>InstantForum.NET v2.9.0</generator><description>SQLServerCentral</description><link>http://www.sqlservercentral.com/Forums/</link><webMaster>notifications@sqlservercentral.com</webMaster><lastBuildDate>Fri, 24 May 2013 17:42:48 GMT</lastBuildDate><ttl>20</ttl><item><title>RE: Multiple dimensions from 1 table or 1 dimension with multiple hierarchies</title><link>http://www.sqlservercentral.com/Forums/Topic1385008-17-1.aspx</link><description>I used my time dimension to create 2 different hierarchies, the first consisting of 4 level, the second 3 level.  In both cases the toplevel is year.  When adding attribute relations 2 relations now converge on the year:________________--&amp;gt;  monthlevel1 --&amp;gt; quarterlevel1date(key)  ________________________________________________   ---&amp;gt;  year________________----------&amp;gt;   monthlevel2No warnings are displayed, but after processing the dimension and browsing the members the 2nd hierarchy makes no sense at all with members randomly missing.Do I need to add a second year column to avoid the convergion (on to year)Should I create seperate dimensions each with its own hierarchy?</description><pubDate>Fri, 16 Nov 2012 13:37:36 GMT</pubDate><dc:creator>blom0344</dc:creator></item><item><title>RE: Multiple dimensions from 1 table or 1 dimension with multiple hierarchies</title><link>http://www.sqlservercentral.com/Forums/Topic1385008-17-1.aspx</link><description>I would go with the single dimension with multiple hierarchies to save processing time,</description><pubDate>Fri, 16 Nov 2012 04:17:28 GMT</pubDate><dc:creator>Scott Murray-240410</dc:creator></item><item><title>Multiple dimensions from 1 table or 1 dimension with multiple hierarchies</title><link>http://www.sqlservercentral.com/Forums/Topic1385008-17-1.aspx</link><description>Being accustomed to Cognos DMR (ROLAP based technology) I am trying to wrap my head around SSAS.My pilot consists of a facttable (Cashflow) with just 2 FK's.FK1 relates to the time dimension. DMR expects a physical calendertable to be used, SSAS allows for generating the time dimension by creating it within SSAS.  Very NiceFK2 relates to 1 dimension table that stores roughly 60 attributes. About half of these attributes allow for grouping by in aggregates.  In DMR we build seperate dimensions. Such a dimension consists - for instance - of 2 high level grouping codes to be used as context filters or high level aggregate groups.I cannot find any best SSAS practice on this..Should I create 1 single dimension and build numerous hierarchies mimicing the structure we used in Cognos DMR?Or should I define multiple dimensions against the same dimension table , each with its own hierarchy?Technically the dimension table stores only natural hierarchiesAs DMR is strictly ROLAP generating a SQL against the relational database without storing pre-aggregates both solution would work.  I am not sure whether SSAS supports this..</description><pubDate>Thu, 15 Nov 2012 01:58:19 GMT</pubDate><dc:creator>blom0344</dc:creator></item></channel></rss>