XMLA / MDX : Where to specify the database name ?

egau

<p>Hello ,</p><p> </p><p>I m actuatlly trying to use OLAP Cube provided with Microsoft SSAS solution. I've tried using example cube directly on internet and I m able to connect / request the cube and I got result in my dataset.</p><p> </p><p>The example I used only had 1 database instance and 2 cube inside.  In my real life example, I got 2 databases (db1 and db2) with several cube inside and I'm trying to use the cube "CUBE1". The problem is that when I try to execute an MDX request in a BIRT dataset , I get this error :</p><pre class="_prettyXprint">"The CUBE1 cube either does not exist or has not been processed."</pre><p>I know it exists and I know it has been processed because I m also connected using Excel to this cube and it works.</p><p> </p><p>That's why I want to know where and how should I specify the database name "db1" that contains my "CUBE1" ?</p><p> </p><p>It tried specifiying it in the XMLA adress :</p><pre class="_prettyXprint">http://localhost/xmla/msmdpump.dll/db1</pre><p>But it didn't work.</p><p> </p><p><strong>Bonus question : </strong>When I create a dataset based on an OLAP cube using MDX, the result in the dataset seems already formated to a crosstable layout (see screenshot attached). It means I can only use it as it is and can't connect it to a real cube and be able to play with it.</p><p>>> Am I wrong or is it only possible to get data from an OLAP Cube layed out like that ? Which means that using XMLA/MDX in birt is really limited because you can't use data later in a BIRT cube or graph ...</p><p> </p><p>Thanks in advance</p>

PuckPuck

<p>To connect to a non-default database needs to be done via the msmdpump configuration as part of IIS.  Inside of msmdpump.ini is a <ServerName> item.  Change this to your serverinstance</p><p> </p><p><ServerName>localhostdb2</ServerName></p><p> </p><p>Typically the msmdpump.ini file can be found in C:InetpubwwwrootOLAP</p><p> </p><p> </p><p> </p><p><strong>Bonus Answer:</strong> Cubes in BIRT can only be based on a dataset.  Data Sets themselves are 2 dimensional, while cubes are multi-dimensional.  To handle this, the connector will attempt to flatten the data, while maintianing some dimensionality, as you can see with the Column, and Row dimension names, pivoted into standard dataset columns.  To bring this back into a cube, would require further flattening of the data.  The connector can do this, with some of the options made available to you.</p><p> </p><p>I believe what you really want to do is have a BIRT cube source data directly from your MSAS cube.  This is not currently possible as described.</p>