Hi, I am a relatively new to JasperServer, but I've set it up and we're using it for our Data Warehouse with success on most of the MDX queries (in analysis views, created via the webinterface which runs on top of Tomcat. Our DB is PostgreSQL 9.0). I have the following query which really takes a *really long* time, and often results in out of heap space for tomcat, despite setting the heap space to higher and higher levels (right now 1.5GB). SELECT [Aff].Children ON COLUMNS, [District].Children ON ROWS FROM WHERE [Municipality].[FOOBAR] So basically Aff is one of my Dimensions, and in that dimension table itself there are about 20 records. That's my "column axis". Then, the District is a larger dimension table with maybe 25000 rows. This is where the problem comes in, because our MDX queries like these are fine when the dimension tables are small in size. The cube or the fact table Center consists of foreign keys to my dimensions e.g.: date_id aff_id district_id municipality_id (I also have a measure which is a basic "sales-count" type of measure) Now, I chose the slicer to be down to the municipality, which is then only relevant to maybe 5-6 rows in my fact table. However, when I bring up the MDX query in my analysis view, I get all 25000 rows from District, even though in practice only a small handful have numbers in them (the measure). This is really killing my Tomcat server as it's running with 100% CPU for many minutes just to come up with this. The browser struggles to display the 25000 row table. Note that the center-fact table has only about 20000 facts in it, so it should not be this computationally intensive.
