Silly question - what happens when i upload a new schema?
First post and been using Pentaho for a couple of days.
History - we've had 4 developers integrating data with Pentaho and now it updates daily. They've also created one cube so far that I can use for Analysis. I have been however using the Data Source Wizard. While this is seems to limit some of the configuration, it does seem fairly rapid to design a new cube.
Question - What happens after I create a cube design? Does Pentaho begin a process to build out the values of the cube there and then? Is there a waiting process before I can use this cube?
Sometimes it seems I can use the cube straight away and other times it just never seems to function when I add measures to the report (refreshes endlessly then says too much data)
Is this because I have created a rough design using too much data (this is millions of transactions) and havent optimized it? Can this type of data be optimized through better design or will it always be slow with so many indivudal transactions?
Loving Pentaho so far, despite some frustrations!
Your intuition is correct. DSW creates fairly simple and rough schema definitions. I would recommend spending some time looking into schema design to come up with something more tailored (http://mondrian.pentaho.com/documentation/schema.php). Also look into optimizing performance with aggregate tables.
There is no "waiting process" after the cube has been published. There's nothing to be built or loaded. It's possible that the slowness you're seeing is the result of some data not being cached, and subsequent reports are able to load some of that same pre-cached data. Again, looking into performance optimizations can reduce the wait.
Good luck and don't hesitate to post questions as you run into new frustrations.