CDG – Community Datawarehouse Generator

CDG is a datawarehouse generator and the newest member of the Ctools family. Given the definition of dimensions that we want, CDG will randomize data within certain parameters and output 3 different things:

  • Database and table ddl for the fact table
  • A file with inserts for the fact table
  • Mondrian schema file to be used within pentaho
While most of the documentation mentions the usage within the scope of Pentaho there’s absolutely nothing that prevents the resulting database to be used in different contexts.


Several times, webdetailshad to prepare dummy data to feed dashboards / demonstrations. The traditional approach is always one of the following:

  1. Use data from steel-wheels (the pentaho sample datawarehouse)
  2. Build specific transformation to return static pieces of data

Both have severe disadvantages. With number one, we found out that not only the datawarehouse has severe inconsistencies, worst of all the end customers most of the time can’t transpose to their business terms like trains and cars.
For number 2, while we can build specific sets, it ends up being time consuming and since it’s not a full datawarehouse we can’t show all the abilities of the solution we’re demonstrating.
So we decided to build CDG – a datawarehouse generator that we can quickly use to build scenarios where the end consumer can feel comfortable with, written with Kettle


Version 1 of CDG is called by editing a kettle transformation. In the source you’ll find a src directory but that is our “development area”. The code you need to use is inside kettle.
Once you open the transformation in kettle/generateDW.ktr you’ll see the following:
CDG kettle transformation
There are only 2 things that need to be changed:

  1. In transformation properties you defined the name of the database and fact table
  2. In Dimension Info you configure the transformation parameters

Change the parameters you want (or just run with the default), run the transformation, and 3 files will appear in the output directory: database and table ddl, insert scripts and Mondrian schema file.

Configuring Totals

Inside the Dimension Infostep you’ll find a mention to the total:


This will be approximate total for all the breakdowns. We need to specify something within the order of magnitude of what we’re trying to show. CDG will then take that value and randomize it.

Configuring Dimensions

In the same file you configure the dimensions. You can have as much as you want, just paying attention to the fact that if you use a lot of dimensions / high cardinality we can quickly end up with a huge database. While there’s nothing particularly wrong with that, it’s then up to you to do specific optimizations like indexes or even aggregate tables. That’s outside the scope of CDG.
Here’s a sample dimension definition:

var countries = [ {countryName: "Italianos", proportion: 30}, {countryName: "Portugueses", proportion: 18}, {countryName: "Alemães", proportion: 12}, {countryName: "Espanhóis", proportion: 10}, {countryName: "Japoneses", proportion: 15}, {countryName: "Coreanos", proportion: 8}, {countryName: "Chineses", proportion: 3}, {countryName: "Outros", proportion: 5} ]; 

The sample provided in CDG is in Portuguese to specifically test character encoding support. The generated files are in UTF-8 and we recommend always using utf-8 in the database too
By defining this object, CDG will create a dimension with 8 members and one level called countryName. You could have other properties in there and CDG would create a mondrian schema with different levels. The provided example has only one.
There’s a special property in there called proportion. That will be used by CDG to do the breakdown of the total. In the example, roughly 30% of the total will be assigned to Italians and so on. In all aspects of the code there’s a random factor in place.

Configuring Date Dimension

The date dimension is always a specific case, since most of the times acts as a snapshot dimension.
Since configuring all possible members of this dimension would be very time consuming, we provide an utility function that generates all the dates between 2000 and 2012 down to the month. This is standard javascript, so feel free to change this function either to change the date range, month names or even adding the day level (be aware that adding the day level will substantially increase the number of values in the fact table)

/* CONFIGURE THE DATE DIMENSION. */ var dateDim = []; var months = [ [1,"Jan","Janeiro"], [2,"Fev","Fevereiro"], [3,"Mar","Março"], [4,"Abr","Abril"], [5,"Mai","Maio"], [6,"Jun","Junho"], [7,"Jul","Julho"], [8,"Ago","Agosto"], [9,"Set","Setembro"], [10,"Out","Outubro"], [11,"Nov","Novembro"], [12,"Dec","Dezembro"] ]; range(2000,2012).map(function(year){ range(0,12).map(function(month){ var m = months[month]; dateDim.push( {"year":year ,"monthNo": m[0], "monthAbbrev":m[1], "monthDesc": m[2]} ); }); }); ; 

Final configuration

In the end of the script there’s the final configuration that will be used by CDG:

/* MAKE THE FINAL CONFIGURATION. DIMENSIONS CAN EITHER BE SNAPSHOT OR REGULAR BREAKDOWNS */ var outputArray = [ {name: "Date", dimension: dateDim, toBreakdown: false, increment: 0.05 }, {name: "Provices", dimension: provinces, toBreakdown: true }, {name: "Countries", dimension: countries, toBreakdown: true }, {name: "Dates", dimension: gender, toBreakdown: true } ] 

In here we define the names and types of the dimension. The property of toBreakdown should be true for normal dimensions and false for snapshot dimensions. If it’s a snapshot dimension, you need to specify the increment property. This value of 0.05 basically means that we’ll have roughly 5% increase each month.
You can then run the transformation.


After running the transformation, we get this output:

cdg/kettle/output ├── cdgsample.ddl ├── cdgsample.mondrian.xml `── cdgsample.sql 
  1. A ddl file to create the database and the table
  2. A file with sql inserts to populate the datawarehouse
  3. A mondrian schema file to use within mondrian / import to pentaho

Schema workbench


The result, after declaring this new datasource and registring the cube in mondrian, is a new cube that we can use.

Issues, bugs and feature requests

In order to report bugs, issues or feature requests, please use the Webdetails CDG Project Page


CDG is licensed under the MPLv2 license.


Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s