By Jeffrey McDaniel on Apr 07, 2014
To accommodate multiple data sources for codes and udfs requires a little coordination. Let's say you have 10 codes from your 1st data source and 10 codes from your 2nd data source that you would like to be used in P6 Analytics in OBI. Run the configuration utility for data source 1, choose your 10 codes using slots 1-10. Go to your \res folder and make a copy of the staretl.properties and save it to a different location (ds1staretl.properties). We're going to come back later to use this staretl.properties file for data source 1. Now for data source 2 lets go ahead and log into the configuration utility. Choose your codes, 1-10 don't really matter but choose codes 11-20 as the ones you want to represent this data source in OBI. Make a copy of this staretl.properties and save in a different location (ds2staretl.properties). Your are saving these backups so you can easily rebuild the codes list at a later time if runSubstitution or configuration utility are executed.
If you have a scenario where you have the same codes in data source 1 and data source 2 then this process is much easier. You would need to make the 'master copy' in just one of the data sources and then add it to the other properties files. If the data sources are truly unique and there are codes that don't exist in one of the other data sources you must follow the steps above.Do not worry about when the ETL process runs for data source 1 that it might not have the codes for data source 2 that are now defined in the properties file. This is adding it into the staging tables and mappings for OBI to use for the data associated with data source 2. When you try to add this code along with data from data source 1 it will not return any results just like any other project or activity that is not associated with a chosen code.