For BI Applications installations where eBusiness Suite is the source system. what should be value of prune days parameter in DAC for daily incremental run ? Out of the box , it is set to 30 days. If the incremental ETL is run daily, do we still need to set it to 30 days.
Prune days decide the number of days of data which it process during the ETL. Ex. If the last ETL load date is 14-Jan-2013 then with Prune 30 it would process data like last updated >= 13-Dec-2012.
The functional reason why we use Prune days is for address back-dated transactions and remote users (like remote users in Siebel). In our existing implementation we are using Prune days set to 7. It can be 1 also. I suggest you to talk with your Functional users and understand the remote users concept and back-dated transactions.
In back-dated transactions case, the last update column in eBusiness Suite tables should still reflect the date those transactions were uploaded. A low prune days value ( ie. 1-2 days ) should still pick those transactions ..right ? or am I missing the concept ?
You are correct on the back-dated transactions. The example I was quoting was especially in Siebel world they have remote users where the data is stored in locally and monthly twice or once they do a data sync. In this case the last updated date would not be equal less than the last etl load date. So to avoid that prune days are set to pick those records.