Question
SYNTEL
GB
Last activity: 23 Jun 2016 4:34 EDT
Reusable nested class structure design to support BIX
For one of my customer when the original Pega platform was built on v 6.1 SP2 one of the business requirements was MI function must have access to each and every information we are keeping in Pega Work DB. In absence of detailed MI requirements project team didn't give additional consideration for efficient MI while designing the Work data structure. In the last stage of project life cycle team configured the BIX Extract rules to pull the information into the staging DB to support business MI function outside Pega. Although the application was designed using reusable nested class structure across work types, the BIX tool generated 500+ DB tables in the destination database which created a big problem for Business MI team to understand the data.
Now the same scenario is repeating itself for other client and we want to prevent it in design stage. Is there any configuration available in Pega 7.x to map the nested class structure of Work type to the normalized DB tables so that the extracted information is immediately ready for reporting.
Many thanks in advance for your invaluable recommendations.
Pegasystems
IN
Now the same scenario is repeating itself for other client and we want to prevent it in design stage. Is there any configuration available in Pega 7.x to map the nested class structure of Work type to the normalized DB tables so that the extracted information is immediately ready for reporting.
It is not clear from this statement whether you are worried about multiple child classes of a framework class in the application or multiple classes being used for properties inside a given class in terms of Pagelist / Pagegroup structures. If it is the latter, Declare Index rules can be used to normalize the data and report definitions can join correctly to fetch the data as required.
I am still not clear on why you ended up with 500+ tables and what kind of application structure resulted in such a large number of takes.
SYNTEL
GB
Sorry, looks like there is a weakness in my language that's why you are not able to understand the issue I am trying to highlight.
My question is not related to building reporting functions with in Pega.
To simplify, when we use BIX to read the information from BLOB, is there a way that the extracted records (single or multiple per work item) can be mapped to normalized destination DB tables to prevent building an extra staging layer in between before that information can be leveraged for BI?
Pega -> Staging DB -> Business MI DB; can we prevent Staging DB via OOTB configuration?
Pegasystems
IN
In a typical ETL process, there is some transformation involved. BIX only takes care of the extraction part of the ETL process primarily because we know the structure and encoding of the BLOB. The following PDN article can give you some insight on how we envision the usage of BIX - https://pdn.pega.com/bix-design-models.
In BIX when you extract the data to a target database (instead of CSV or XML), we already provide an option to get the DDL which is normalized. The user can apply the DDL to their environment or, if they do already have the target tables in place, they can modify the BIX extract rule to specify the table name and column names appropriately for BIX to insert the data as part of extraction.
I suggest you read the BIX user guide to understand how BIX works - https://docs-previous.pega.com/bix-72-user-guide