Question
What is the Best Practice for Data Layer for Integration?
What is the best practice for data layer for Integrations? Do we create a separate Data Class and Ruleset for each Integration, or do we only create the Integration Class and Rulesets for the Integration and reuse the application Data Class and Ruleset for data?
E.g. if we are consuming a SOAP service called 'SampleSOAP', do we create 'SampleSOAPInt' ruleset for the integration and 'SampleSOAP' ruleset for corresponding data rules and then plug this component into the 'SampleImplementation' application ruleset? Or do we only create 'SampleSOAPInt' ruleset for integration and use 'SampleImplementation' application ruleset for the corresponding data rules?
***Updated by Moderator: Marissa to add Categories***
I am actually referring to consuming a SOAP Service from Pega, i.e. Pega is the consumer. This would involve a Connect-SOAP and related rules.
Hi Sandeep,
Since you are talking about consuming a SOAP service, I assume you are refering to Service-SOAP here. In case of Services since the consumer is the external application or system, it makes sense to have a separate ruleset for housing the service-soap rule, package and also the associated activities and their classes. Since in most of the scenarios, the calling application would like to have the work object related information or process related information to be returned, this ruleset should have access to your Application FW ruleset assuming that you put most of your process rules in the FW layer as per best practices.
So in your case it would be SampleSOAPSvcs ruleset sitting on top of your SampleAppFW ruleset. The data translations can either be at the Svcs ruleset or you can reuse some of the existing mapping components from the FW layer as well. This largely depends on the kind of requirements you have for the service SOAP.
The scenario for SOAP Connectors are however entirely different. Hope this clarifies.