Top-Down Design 
When you apply the top-down design approach, you first define the overall integration scenario at a high level, in particular, its separation into individual process components and how these are connected with each other. This gives you a bird’s eye view of the integration. You do this in a model, as we will show below. Based on the model, you then specify the other relevant integration content objects like interfaces, data types, and mappings in more detail.
These are the main tasks when you design integration content top-down:
Design Task |
Design Objects |
Modeling the integration-relevant aspects of a cross-component process – or how process components interact with each other |
Process models You define a process model to start with (for example, a process integration scenario). Based on the model, you create the corresponding integration content that specifies the integration in more detail (interface objects, mapping objects, and communication channel templates). |
Specifying how one process component (or application component) exchanges messages with another |
Interface objects There are different interface object types available that describe these aspects – from the communication mode of message exchange down to the detailed data structure of a message. |
Specifying how data structures are transformed into each other |
Mapping objects |
Pre-configuring connectivity/adapters |
Communication channel templates You use communication channel templates to specify those adapter settings for connectivity that are already known at design time. |
After having finished the design tasks in the ES Repository, you need to generate proxies for the interface objects in the corresponding back-end systems The interface objects defined in the ES Repository are merely metadata that are independent from any programming language. Using proxy generation, you convert non-language-specific interface descriptions into executable interfaces.