Show TOC

Model-based ConfigurationLocate this document in the navigation structure

Use

Using a process model from the ES Repository as configuration template is the best way to use synergies between design and configuration time activities. This allows a semi-automatic configuration which considerably reduces configuration time, costs, and effort. In this section we show how this works.

At design time (in a process model), you specify relations between process components. In a process components interaction model in particular you specify which interfaces, mappings, and possible communication channel templates determine the interaction between two process components in detail.

At configuration time, you map the interaction of process components to system-to-system-interactions. In the preceding sections, we explained how you do this manually: You manually defined the routing and the other processing details for each incoming message.

However, assume that at configuration time you use a process components interaction model as configuration template and first assign the involved process components to communication components (systems). Then you have all the information you need to derive all sender system/receiver system relations that are relevant for the interaction between the process components. In other words, you know the object keys of all the relevant receiver determinations, interface determinations, and sender and receiver agreements as explained under Object Key in Configuration Objects .

At this point, the model configurator comes into play: Based on assignments between process components and communication components (systems) for a specific model, the relevant receiver determinations, interface determinations, and sender and receiver agreements are calculated and automatically generated. This means that the tool calculates the object key, and checks if there already is a configuration object with this exact key. Depending on the result, it either creates the object or re-uses an existing one. All generated configuration objects are then grouped in a configuration scenario which is a grouping entity in the Integration Directory.

The following figure highlights how the design and configuration time entities are related to each other, how the model configurator comes into play, and how a receiver determination is generated out of the setting (as an example):

Figure 1: How the Model Configurator Evaluates a Receiver Determination for the Given Example

Using the model configurator saves you the trouble of manually creating all configuration objects. If you configure the integration for large system landscapes, it can quickly turn into a nightmare to manually find out all relevant configuration object keys.

After generating the objects, the configuration objects generated have to be further specified manually by adding those parts that cannot be automatically determined, for example, routing conditions or specific security settings in sender and receiver agreements.

You can use the model configurator with the following model types as input: