Designing the System Structure › Choose a Client/Server Style › Distributed Processing › Where to Locate Logic with Distributed Processing Style

Where to Locate Logic with Distributed Processing Style

The distributed processing style often is the best solution for user task support, flexibility, and performance.

Try to design the end-to-end logic of the application so that it runs in one or more modules on the client.

When access is required to shared databases, locate the databases together with the business logic, integrity rules, and security rules on the server. Access to the data is always through the server logic layer not direct to the database. In practice this often means that the information processes reside on the server while the procedural logic is on the client.

You can consider accesses to the server rather like traditional accesses to sub-routines. The overall logic should be visible and testable on the client with calls to the server modules as necessary (by CA Gen dialog link flows). By adopting this style, you should find that your server modules can be reused by new applications without change.

Also, by locating such rules in one place (next to the data), changes to business rules can be more easily implemented and with minor disturbance to the using client applications.