Mapplet Architecture

Figure 5-1 shows a high-level architectural view of the ArcGIS Server architecture as it relates to web-based applications. We can see from this image how the mapplet application will need to fit into the ArcGIS architecture.

Figure 5-1: . ArcGIS Server architecture diagram

We can think of the mapplet as a multilayer application. There are three major server objects deployed within the ArcGIS Server. The Map Server and Geocode Server are coarse-grained ArcObjects that are provided by the ArcGIS Server out of the box. These coarse-grained objects consist of hundreds of pre-existing COM objects that perform a multitude of GIS-related functionality. These server objects access map data stored in a variety of ESRI data formats. The other object is the Hotel Server, which wasn’t provided out of the box. The Hotel Server is a utility COM object created using Visual Basic code, and it was written to access the hotel data in an SQL Server database. The Server Object Manager (SOM) manages the life cycle of the Map Server and Geocode Server, and brokers the requests to these server objects.

The web application layer of the mapplet is written in C# and deployed as ASP.NET pages. The ASP.NET web pages use the ESRI-provided WebMap control for ASP.NET applications. For mapping- and geocoding-related functions, the WebMap control accesses the Map Server and Geocode Server objects. For hotel-related queries, the C# classes interact with the compiled Visual Basic object Hotel Server on the server side, which in turn makes calls to the hotel database.

Most of the classes you’ll see in the UML models that we present in this book are implemented as C# classes to encapsulate the business logic of the application. The ASP.NET code runs on the server side within the IIS web server, and generates the HTML and JavaScript code that is sent to the browser. Some data is cached in the web browser, and some functionality is implemented in the browser via HTML and JavaScript.