The Open GIS Architecture

Michael A. Sokolewicz
Fraunhofer Institute for Computer Graphics
Graphic Information Systems Department
Wilhelminenstrasse 7
64283 Darmstadt, Germany

email: mike@igd.fhg.de URL: http://www.igd.fhg.de/www/igd-a5/people/mike/me.html

Keywords:: Graphic Information System; World Wide Web; Distributed Environment; Metadata

Information Technology as a whole is going through drastic changes. The explosion of interest in the Internet and the introduction and wide acceptance of the WWW are all signs of this. These changes offer the area of Graphic Information Systems (GIS) some large and not-to-be-ignored possibilities. In order to take full advantage of these possibilities, however, the current isolationist approach to GIS applications has to change. The Open GIS Architecture (OGA) was developed in order to bridge the gap between existing GIS databases.

Architecture advantages:

reduced duplication of effort, especially in data acquisition
access to larger body of information in distributed systems
database-software independence
hardware-base independence

At the heart of the OGA is a typical client-server scheme. In its simplest form, an OGA server can be used to service requests that come from some outside access point, such as a normal WWW Gateway.

In the middle of the figure above, however, is the pivot of the entire architecture, and the part that makes it open: the OGA Node-Server. A node-server is a server which is in touch with other servers. In this case, the node-server has contact with a constellation of OGA Leaf-Server/GIS Database Engine combinations. If a request comes in (either from the WWW client or some other OGA client) which is specific to Darmstadt, the node-server will know to route that request to the Darmstadt server. If a request for general information arrives at the node-server, it will be split or copied by the node-server into pieces specific for each leaf-server and sent where appropriate. The node-server then collects the query replies and returns them to the caller.

Each of the leaf-servers translates an incoming query into the native language of the database; the reply is then converted into a standardized format for transmission; upon receipt, the OGA Client converts the information into the format required at its end. This makes the actual process independent of the software/hardware used at either end of the chain.