TREX is based on a client/server architecture. The client component is integrated into the application that uses the TREX functions, and allows communication with the TREX servers. The server component processes the requests; it indexes and classifies documents and answers search queries.
The client component is subdivided into the Java client and ABAP client.
The server component is subdivided into the following servers:
The graphic below shows the individual components and the communication between components: