Hardware configuration

This topic describes the hardware on which CATALYST Enterprise is installed and configured, the concept of distributed processing and how the CATALYST Enterprise system achieves this, and the system interactions that occur to facilitate distributed processing.

The CATALYST Enterprise system involves multiple computers and distributed, parallel processing of aerial, satellite, and SAR imagery. The system has a central data repository to store the input and output data created by the various modules. It also includes computers designated as "processing nodes", which perform all the data processing. Typically, one of these nodes is designated as the controller and handles the task of distributing jobs and managing processing of jobs currently in the system.

Workstations are also typically configured to allow users to perform tasks such as manual collection of ground control points (GCP) and tie points (TP), quality assurance (QA), and mosaic preparation.

The following image illustrates a typical hardware configuration for CATALYST Enterprise.

Figure 1. Typical hardware configuration for deployed CATALYST Enterprise system

CATALYST Enterprise architecture

The following table describes the hardware components of a typical CATALYST Enterprise system.

Component	Description
Client computer	Computers assigned to the network to provide users with access to job results and perform QA.
Controller	Houses system database, webserver, and licensing server. Handles system requests and load balancing. Automatically and efficiently distributes jobs to multiple processing servers.
Processing servers	Processing nodes that perform the computational tasks assigned. Easily scalable. Responsible for CATALYST Enterprise tasks. Generally multicore, graphics processing unit (GPU), or both.
Data server	Supplies data High-speed switch (fiber) used between data server and processing servers.
QA workstations	Provides end-users with access to job results and perform QA tasks.