4.9 Parallelism using Distributed Processing

Data Explorer provides the capability of distributing the execution of a visual program or a program generated using the scripting language over multiple workstations on a network. Distributing the execution provides parallelism and enhanced resource utilization. Parallelism is achieved by the simultaneous execution of different portions of the visualization on each of the workstations. Enhanced resource utilization can be achieved, for example, by assigning computationally intensive portions of the visualization to the more powerful workstations, or transformation and realization functions that are applied to data located remotely can be distributed to the remote workstations, reducing the amount of data transfer.

Distributed processing is achieved in two ways: using "outboard" modules or placing groups of tools into "execution groups." These two methods can be used independently or in combination. An outboard module is a user-written module controlled by the Data Explorer executive but is external to the Data Explorer server program. They can be invoked from either a visual program or a script program. Execution groups are a set of tools that can be assigned to a workstation. Once groups are created, they can be assigned to the workstations over which the visualization is to be distributed. More than one group can be assigned to each workstation. (See also 9.1 , "Using Distributed Computation".)