Selecting hardware for Dispatcher Server components

Selecting hardware for the module and translators The module is where all the intense computational process of translations occur. It must be sized sufficiently to handle the expected translation throughput. The module hardware requirement is governed by the translator hardware requirements, number of translators installed, and the number of maximum potential translation tasks (number of users simultaneously submitting tasks). An enterprise class server machine is highly recommended for the module, as these machines have the ability to scale the number of CPUs and memory much more than a normal workstation. Selecting hardware for the scheduler The scheduler provides the communication from Teamcenter clients and manages the queue and job distribution to the modules. The scheduler is only a pipeline mechanism to manage the translation requests used by modules. The scheduler requires minimum machine resource to run; however, it does hold an object in memory for each translation request that it is managing. Selecting hardware for the dispatcher client The dispatcher client sends the translation task from Teamcenter to Dispatcher Server. Only one dispatcher client is required per Teamcenter server installation. The dispatcher client for Teamcenter manages the data extraction and data loading into Teamcenter. While the data extracting and loading are not computationally intensive, it does require a machine with communication pipeline to the server as well as enough disk space and memory needed for the operations. Although not required, the recommended approach is to have the Dispatcher client running on the same machine as the Teamcenter server.