The environments available for the project were:
After the process of refining the BPE program, it was necessary to develop the Clustor plan for BPE. This was a straight forward process involving the three steps of 'Preparator', 'Generator' and 'Dispatcher'. All three are most conveniently used through their TCL/TK (pronounced Tickle) interface.
The Preparator has a fairly easy to use 'Wizard' which deals with the areas of 'Parameter Description', 'Preprocessing Commands', 'Input Files', 'Substitution Files', 'User Command', 'Output Files' and 'Post Processing Commands'. This was very straight forward. After saving the plan, move to the Generator.
The Generator shows the details that are generated by Clustor from the user created plan.
The Dispatcher shows each of the jobs, and their status of waiting (to be run), running, done or failed. To begin running the jobs, use the Start button. Statistics are also available on jobs, servers and parameters. Details of Clustor status, Job status (the number waiting, executing and done) and Node Status (the number ready, busy or down) are all shown.
Also necessary is a clustor.nodes file. This tells Clustor the computers run the jobs on. This file can include a password or will prompt for one. A useful feature of '$nodename' (see example file) allows you to reuse the password entered for the first node.