Getting started with BeanCounter

- Log into the BeanCounter cluster on a unix command line type program.  Use and ssh. You can use putty if you dont have another program, its free!

- Type in your username and password when prompted.


Beancounter uses the LAM version of MPI. To start you need to have a file that contains a list of all nodes in the cluster. For example, you might have the file "bootschema" that contains the following:

Next start up lamboot which starts lamd, the LAM message server, on each machine in the cluster. Use the command: To compile a LAM MPI program use: mpic++ -o outputfile source.c++

This produces a binary file name "outputfile". To run the file use: mpirun C outputfile. In this case the program will be run on all available CPU's listed in the bootschema. You can also specify how many processors should be used using -c: mpirun -c 20 outputfile would run the program on the first 20 CPU's. Note that in our cluster, each node is a dual processor (except for the master) so you can generally specify twice as many processors as nodes without assigning multiple processes per processor.

After your program is finished, use lamhalt to finish your session:

lamhalt bootschema

- Continue with instructions to run Genetic Algorithm

- Continue with instructions to run HeatBug Simulation