This automated performance testing suite was created by Divya Jaganathan for
the Parallel Computing Summer Research Institute, 2018, with guidance from Mark
Petersen, LANL.

To run a performance test, start in any MPAS-Ocean run directory and add links to:
- ocean_model
- metis
- from this directory:
    submit_performance_test_to_queue.py
    performance_test.py
    plot_from_files.py

There are three ways to run this testing suite:

1. From a front end node, this command will submit a single job to the queue, of the size required.
./submit_performance_test_to_queue.py -M 128 -m 16 -r EC60to30

2. From a compute node, run this directly
./performance_test.py <Maximum Tasks> <Minimum Tasks> <machine_long_name> <cores_per_node> <resolution_name>

3. The above steps creates plots with just those tests. After running several jobs, 
   the data can be combined into a single plot with plot_from_files.py.  This file is case specific, and
   requires the user to alter lines by hand.
