Cluster testing

From CajunBot Wiki

This script starts distributed testing on cluster nodes. When provided with an rndf file and a sample mdf file, the script generates checkpoints in an iterative manner and runs each sequence to determine the outcome (failure, success). The rndf is required to determine which checkpoints should be created and the mdf is desired for reference of speed limits.

A list of machines should be located in the testing folder for reference, by hostname or optionally username@hostname. Required ssh keys should be set up so that the script can log in without having to request a password.

Usage: cluster_testing <options>
        [-d #min]        Delay to check if all machines are still computing [1]
        [-l file]        Newline separated file of hosts to run tests on
                         [/home/cbserver/<user>/testing/machine_list.txt]
        [-m file]        MDF file (for getting speedlimits, not required)
        [-r file]        RNDF file (for getting checkpoints, required)

Additionally, here are the default values for some of the variables (some can only be changed by changing the script):

testing_dir=$HOME/testing                                                                                                                                                                               
rndf_template=$testing_dir/rndf_file.rndf                                                                                                                                                               
mdf_template=$testing_dir/mdf_template.mdf                                                                                                                                                              
testing_list=$testing_dir/checkpoint_list.txt                                                                                                                                                           
machine_list=$testing_dir/machine_list.txt                                                                                                                                                              
running=$testing_dir/running_on_                                                                                                                                                                        
delay=1