array job size broken in newest version

[jerowe]

max_array_size does not work when also using commands_per_node

[jerowe]

This is because it will create the indices based on each task. array:1-20000:100 with 200 tasks * 100 commands per node, and this is not a valid slurm command.

Instead we will have to export a skip_index variable array:1-200, and invoke hpcrunner.pl execute_job --skip_index 100.