Massive Parallel Task pool not destroyed after use.

[cianciosa]

Issue

When running the massive parallel as a sub workflow, the intent is that this would flow would be called multiple times in a loop. However when trying to do that, I get an error on the second iteration.

2023-07-13 12:58:43,221 fastran__ips_massive_parallel_10 ERROR    Uncaught Exception in component method.
Traceback (most recent call last):
  File "/global/common/software/atom/perlmutter/adaptive/conda/lib/python3.8/site-packages/ipsframework/component.py", line 139, in __run__
    retval = method(*self.args, **keywords)
  File "/global/common/software/atom/perlmutter/adaptive/conda/lib/python3.8/site-packages/fastran/driver/ips_massive_parallel.py", line 100, in step
    pool = services.create_task_pool('pool')
  File "/global/common/software/atom/perlmutter/adaptive/conda/lib/python3.8/site-packages/ipsframework/services.py", line 1871, in create_task_pool
    raise Exception('Error: Duplicate task pool name %s' % (task_pool_name))
Exception: Error: Duplicate task pool name pool
2023-07-13 12:58:43,257 workers__massive_parallel_runner_8 ERROR    Uncaught Exception in component method.

The cause of this is that the task pool is never destroyed after use.

Suggested fix

The code here

ips-fastran/src/fastran/driver/ips_massive_parallel.py

Line 161 in b1b7207

exit_status = services.get_finished_tasks('pool')

should be changed from

print('ret_val = ', ret_val)
exit_status = services.get_finished_tasks('pool')
print(exit_status)

to

print('ret_val = ', ret_val)
exit_status = services.get_finished_tasks('pool')
services. remove_task_pool('pool')
print(exit_status)