systemctl stop should not fail the apply of the paunch command, when it attempts to stop & remove a container to be recreated again

[Bogdan Dobrelya]

While testing https://review.openstack.org/#/c/635161, I sent SIGSTOP to the running mysqld and reran paunch for the updated TRIPLEO_CONFIG_HASH. As expected, paunch waited for a 60 seconds for the graceful stop attempt of the container:

Deleting container (changed config_data): mysql
Stopping and disabling systemd service for tripleo_mysql

(waits for 60 seconds here...)

Job for tripleo_mysql.service canceled.
systemctl failed

^^ although, the container has been removed in the end, that failure was undesired for the paunch, when it attempts to re-apply a config change and meets an unresponsive container. Perhaps, the exception should be caught and suppressed in paunch for that a particular case.