Request generation might end too soon

[phip123]

Currently the requests are executed in a sequential way, while the last ClientGroup will wait till it finishes all requests.

galileo-experiments/galileoexperiments/experiment/scenario/run.py

Lines 102 to 108 in b6a6f44

	def requests():
	for idx, group in enumerate(client_groups):
	if idx == len(client_groups) - 1:
	# TODO the wait here might return sooner than other profiles => implement solution that waits for all
	group[2].request(ia=('prerecorded', 'ran')).wait()
	else:
	group[2].request(ia=('prerecorded', 'ran'))

In case the last ClientGroup has a profile much shorter than the others, the experiment will end too soon.

Solution: Dispatch all requests concurrently and wait for all of them to finish.