Comment 4 for bug 2031897

Christian's observation is right, the pipeline gets set up and both processes get started, but the code never looks at the returncode of the first process (p1) and only returns the returncode of the second process (p2): https://opendev.org/openstack/cinder/src/commit/3b9cc13a94668018d8e0aec2dc997262c9c654a8/cinder/backup/drivers/ceph.py#L616

A simple solution would be to check that p1 actually exited (whether successful or because of a timeout) and then stop the pipeline to avoid defunct processes. The subprocess module offers the wait() function for this: https://docs.python.org/3/library/subprocess.html#subprocess.Popen.wait