Comment 2 for bug 1460112

Oh nevermind, it is a cast because there is a host and the service group API is up, so it sends the delete request to the compute manager:

http://git.openstack.org/cgit/openstack/nova/tree/nova/compute/api.py?id=2015.1.0#n1775

which eventually gets us here:

http://git.openstack.org/cgit/openstack/nova/tree/nova/compute/manager.py?id=2015.1.0#n2638

So it makes sense that we have a race here since grenade isn't waiting for the instance to be gone before it tries to delete the security group which is still associated with the instance.

We could add a 1 second sleep to grenade and it would probably fix this, or loop on a server get until we get a 404.