Comment 25 for bug 1843708

> a quick question about the actual SRU. I don't have enough context as I am not a huge nova user and I totally agree this is a bug, but I just wanted to make sure if, by any chance, the current behaviour couldn't now be the 'expected behaviour' on bionic systems?

When executing the "openstack server rebuild" command (which, generally, destroys the current VM disk and rebuilds it from the image), the user is explicitly specifying this new key on the commandline. In the current version, this request is ignored.

This parameter is optional, so I think it should be reasonably safe to assume that if the user really wanted to use the old key, they would simply not specify a new key and would leave it absent instead.

While in theory someone could have accidentally relied on this behaviour, it seems highly unlikely to me and if they did, it would be having explicitly asked for the behaviour they will now get. So this feels OK to me.