/kind bug

What steps did you take and what happened:

1. Create a cluster using an image that doesn't exist.
2. The OpenStackCluster becomes ready (provided no bastion is used or bastion is using an existing image)
3. There is no visible progress and no explanation for what is wrong at the OpenStackMachine level. Worse, there is a misleading message that the OpenStackMachine is waiting for bootstrap data, which is not true.
4. There is no explanation on the OpenStackServer. It is just "not ready"

What did you expect to happen:

There should be a condition on the OpenStackServer that would propagate to the OpenStackMachine, with a message explaining that the image cannot be found. The OpenStackMachine should not say that it is waiting for bootstrap data when the data is available.

Anything else you would like to add:

This issue is also present in v0.10, just that there is no OpenStackServer involved then.

Environment:

• Cluster API Provider OpenStack version (Or git rev-parse HEAD if manually built): v0.11.2
• Cluster-API version: v1.8.5
• OpenStack version:
• Minikube/KIND version:
• Kubernetes version (use kubectl version):
• OS (e.g. from /etc/os-release):