We need to make sure that always only one OpenStackLightspeed instance is managed by the OpenStack Lightpseed operator. Otherwise, there might be some collisions when there are two instances that are in conflict (e.g., they have different values for the LLM URL). We can resolve this in two ways:
- Detect only
OpenStackLightspeed instances with specific name (e.g., openstack-lightspeed).
- Add detection logic for multiple
OpenStackLightspeed instances (either in the reconcile loop or in a webhook).