There is a race condition between the operator started and the crd definition to be available
When you deploy the operator you wait for the operator to be ready.
However, when you are running in some kubernetes not very fast (like a single user instance) there is a small amount of time from the moment you see the operator running and when the crd definitoin is ready
As a result when you create the crd instance, the cretion fails.
In this case you should retry a few times and wait a bit. The definitino will be available shortly but you cannot assume it is ready when the operator is running.
There is a race condition between the operator started and the crd definition to be available
When you deploy the operator you wait for the operator to be ready. However, when you are running in some kubernetes not very fast (like a single user instance) there is a small amount of time from the moment you see the operator running and when the crd definitoin is ready
As a result when you create the crd instance, the cretion fails. In this case you should retry a few times and wait a bit. The definitino will be available shortly but you cannot assume it is ready when the operator is running.