The more I think about this, the more having this data associated with stored Node Group Templates appears wrong.
For instance, foreign key constraints - we've had issues w/ node groups templates being deleted while they are still in use in cluster templates. Having IP Pools linked to node group templates creates a foreign key on an external service, and will result in more complex code to do validation - validation on creation, validation on deploy. The code to address failed validation on deploy is either complex or manual intensive.
The more I think about this, the more having this data associated with stored Node Group Templates appears wrong.
For instance, foreign key constraints - we've had issues w/ node groups templates being deleted while they are still in use in cluster templates. Having IP Pools linked to node group templates creates a foreign key on an external service, and will result in more complex code to do validation - validation on creation, validation on deploy. The code to address failed validation on deploy is either complex or manual intensive.
@ruhe where is the bp for this?