The default node pools cause trouble with managing the cluster. It seems like a better architecture to manage the node pools explicitly. Does it make sense to use the remove_default_node_pool param...