Utilizing Docker Swarm's built-in loadbalancer

Hey, I had a question about utilizing the docker swarm’s built-in load balancer/proxy to resolve nodes that are up and running.
I have read the post on https://github.com/cockroachdb/cockroach/issues/17133, which is also linked to Docker Swarm Scaling.

After reading them I am still a bit unsure about what is the cleanest way to deploy cockroach DB in the docker swarm stack in a way that would utilize swarm’s load balancer to resolve the nodes.

While it would be nice to be able to utilized certain swarm features like replicas, it is not a priority(at least for me). All I want is to either set up a cluster (by using multiple services like cockroachdb1, cockroachdb2, cockroachdb3, etc or by creating a single cockroach DB service with multiple replicas) so that it is resolvable behind a single DNS name.

(I am still relatively new to both docker swarm and cockroach DB so I might be missing some basic point)