I got started with the nats-operator helm chart. Everything works well there. As far as the streaming operator:
it would probably be best to use statefulsets for the actual cluster itself, as this will support volume claims much better. Otherwise, seems like you can only use a single node when running with DB or volume persistence.
would be awesome to turn this into its own helm chart as well, as that would definitely help with parameterization and other such issues (which I believe are still in the TODO.md in this repo).
I got started with the nats-operator helm chart. Everything works well there. As far as the streaming operator:
TODO.md
in this repo).Thanks for all the excellent work so far!!!