DNS propagation via CLI does not work if you have a subdomain and the bare domain on the same instance

[luckow]

We use a GitLab pipeline for our staging environments. Every rollout on the stage therefore works via the Cloudron CLI.
Everything is fine, if you only have a stage.example.org Domain on your Cloudron instance. The moment you added example.org as a second domain (or third ...) to the same instance, the CLI chooses example.org instead of stage.example.org for the DNS propagation.

      "icon": "file://.gitlab/cloudron-icon.png",
      "addons": {
        "localstorage": {},
        "mysql": {},
        "sendmail": {}
      },
      "dockerImage": "registry.example.org/vendor/project:129371"
    },
    "subdomain": "review-slug-hot19w.stage",
    "domain": "example.org",
    "portBindings": {},
    "fqdn": "review-slug-hot19w.stage.example.org"
  },
  "taskId": "28952"
}

Everything works as expected, if you use the mouse and the Cloudron dashboard. IMHO because you choose the right (in this case) sudomain stage.example.org

BTW: it's not a question, it's a bug report.

[girish]

@luckow good catch. In fact, the sort function usage was entirely wrong.

Fixed in https://git.cloudron.io/cloudron/cloudron-cli/-/commit/87dcb5f8cc8ae29f877a6f5b96c865ec48dd8846

I published a new CLI 5.5.0