Sorry for the late reply here — I really thought about this and environment variables were the only thing that sprung to mind as a “missing” feature of dbt Cloud.
There will be differences in deploying a project via dbt Cloud versus Airflow, but they aren’t really “features” per se. Things like:
- Jobs in dbt Cloud are configured through a UI, whereas tasks in Airflow are defined in your code (as far as I know)
- You can chain together complex workflows in Airflow, whereas dbt Cloud is purpose-built for running dbt jobs only.
In exchange though, you get a ton of things that make running dbt jobs much nicer, like:
- Access to the IDE, which might be a nicer development experience for team members
- Automatically generate documentation for your project, in a way that members of your account can access securely (with lots of read only accounts)
- Run your project automatically on pull requests
By the way, I’m going to unlist this topic — we’re trying to use Discourse for more long-form discussions, and “How to” posts, particularly those around dbt Core. Check out more details here.
Happy to continue the discussion on dbt Slack in the #dbt-cloud channel