I have a service which generates DAG according to a template and uploads it to a git repo. There's an Airflow container, using that repo as its DAGs source. How can I, as efficiently as possible, run one-time task that runs after the very first successful DagRun of each generated one (kind of setup task for that DAG)? What I have tried: For each DAG, create another one that is scheduled on dataset which the original one outlet and then pausing itself. However, this DAG can never be deleted as it's defined in the same file as the generated one. Note that I would like to avoid separating it into two different files. Continue reading...
