k-nithin commented on issue #59618: URL: https://github.com/apache/airflow/issues/59618#issuecomment-3892931449
> agree that the current name last_automated_run is misleading given that it may not actually represent the most recent DAG run, especially considering scheduler behavior and Airflow’s distributed nature. > > As a first step, I’m planning to: > > Audit the call sites to better document the expectations vs. reality of what is being passed. > > Propose a clearer, semantics-accurate name that reflects “the run currently being processed by the scheduler” rather than “latest”. > > Separately, I’ll also look into whether the logic around updating next_dagrun can be made more robust to avoid regressions when newer runs are created concurrently. > > I’ll keep the initial PR focused and small (primarily naming + documentation), and follow up with logic changes if that’s acceptable. @dstandish I've audited the call sites for calculate_dagrun_date_fields and confirmed that last_automated_run does not consistently receive the latest run. It receives the run object currently being processed by the scheduler (which might be an older run in a backfill). As a first step, I've renamed the parameter to reference_run locally to better reflect its usage as a reference point for calculation rather than strictly the "latest" automated run. Does reference_run align with your thinking? I can proceed with this rename in a PR if you agree, or adjust if you prefer something like base_run. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
