openclimatefix/airflow-dags

AI Summary: This issue requests an upgrade of the 'solar-consumer' dependency to version 1.4.15 within the Airflow DAGs. The upgrade is intended to improve versioning information in logs. The specific locations for the upgrade are provided in the 'Possible Implementation' section.

AI Summary: This issue identifies a bug where the model name 'pvnet_da' is incorrectly used in a check within the forecast GSP DAG. The correct model name should be 'pvnet_day_ahead'. The fix involves changing this specific string in the provided Python file.

AI Summary: This issue requests an update to an informational message within an Airflow DAG. The message should be enhanced to explain how the forecast blend service handles Met Office data failures by incorporating backup models and provide a link to the Met Office service status page. This is a straightforward text update.

AI Summary: Modify the `consume-pvlive-dag.py` Airflow DAG to increase the historical data retrieval from PVLive to 12 hours by changing the `BACKFILL_HOURS` parameter. This addresses data gaps and ensures more complete historical data.

AI Summary: Modify the existing Slack callback function in the `airflow-dags` project to include country flags (UK, NL, India) based on the provided country parameter. This involves creating a new function `get_slack_message_callback_no_action_required` that returns a list of `send_slack_notification` calls, each including the appropriate flag emoji.

AI Summary: Add emoji country flags to existing Slack messages within specified Airflow DAGs (UK, Netherlands, and India) to improve message clarity and readability. This involves modifying the `on_failure_callback` functions in the relevant DAG files to include appropriate country flag emojis.