mpgreg commented on issue #34206:
URL: https://github.com/apache/airflow/issues/34206#issuecomment-1837468294
This is something that will be necessary for MLops use cases.
@jscheffl if I'm understanding your suggestion correctly couldn't we just
use extras to store dynamic content?
I tried this and for some reason my downstream DAGs' tasks cannot read
extras from `triggering_dataset_events`. I can see the triggering event but
extra is empty.
```
[2023-12-03, 12:20:40 UTC] {logging_mixin.py:154} INFO - defaultdict(<class
'list'>, {'USAGE_FEATURES': [DatasetEvent(id=2, dataset_id=1, extra={},
source_task_id='build_features', source_dag_id='feature_engineering',
source_run_id='dataset_triggered__2023-12-03T12:20:31.045106+00:00',
source_map_index=-1)]})
```
even though the task in upstream DAG clearly put out the correct
pipeline_outlets
```
['airflow.datasets.Dataset@version=1(uri=USAGE_FEATURES,extra={\'cutoff_date\':
"\'2022-11-15\'", \'cost_categories\': [\'compute\', \'storage\']})']```
Does anyone know if it is supposed to work this way?
--
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: commits-unsubscr...@airflow.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
