anishgirianish commented on issue #55956:
URL: https://github.com/apache/airflow/issues/55956#issuecomment-3938201485

   Hi,@uranusjr, I'd like to work on this, if this is still open.
   
   On boolean vs integer, AssetDagRunQueue is keyed by (asset_id, 
target_dag_id) and deduplicates on insert, so it's a boolean signal, not an 
event counter. An integer limit would need the queue rearchitected to track 
individual events. Boolean "each event triggers its own run" vs "all grouped 
into one run" covers the use case and is simpler.
   
   For where the flag lives, DAG-level feels right. Asset has no scheduling 
parameters, and two DAGs depending on the same asset might want different 
grouping. Something like DAG(schedule=[asset], group_asset_events=False) 
defaulting to True for backward compat.
   
   Re partitions, these use separate infrastructure (AssetPartitionDagRun + 
PartitionedAssetKeyLog vs AssetDagRunQueue), and partitioned DAGs are excluded 
from the regular asset-triggered path. They shouldn't conflict.
   
   Let me know if this direction makes sense.


-- 
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]

Reply via email to