dstandish commented on PR #54284:
URL: https://github.com/apache/airflow/pull/54284#issuecomment-3175116567

   > Don't you dare looking at (max_tis+1)'th task unless you schedule the 
max_tis most prioritized ones, except tasks that are starved at the moment.
   
   The way you phrase this here, it feels like this approach is faithful to 
that.  If the TI comes out in the query, and ends up being not queuable, then 
it falls into that "starved" category.  That's why nulls are first.  Do you see 
it differently?
   
   Priority is already a pretty weak feature, it seems, because it can't look 
ahead to what's coming -- it's only evaluated per scheduler loop.


-- 
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