uranusjr commented on issue #31743:
URL: https://github.com/apache/airflow/issues/31743#issuecomment-1600433143

   In general, you need to restart services when you change things like this. 
If you modify a plugin, for example, you need to restart the scheduler and/or 
the webserver. This is no different; if you change a trigger, you need to 
restart the triggerer service. Airflow’s service processes are all built with 
this design in mind, restarting is an expected operation, should not cause any 
data breakage, and can be performed liberally.
   
   Maybe some notes should be added to documentation about this, at the same 
place where we walk through how to implement a trigger. I think it’s generally 
accepted that plugin changes require rebooting the processes, but triggers may 
be greener and less understood.
   
   I also wonder whether it is possible for Airflow to detect changes like this 
and automatically suggest restarting, instead of failing later. But this may be 
much more difficult since Python code changes are not that easy to detect in 
the first place. This topic is much more involved and should be explored 
separately if desired.


-- 
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