kfaraz commented on issue #18764:
URL: https://github.com/apache/druid/issues/18764#issuecomment-3561338239

   > Printing the server B action in the load callback routine actually yields 
MOVE_TO (basically the action has not been removed yet from the queuedSegments)
   
   This might be true, but `ServerHolder` for B is not relevant in the next 
duty run since each run of the duty uses a fresh set of `ServerHolder` 
instances (which is a snapshot of the time the duty run starts). This is why 
the callback does not use the `ServerHolder` and refers to the peons directly 
(peon objects remain in scope as long as the underlying server is visible to 
the Coordinator), since the `ServerHolder` would have a stale snapshot state by 
the time the callback actually gets run.
   
   The race condition that you have encountered, is it happening within the 
same duty run or across two duty runs?
   
   


-- 
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]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to