abhishek-chouhan commented on PR #14092:
URL: https://github.com/apache/druid/pull/14092#issuecomment-1513628841

   @abhishekagarwal87 Thanks for taking a look at this.
   > why not just do a retry (along with cache invalidation) on any 5xx?
   I was in 2 minds about this as well. It seemed like giving the callers an 
option to override this behavior, would provide the flexibility such that if in 
the future they would want to handle a 5xx selectively, without retry, then 
they'd have this option. Otherwise there would be no way to handle or expect a 
5xx response from the node, internally the client would always think this as 
inappropriate and make additional requests, before handing it back to the 
caller. I do agree that it makes it a bit more complicated for classes like  
LookupReferenceManager. 
   Let me know what you think, if we feel like this would not be the case, we 
can make changes to retry for 503/504. Or maybe we could make this a default 
behavior, but then provide another API with a boolean which disables this 
behavior for that specific call.


-- 
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