jadami10 opened a new issue, #16358:
URL: https://github.com/apache/pinot/issues/16358

   https://github.com/apache/pinot/issues/8490 did a great job adding an 
interface/class for detecting connection failures between broker <> server to 
mitigate query errors.
   
   We've occasionally seen issues with the retry logic being implemented as 
"mark the server healthy again, and see if it works". For high throughput use 
cases (thousands of RPS), this can cause thousands of queries to fail every 
time we hit the exponential backoff. Likely it makes more sense for the broker 
to first attempt to connect to the server to see if it's even responsive before 
marking the server as healthy again.


-- 
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