mxm opened a new pull request, #554:
URL: https://github.com/apache/flink-kubernetes-operator/pull/554

   When the rates drop to zero, we set near zero values to continue to use the 
same scaling logic. There is a gap in this logic when only the input rate drops 
to zero but the output rate is still measured to be greater than zero. This is 
because we do not pass on the output rates directly but the output / input 
ratio which becomes very large in this special scenario.
   
   In this change we cap the output rate in case we use near zero input rates. 
This has been tested to effectively prevent scale up spikes in pipelines which 
only occasionally receive data.
   


-- 
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