himanshukandwal opened a new issue, #2646:
URL: https://github.com/apache/helix/issues/2646

   ### Describe the bug
   In WAGED n - n + 1, we filter out the host from the map-field of idealState 
if it doesn't have capacity to hold the partition. The instance capacity is 
determined from the resource-partition weights, instance currentState, and 
pending messages to build WagedInstanceCapacity data structure. This 
data-structure is computed every time during a pipeline run as any cluster 
event can cause capacity change. In case of large clusters, this computation 
takes ~80% of total time and hence, it's important to improve its performance. 
   
   ### To Reproduce
   This issue is quite apparent in all the integration and release 
certification tests. 
   
   ### Expected behavior
   The additional calculation will take some portion of time but should not 
incur high latency in the overall pipeline runtime.


-- 
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