msokolov commented on issue #16618: URL: https://github.com/apache/lucene/issues/16618#issuecomment-5512617553
The speedup/slowdown is pretty remarkable, but I wonder if you could share total CPU comparing the "full rebuild" with the graph reuse optimized merge? I suspect we are saving on CPU usage, but losing on wall clock time due to the reduced concurrency. That wouldn't be the worst sin, and I think we'd rather have the fix for the recall drop even at the expense of some merging latency, and while we should fix this, I don't know if we can consider a bug so much as an opportunity for further optimization. -- 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]
