jpountz commented on PR #895: URL: https://github.com/apache/lucene/pull/895#issuecomment-1128736566
The current calculation makes sense to me. Merge policies like to organize segments into tiers, where the number of segments on each tier is typically also the number of segments that can be merged together. So it doesn't make much sense to perform multiple merges on the same tier concurrently. The way I'm reading the current formula is that we are scaling the number of merge threads with the number of processors, but stop at 4 anyway because it already allows Lucene to perform merges on 4 different tiers concurrently, which is already a lot given that tiers have exponential sizes, and that TieredMergePolicy has a max merged segment size of 5GB. -- 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: issues-unsubscr...@lucene.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For additional commands, e-mail: issues-h...@lucene.apache.org
