jbellis commented on PR #12372: URL: https://github.com/apache/lucene/pull/12372#issuecomment-1594763703
I'm using the million-row sift dataset via this harness https://github.com/jbellis/hnswdemo/tree/benchmarking I believe what is happening is that allocation is basically free and there is enough slack across the JVM in these synthetic benchmarks to soak up the extra GC required -- I do not see any changes when run normally, either. So I forced it to a single core with ```$ taskset -c 0-1 ./gradlew runTexmex -PsiftName=sift``` (it is a hyperthreaded core so i give it logical cores 0-1) Also if you take a look at the jfr files it is very clear that a significant amount of allocation is gone. -- 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
