benwtrent commented on PR #13586: URL: https://github.com/apache/lucene/pull/13586#issuecomment-2237254527
🤔 my benchmarking is suspicious. I wonder if I am doing something wrong. I have a 4GB index, on a 4GB machine, 1GB set aside for the JVM. So, QPS should be about the same. However, I am getting far worse numbers and some weird artifacts (this is 5 runs in a row with the exact same parameters). baseline: ``` completed 1000 searches in 4776 ms: 209 QPS CPU time=4741ms completed 1000 searches in 4923 ms: 203 QPS CPU time=4904ms completed 1000 searches in 4557 ms: 219 QPS CPU time=4533ms completed 1000 searches in 4667 ms: 214 QPS CPU time=4655ms completed 1000 searches in 4702 ms: 212 QPS CPU time=4700ms ``` Candidate: ``` completed 1000 searches in 5151 ms: 194 QPS CPU time=5106ms completed 1000 searches in 5383 ms: 185 QPS CPU time=5371ms completed 1000 searches in 8719 ms: 114 QPS CPU time=5894ms completed 1000 searches in 7343 ms: 136 QPS CPU time=5554ms completed 1000 searches in 5285 ms: 189 QPS CPU time=5272ms ``` I can restrict memory even more to try and force a significant difference, but if this code is so much worse even when everything already fits in memory, something is wrong. -- 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
