sgup432 commented on PR #15124: URL: https://github.com/apache/lucene/pull/15124#issuecomment-3321452682
@benwtrent I have removed boolean query changes now. Ran the benchmarks again. **Baseline (100% write)** ``` Benchmark Mode Cnt Score Error Units QueryCacheBenchmark.concurrentPutOnly thrpt 9 4057602.444 ± 380153.561 ops/s ``` **My changes (100% write) - 47% (1.9x) slower than baseline** ``` Benchmark Mode Cnt Score Error Units QueryCacheBenchmark.concurrentPutOnly thrpt 9 2133320.577 ± 335226.223 ops/s ``` **Baseline - 60% read, 40% write (6 read threads, 4 write threads)** ``` Benchmark Mode Cnt Score Error Units QueryCacheBenchmark.concurrentGetAndPuts thrpt 9 8074623.026 ± 1229792.902 ops/s ``` **My changes - 60% read, 40% write - ~10% slower than baseline** ``` Benchmark Mode Cnt Score Error Units QueryCacheBenchmark.concurrentGetAndPuts thrpt 9 7263435.185 ± 828131.018 ops/s ``` Overall mixed load benchmark looks fine to me, considering I was only testing with boolean queries. In a lot of cases, we might be directly caching `TermInSetQuery` or `TermsQuery` for which we already cache ramBytesInUsed, so this 10% slowness will also go away overall. -- 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]
