gnodet commented on issue #12667: URL: https://github.com/apache/maven/issues/12667#issuecomment-5444753360
## Benchmark: maven-4.0.x branch with cherry-picked PRs Tested on a branch based on `maven-4.0.x` with the 5 Maven PRs cherry-picked (#12652, #12653, #12654, #12655, #12662). The resolver optimizations from apache/maven-resolver#2014 are already included via resolver 2.0.22. The install/deploy plugin PRs (#427, #684) are **not** included (separate repos). **Test machine:** Apple M4 Pro, JDK 21 **Project:** 4,383-module diamond-graph reactor (`clean install -DskipTests -q -B`) | Configuration | Run 1 | Run 2 | Run 3 | Avg | vs RC6 | |---|---|---|---|---|---| | Maven 3.9.16 | 1:51 | 1:55 | 1:53 | **~1:53** | — | | Maven 4.0.0-rc-6 (unpatched) | 3:24 | 3:37 | 3:29 | **~3:30** | baseline | | Patched maven-4.0.x + resolver 2.0.22 | 2:25 | 2:28 | 2:22 | **~2:25** | **-31%** | ### Notes - The patched build shows significantly higher CPU utilization (~550% vs ~383%), suggesting the AsyncDrainWriter lock contention fix (#12654) is enabling better parallelism. - The remaining ~32s gap vs 3.9.16 could partially be closed by the install/deploy plugin optimizations (O(N²) → O(N) reactor scanning). - The `DefaultModelObjectPool` changes from #12652 were dropped during cherry-pick (file doesn't exist on `maven-4.0.x`). _Claude Code on behalf of Guillaume Nodet_ -- 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]
