drogonframework/drogon

jemalloc benchmarks

Open

#635 建立於 2020年11月25日

在 GitHub 查看
 (6 留言) (4 反應) (0 負責人)C++ (10,462 star) (1,014 fork)batch import
enhancementgood first issue

描述

Hey, I recently benchmarked using jemalloc on my NUMA servers, and the difference was quite large. Using the system allocator I got 710k req/s, while with jemalloc I got 830k req/s. That's 17% faster just by dynamically linking with a library.

Anyone else tried this?

貢獻者指南

技術棧
cpp
領域
backendperformance
議題類型
research
難度面向新貢獻者的預計實作難度，1 表示很小改動，5 表示專家級工作。
4
預計時間有經驗貢獻者完成調查、實作、測試並準備 pull request 的粗略時間範圍。
1-3 hours
活動狀態議題目前的可參與程度：新鮮、活躍、陳舊、阻塞或等待維護者輸入。
stale
清晰度議題是否清楚說明預期改動、驗收標準和下一步。
needs investigation
前置要求
C++ knowledgeBenchmarking experienceUnderstanding of memory allocators
新手友善度1-100 的估計分數，表示該議題對首次貢獻者的友善程度。
20
研究方向
The issue asks for benchmarking results comparing jemalloc vs system allocator on the Drogon framework. To investigate, set up a Drogon server on a NUMA machine, compile with jemalloc linked dynamically, and run benchmarks (e.g., using wrk or similar). Compare requests per second and latency. Look at existing benchmark scripts in the repository, if any. Since no specific benchmark code is provided, you would need to create a minimal server and run tests.