simonw/datasette

Research: demonstrate if parallel SQL queries are worthwhile

Open

#1,727 建立於 2022年4月27日

在 GitHub 查看
 (32 留言) (0 反應) (0 負責人)Python (8,637 star) (625 fork)batch import
help wantedresearch

描述

I added parallel SQL query execution here:

My hunch is that this will take advantage of multiple cores, since Python's sqlite3 module releases the GIL once a query is passed to SQLite.

I'd really like to prove this is the case though. Just not sure how to do it!

Larger question: is this performance optimization actually improving performance at all? Under what circumstances is it worthwhile?

貢獻者指南

技術棧
pythonsqlite
領域
backenddatabaseperformance
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。
3
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。
half day
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。
active
清晰度議題是否清楚說明預期改動、驗收標準和下一步。
mostly clear
前置要求
Python basicsSQLite conceptsPerformance benchmarking
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。
40
研究方向
Review the existing parallel query implementation in issue #1723. Design a benchmark that runs a set of SQL queries sequentially and in parallel, measuring total execution time and CPU utilization using Python's time and resource modules. Compare results on multi core machines and document findings in a comment.