pingcap/tidb

executor: avoid joiner to copying data when filter rows

Open

#12,030 创建于 2019年9月4日

在 GitHub 查看
 (1 评论) (0 反应) (0 负责人)Go (40,090 star) (6,186 fork)batch import
help wantedsig/executiontype/enhancementtype/performance

描述

Feature Request

Is your feature request related to a problem? Please describe:

We can use the sel field in chunk.Chunk instead of copying data in baseJoiner.filter when filtering data in join executor

Describe the feature you'd like:

We can use the sel field in chunk.Chunk used in baseJoiner.filter

Describe alternatives you've considered:

Teachability, Documentation, Adoption, Migration Strategy:

贡献者指南

技术栈
go
领域
backenddatabaseperformance
议题类型
feature
难度面向新贡献者的预计实现难度，1 表示很小改动，5 表示专家级工作。
3
预计时间有经验贡献者完成调查、实现、测试并准备 pull request 的粗略时间范围。
half day
活动状态议题当前的可参与程度：新鲜、活跃、陈旧、阻塞或等待维护者输入。
stale
清晰度议题是否清楚说明期望改动、验收标准和下一步。
clear
前置要求
Basic knowledge of GoUnderstanding of TiDB's executorFamiliarity with chunk.Chunk and sel field
新手友好度1-100 的估计分数，表示该议题对首次贡献者的友好程度。
50
研究方向
Study the `baseJoiner.filter` function in `executor/joiner.go` and the `sel` field in `util/chunk/chunk.go`. Understand how data is currently copied and how using `sel` can avoid that. Implement the change to use `sel` in `filter`, ensuring no performance regressions. Add tests to verify correctness.