tikv/tikv

Better memory observation for tidb_query executors

Open

#5,338 建立於 2019年8月27日

在 GitHub 查看
 (7 留言) (0 反應) (0 負責人)Rust (15,922 star) (2,189 fork)batch import
difficulty/easyhelp wantedsig/coprocessorstatus/mentorstatus/proposal

描述

Feature Request

Inspired from https://github.com/pingcap/tidb/issues/11840. Currently we have no existing way to check how much memory is used in some memory-heavy executors like hash aggregation and top n. Adding metrics to these executors should help diagnose these kind of problems without the need to build a specialized binary.

貢獻者指南

技術棧
rust
領域
backendperformanceobservability
議題類型
feature
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。
3-5 days
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。
mostly clear
前置要求
Rust programmingKnowledge of TiKV codebaseUnderstanding of memory profiling
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。
25
研究方向
Study the existing memory metrics in TiKV, particularly in the tidb query executor components. Look at the hash aggregation and top n executors in the tikv/src/coprocessor directory. Examine how metrics are implemented elsewhere, for example in tikv/src/server/status.rs using the prometheus crate. Identify where to add memory monitoring and propose specific metrics (e.g., cumulative memory usage, peak memory).
Better memory observation for tidb_query executors · tikv/tikv#5338 | Good First Issue