keystonejs/keystone-classic

[0.x] Keystone OOMkilled on large CSV export

Open

#2,592 建立於 2016年4月8日

在 GitHub 查看
 (1 留言) (0 反應) (0 負責人)JavaScript (14,656 star) (2,288 fork)batch import
bughelp wanted

描述

When exporting a keystone list as csv, the entire backing collection is loaded into memory before sending the response. This renders the export feature useless for a large collection; keystone will exhaust the available memory on the machine until it gets OOMkilled.

Instead, the export should be streamed back to the client so that we never load more than a cursor's worth of data in memory.

貢獻者指南

技術棧
javascriptnodejs
領域
backend
議題類型
bug
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。
stale
清晰度議題是否清楚說明預期改動、驗收標準和下一步。
clear
前置要求
Understanding of Node.js streamsBasic knowledge of CSV generationFamiliarity with Keystone API
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
Investigate how the current CSV export is implemented. Look for a file handling CSV export (likely in the keystone admin or export module). Implement streaming using Node.js Readable streams or async generators to send data in chunks instead of loading entire collection. Consider using a cursor for database queries to avoid large memory. Reference the issue comments for any additional context.
[0.x] Keystone OOMkilled on large CSV export · keystonejs/keystone-classic#2592 | Good First Issue