知识卡片
数据湖延后建模的代价
数据湖先保存原始运营数据,之后再生成面向报告或机器学习等不同任务的分析模型,已有原始数据也能支持未来新增模型。代价是要兼容原始模型多个版本,维护多套转换;输入缺乏质量和结构约束时,易进难用的数据会变成沼泽。延后转换没有消除领域理解与治理责任。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
原书出处 · 1 处
- 数据网格 · 查看原书依据
A data lake–based system ingests the operational systems’ data. However, instead of being transformed right away into an analytical model, the data is persisted in its raw form, that is, in the original operational model. Eventually, the raw data cannot fit the needs of data analysts. As a result, it is the job of the
参考来源
- 位置:《Learning Domain-Driven Design》第16章《数据网格》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Learning Domain-Driven Design/_epub-src/OEBPS/ch16.html)
- 结论依据:原文比较原样落盘和立即转换,并说明多版本ETL与数据沼泽问题。
- 原始内容:A data lake–based system ingests the operational systems’ data. However, instead of being transformed right away into an analytical model, the data is persisted in its raw form, that is, in the original operational model. Eventually, the raw data cannot fit the needs of data analysts. As a result, it is the job of the