HANA Hot News: HANA建模时关于使用join的限制

最新推荐文章于 2024-11-10 07:04:05 发布

原创最新推荐文章于 2024-11-10 07:04:05 发布 · 725 阅读

1 ·

CC 4.0 BY-SA版权

文章标签：

#HANA

Data Warehouse&BI 专栏收录该内容

79 篇文章

订阅专栏

本文详细阐述了在HANA Modeler中进行星型模型建模时的若干规则，包括事实表、关键属性表以及逻辑连接等核心内容。重点强调了事实表与关键属性表的唯一性、多对一或多对多关系可能导致的行数意外增加等问题，并提供了避免这些常见错误的使用案例指导。

Note 1702641 - Limitations for star schema modeling in HANA Modeler

Currently (SP04) the following rules do apply.

1. The central table of an analytic view is the table the measures point to. This is also called the fact table. There can only be one fact table in the data foundation of an analytic view. I.e. all measures must point to the same table.

2.The central table of an attribute view is the table the key attribute(s) point(s) to. All key attributes have to point to the same table in the data foundation of an attribute view (in case there is more than one key attribute).

3. If the logical join from an analytic view to an attribute view is a multi column join, all join conditions (lines) must point to the same physical table of the attribute view (this does not necessarily need to be the central table of the attribute view, although in the majority of the use cases it will be that table)

4. The cardinality can be 1:1, 1:n, n:1 n:m. Cardinality 1:n or n:m might lead to unexpected multiplication of fact table rows (a warning during deployment is shown in that case)

5.A join path is the set of joins and tables that links the fact table to an attribute / table. A table must not appear twice in any join path
Please refer to attached document "Join_issues_in_HANA_Modeler.pdf" explaining which use cases are allowed and which are not.