Understanding legacy model?

[ovonel]

I have this legacy model:

I have a few questions...

Why is it bidirectional? If we only want the filter to flow to the Fact, what could be a reason to have a <> relationship?

I know this is used for RLS, but isnt it redundant to have so many tables to handle it? or is this standard?

Having so many tables... where is best to put the RLS filters? on which table?

[v-jianboli-msft · Accepted solution]

Hi @ovonel ,

A model relationship propagates filters applied on the column of one model table to a different model table. Filters will propagate so long as there's a relationship path to follow, which can involve propagation to multiple tables.

With bidirectional cross-filtering, report creators and data modelers now have more control over how they can apply filters when working with related tables. Bidirectional cross-filtering enables them to apply filters on both sides of a table relationship. You can apply the filters by propagating the filter context to a second related table on the other side of a table relationship.

For more information and for examples of how bidirectional cross-filtering works, check out the Bidirectional cross-filtering for Power BI Desktop whitepaper

Model relationships in Power BI Desktop - Power BI | Microsoft Learn

Best Regards,

Jianbo Li

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.