Applying Joined Filters for Dashboard Views from Multiple Data Sources

Hi Tableau experts,

I’m quite a beginner in Tableau. Apologize in advance for my long question…

I’ll be glad to get your help on the following issue I’m facing, which I couldn’t find an appropriate solution for it, even after I tried using various solutions suggested here…

I’ve created few sets of views from 3 different data sources (each view from its relevant one) and gathered them into one Dashboard.

Dashboard requires joined filters on few dimensions, that are available in all 3 data sources, containing identical data (for example: Date, Region, Organization, etc.),

but dimensions names are slight different from one data source to another.

Note that I’m using data sources via Hortonworks Hadoop Hive Server, on Tableau 2018.1.2, 64-bit.

These are the methods I tried to handle this need of applying joined filters, that eventually any of them didn’t provide the appropriate solution:

1) Data blending among all 3 data sources by editing relationships between all relevant joined dimensions that required the joined filters, as well as activating them as linking fields.

Once I’ve done that, all Count Distinct calculations I have in the views turned inactive (red capsule with error message: “cannot blend the secondary data source because one or more fields use an unsupported aggregation”).

Note that the Count Distinct calculation is from the Primary data source, only the (joined) filter I applied in this view is from the secondary data source.

I tried to use other suggested formulas to calculate the count distinct, other than the simple one, but neither of them worked…

Example: {fixed [Customer Name] : max([Number of Records])}. Doesn’t work… Still provides the same error.

2) I cancelled all data blending/relationships/linking fields, and went to another direction: I joined all 3 data sources, again with all relevant dimensions,

tried left join from primary to secondary data sources, and then noticed this joining caused significant differences in the given data, comparing view from single data source to view from joined data sources.

3) I cancelled all joins and kept every view to get its data from its relevant data source. No joining. No blending. Then tried to created joined filters by creating parameters for every set of views from each data source.

This method works! Finally I manage to filter common dimensions from different data sources across all views, as well as adding the option to select ‘(ALL)’ data in the created parameter filter.

BUT… this option turned to provide incomplete solution, as it doesn’t provide the option for multiple values selection (dropdown), as the regular structured Tableau filter provides, which is MUST for us.

I tried to use suggested solutions for applying multiple selection option in parameter filter, but they were totally unfriendly and way far from the good structured one…

Can you please kindly help me to get an appropriate solution for this need, that I was sure can be solved easily by the strong capabilities of Tableau?

Note: Really sorry, I cannot attach here any of my files due to information security.

Thanks in advance