I am building some dashboards that will be shown to external users.
We need to have one column in our data that determines whether or not those users can see a given entry in the table.
What’s most critical, however, is that they cannot ever see this column NOR can they change the filter that uses this column and then see the hidden rows.
How can I ensure that this private column stays secured yet functional?
Excluded solutions: separate tables is not an option.
Hello,
Typically this scenario is called row-level filtering and is accomplished in Dundas BI via a security hierarchy and a custom attribute on the user’s accounts. The custom attribute can contain different values for each users so they see different results.
You can read more about this functionality and how to use it here.
Hopefully this helps you solve this.
It looks like that may work. I’ll have to try it out fully to see, I guess. In my case, I just need to filter at the cube level (this is a new realization, since I posted the initial question) so perhaps I can just set a filter transform and then exclude the column from the cube output? Would that suffice to keep it fully isolated from the user, or do I still need the security hierarchy feature in that case too?
I’ve also just realized that I have another case that I need to apply the security hierarchy feature on, to isolate it. Glad I’m going down this road!
If you filter at the data cube level and do not make a public parameter, then yes it’s essentially locked to anyone who does not have write access to that data cube. That is definitely an option, but the only caveat there is the value of the parameter would be unchanging per user.
The custom attribute + security hierarchy functionality is there to allow different values for the filtering for different users, if required.