Hi, I’d like to ask for you awesome help once again 🙂 I’m struggling with filtering the data in a child workspace. Resp. it seems that even though filter is working, it won’t restrict some values. • On the first screenshot is LDM with identification of the most important data sets.

Company ID

is the column used for the workspace filter row level security. • Second screenshot is showing filtered child workspace without selected Company, coming from the most left data set (ideally, the filter wouldn’t be present on the dashboard at all, but it’s needed for the demonstration). The Limits are showing correct info for the company defined in the filter. Active Contract values are correct as well (you can see the scale is in similar range as PPU limit). But Values are showing aggregated numbers for all the companies, not respecting filter associated with the workspace. • Third screenshot is showing the dashboard when particular company is selected (and which is the only one available in the filter to be selected). When you actually use the filter, the Values suddenly show correct numbers. And Limits keep showing correct numbers like before. What might be causing this behaviour? The only difference between the datasets in the LDM is that

KBC Usage Metrics Values

data set is not linked directly to

Company

data set. I believe that’s the culprit, or rather the fact that the underlying table does not contain

Company ID

column like the other fact tables, hence it’s not filtered by the child workspace filter. It that a correct behaviour? I’d thought that the most left (Company) data set is filtered, hence only related rows might be accessed in the child data sets. As far as I remember, that’s how it worked with MUFs. Thanks in advance for any info and help.

Hello Martin, This is actually expected behaviour. The Workspace data filters in CN are set on specific column. Therefore, in your case, the

Company ID

is always filtered. Also, this rule is automatically applied to all the datasets where database column matches with the column name defined in the Data Filter. So as you’ve already correctly observed, that is not the case for the third dataset in question. The reason why you see “correct values” as per the third screenshot is that the particular company filter was applied and thus the subsequent inner join made sure that

Values

are filtered as well accordingly. BTW, there will be some changes and improvements regarding data filters functionality in next GoodData CN release. I would recommend you to follow our GoodData CN release notes and stay tuned for further updates.

Hi Julius, thanks a lot for the info. Can you tell me approximate release of the new version (I know you can’t promise anything) or if this “issue” would be somehow solved in the new version? I’m just trying to figure out if it make sense to wait, or if we should make our model bit more messy 🙂 Thanks once more.

Hi Martin, as I’ve said, that is expected behaviour that only inner join would show the filtered results for the example (not directly connected) datasets we were discussing. Regarding the release date, I really don’t want to give you any promises but we hope this will be next week. However, please note that this can always change. In any case, you should have the information soon 🙂

I know, that’s why I’ve used the double quotes 🙂 Anyway, thanks for your answers and let’s see what the next version brings us. Have a good one!