Have a QuickSight question or problem you need help with? Join us for our new QuickSight monthly live Q&A series. Our team of expert Solution Architects will coach you through solutions to solve any QuickSight challenge you may be facing.
Can you please confirm if the limitation described here still exists?
I’ve run into this limitation in the past but when I tried to reproduce the issue recently, I didn’t get any errors. I also didn’t see any announcement about the removal of the limitation.
After performing the joins, I decide to include the following fields in my child dataset and exclude everything else (all included fields are from UserPrograms):
I then update my datasets as follows:
I add a new OrganizationName field to the Organizations parent dataset and refresh my child dataset. The refresh is successful. I see the OrganizationName field in the child dataset but it is excluded by default.
I add a new Organization Id field (string) to the Organizations parent dataset and refresh the child dataset again. This time it fails because the child dataset already contains Organization Id.
When I edit the child dataset, I see both instances of Organization Id are now excluded. If I have RLS based on Organization Id, this means that RLS also doesn’t work anymore. I have to manually include the correct instance of Organization Id and refresh the dataset again to fix the issue.
Do you have any recommendations on how to update parent datasets without breaking the child datasets? I feel like in this case QuickSight should be able to handle update #2 just fine. It’s already appending the parent dataset name to the field name, so why does it still consider it as a duplicate?
The other issue here is that instead of renaming the new instance of Organization Id, QuickSight renamed the original one which is used in my RLS rules. In the case of an inner join, it doesn’t make a difference but it matters for a left outer join because I want the RLS rules to look at the correct instance of Organization Id.
We have some static images and we need to put it in QuickSight. However since it is in prod environment, security team is not allow to put in public URL link. Can we put the images in S3 and put a link in QuickSight.
Even though we already answered this question in the live Q&A, I wanted to make sure to also include it in the comments.
As of right now, only https URL scheme are supported, but we do have a feature request in review on directly providing a link from S3 instead of needing to use cloudfront.
I answered this question in the Live Q&A sessions, but you can set up filters with parameters to show only a specific set of the data based on the row and/or use filter controls to build interactive dashboards when you select specific rows in your dashboard.
If you have a specific case where you are trying to set this up, please let us know and we are happy to help!
The main issue we saw was the deleteDashboard API call consistently failed with the error of:
d: Cannot delete a dashboard version when it is published
The AWS CLI and python boto3 API calls to this functionality had no such complaint, and the documentation does not seem to address the ‘published’ status of dashboards or any real way to change it.
We are seeing an issue where we cannot filter on fields where the data contains a backslash \ . Embedded quicksight does not show an error but returns no data. Is this a known issue?