Is it possible to have some kind of notification popped up or displayed in a tab, or something showing what changed, maintenance message, or other notes? Something similar to what is available in Redshift would be brilliant - picture attached.
It would be great to have an option to join multiple datasets by the same column name as one column to avoid getting different columns when creating such a join. Alternatively, having an option to retain some historical datasets without refreshing them, allowing them to hold specific data for that purpose, with the option to exclude them entirely from all future refresh/incremental refresh. This is useful when dealing with large datasets, as we do.
Filtering between tabs - When you activate filter actions on one tab, it would be cool if it could work on other tabs with the same database.
Yeah, I am doing that; that’s why I propose that it can be done better.
Not true. Join always creates separate columns regardless of the names of the columns. Yes, I could use if-else and create a third column, but this is kind of a lame solution.
That’s what I am doing, and it’s not better than this. I mean, if QS can remember and activate filters on one tab, why not hold those values across other tabs?
Okay, let me rephrase what I’m aiming to achieve with Redshift. Currently, I’m addressing this using Athena and a batch job wich dumps data to Athena table and then I read/display those…
Let’s assume I want separate datasets for 2022, 2023, and now for 2024. I want to ensure that 2022 and 2023 won’t be queried again when 2024 begins. This is crucial not only during a full refresh but especially when I set up incremental refresh.
What would be the best way to accomplish this using only redshift and QS?
Why do you use different datasets, not only one dataset with filter, where clause in SQL or analysis filter?
Can i assume the table structure is the same for all?
Hi, so let’s say that I changed something, like adding a new dataset as a left join to my main dataset. I think it’s possible to QS the main DataSet as it is and just append this new one to it. As of now, it’s performing a whole refresh, which in our case takes more than 7 hours and completely blocks traffic on RS for others or at least slows them down. So my main proposition would be to have an option where it would be possible to append a new dataset to QS to the main one without a full refresh of the main DataSet. Additionally, this would be beneficial for other smaller changes as well. And one more thing, when I add a new incremental refresh schedule, it will perform a new full refresh. What’s more frustrating is that when you change something in DataSet and save it, a full refresh is already triggered, and then adding an incremental schedule launches a second query which is really not needed.
Hello @matjazz, are you utilizing custom SQL queries in QuickSight or are you pulling your full datasource table directly from Redshift? If your refreshes are taking 7 hours, I would highly recommend splitting your data into multiple, more concise datasets.
Use custom SQL queries so that you only pull exactly the data that you require for visualizations. You can utilize multiple datasets on an analysis/dashboard so as long as they are grouped for specific visualizations, this will greatly reduce the load on your Redshift resource. Plus you can use WHERE clauses to only grab data for specific constraints and times that you need. You can also run your joins in the query to reduce query times rather than joining datasets directly in QuickSight.
I will mark this as a solution for now, but if you have further questions, please let me know. I will also add a tag to mark this item - Is it possible to have some kind of notification popped up or displayed in a tab, or something showing what changed, maintenance message, or other notes? Something similar to what is available in Redshift would be brilliant - picture attached. - as a feature request for our support team. Thank you!
Oh, and one final note. If you use the Navigation Action to switch between tabs and set a parameter value that is linked to your filters, that will extend across all sheets that have filters linked to the same parameter.
