Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Hi all,
What are your recommended good practices when it comes to managing QVFs that are solely for creating QVDs in Qlik Sense?
I normally drop the tables after the QVDs are created, but I am looking for recommendations around the QVFs itself.
Do you normally publish QVFs these in a hidden stream? Or leave them in your own stream in an environment? Are there any naming conventions that are good to apply?
Please, I welcome good suggestions for the benefit of me as well as the community 🙂
Hi, in our company we add a prefix "x." for Stream name that contains only generators of QVD. These streams are only visible for developers. Should be published with scheduled task to reload.
We publish these to a QVD Generators stream that only the admins have access to. We do a lot of transformation so we have generators for Extract, Transform and Load all separated out. Each with its own task so it's easier to start at a particular point.
We clear out the qvd's after the final load steps though so the Extract and Transform versions don't take up extra disk space.
Hi swallac104,
This is interesting, what technique do you use for clearing out the QVDs after the final load? 🙂
Thank you all for the replies I appreciate the sharing!
Something like:
FOR i = NoOfTables()-1 to 0 step -1
DROP TableName($(i));
NEXT
Overwrite them with an empty INLINE statement. The newest version of QDF includes a new subroutine called reduceQVD() to do this for you, so all you have to do is loop through the directories and call it for each table. Leaves the actual QVD file but it's just the header so it's like 1k in size.
I think he was talking about the tables in the data model in the QVF, not the actual QVD-files.