Section access works from CSV but not QVD

[DF_DB]

I have an app that generates a CSV for Section Access.  I load from that CSV into my Qlik apps and all works as expected.

I am now attempting to automate the SA file creation by reading and combining data form OKTA and our HR system and creating tables in BigQuery.  I have a second Qlik app that reads that data and creates a Section Access QVD.

When I load the CSV and QVD side by side to compare the results, they are identical.

However, when I load from CSV, it works.  When I load from QVD, it fails.

Anyone else experience this? Data in both look the same.  Data types look the same.

[PhanThanhSon]

Hi, 

have you try upper(trim(Field)) your Reduction fields?

[DF_DB]

It was already trimmed/uppered.

If I take the exact same data and store it to QVD and to CSV, then load it the same way, the QVD fails and the CSV works.

[LorantPataki]

Hi there,

I'm using QVDs for section access and it works. However, I'm using the Upper() function in the load script even if the data in the QVD is already upper case and I'm aliasing each column in capitals. For example:

Section Access;

Load

Upper(ACCESS) as ACCESS,

Upper(USERLOGIN) as USERLOGIN,

etc.

From ... (qvd);

Have you tried aliasing even if the column names are exactly as in the QVD?

[LorantPataki]

... one more thing - if your column names in the QVD are not upper case, you should still write them in ALL CAPS in the load script

[henrikalmen]

Another thing to compare is whether the fields are in the same format after loading. Let's say the field USERLOGIN is a number (not likely though, but what do I know) and loading from CSV interprets the value as text but loading from QVD interprets the value as number. It's a longshot, but could be worth checking. Load the tables from qvd and from csv, without it being within section access, and see what they look like.

You could for example do something like:
qualify *;
csv: load * from csv;
qvd: load * from qvd;

and then compare them in the data model and by putting them on a sheet as tables side by side.