When loading a parquet file in Qlik Sense (Client-Managed May 2024 Patch1) we see some weird behavior if there is a NULL value in the parquet file. It looks like Qlik just randomly fills in a value (that does occur in that field) in that cell...
All non-NULL values are just loaded correctly...
Tested both in table object as in exporting to qvds...
Anyone knows what the reason could be?
Next steps we'll test is replacing NULL values with a N/A string when creating the parquet in spark and see if that resolved the issue.
