Large JSON responses results in incomplete schema using REST connector

[NoNameFound]

Hey everyone, 

I've encountered an issue when loading a large JSON response through the REST connector. 
The response I get contains objects that may have a sub-array or not ("someinfo":{"id":1,"text"="bla"} / "someinfo":"") and is only a single page, so no pagination. 

When the first few rows do not contain a sub-array, but the last few do, Qlik straight up ignores that and does not get the schema right. 

My question is: How far does Qlik look into the response to infer a schema and how can I make sure that Qlik gets the schema right for the entirety of the response? I've already tried to turn up the symbols preload count to "unlimited" (because I hoped it might help), but to no avail. 

Thanks. 

[rafaelencinas · Accepted solution]

Hi @NoNameFound !

I have this same issue with larger JSON Files, 

In my case, when I create a Data Connection, I tried to found a record with all attitributes filled, but , when its not possible, in Load Statement I add manually the fields and sub queryes, or I try to mix different Load Statements.

Bye!