App size limits in Qlik

[akolwalkar]

Hi,

I created a Qlik app that could be considered as a large app (46Gb), and I have maintained it inside a space that supports large apps. Ever since I have built the app, Qlik fails to reload it, irrespective of whether its triggered from the catalog or via a scheduler. The error I get is: "Failed to create a session for reload: insufficient privileges". I am a tenant admin, with the rights to reload apps in that specific space, and I also do not have any section access security in my app script loader. The only possibility I am thinking is this has something to do with my app size, so any insights would be appreciated. Thanks!

[FedericoDellAcqua]

46GB is a lot! probably you could considerate to use ODAG? 

By the way, check if the user (even if it is TenantAdmin) has also a professional license. 

Federico

[marcus_sommer · Accepted solution]

Helpful may be to reduce the size maybe to 1/5/10/20 GB to see if the error always remains or occurs only by a larger data-set. If the general access is successful but failing by the larger data-sets it hints to any configurations - directly to the sizing or indirectly for any kind of timeout. 

[akolwalkar]

Thanks! I was able to address this by reducing the app size; but also learnt that there's no way to fix it within the existing app. I had to create a new app, copy the script over and limit the size of the query so it loaded a smaller dataset on the first run. Kind'a frustrating that Qlik doesn't recognize this in the existing app, and just refused to run any subsequent queries there, while throwing an error that didn't relate at all. 

[marcus_sommer]

I don't know which features Sense provides but in View there were a possibility to open an app without any data as well as to reduce the data-set of an opened app against the chosen selection - and to store the app in this state. Then may come a reload with the debugger by limiting the max. number of records or maybe adding some FIRST N statements to the biggest loads.

By a corrupt app (happens occasionally by any network/storage issues whereby the risks increase by larger ones) it wouldn't help but if the application is ok. it might be a way to reduce the app-sizing gradually.

Beside this by data-sets of such amount it's recommended to apply the common performance optimization measurements (only necessary data, field-cardinality, field-formatting, star-scheme, incremental approaches, ...).