Direct Discovery: How To Make A Successful Deployment?

The key to success with using Qlik Direct Discovery is to understand the feature and it natural limitations. 

Direct Discovery: Technical Documentation describes the main differences between Direct Discovery and traditional Qlik in-memory data. The Direct Discovery - Security Considerations additionally highlight basic security aspect that can be relevant to consider before deployment.

Direct Discovery is intended and suitable under the following conditions:

• Build aggregated charts on homogenous large data sets
• Look at detail records in a table box on large data sets
• Reflect updated records without reloads (not new records)

Direct Discovery has the following natural limitations:

• Not as fast as in memory apps
• Not a solution for scalability issues where the real problem is wrong data model or UI design
• Not designed to convert all tables in apps into Direct Discovery mode
• Not a real time solution
• Will always be slower compared to SQL query run times on source due to associative model calculation times

There is reason to consider Object Calculation Time Limit and Max Symbols in Charts setting on QlikView Server side as Direct Discovery feature queries an external system from QlikView, the chart calculation time is dependent on the performance of the underlying system, and the query can return many distinct values leading to a very high number of symbols to display. 
 
When Ajax client is used, the QVS time out setting on the config.xml file may require adjustment to allow QlikView enough time to get query result before the client closes the connection. 
Consider increasing QVS Time Out Setting in the Config.xml file
<QvsTimeout>60</QvsTimeout>