topic Re: Qlik shows no data before coming back normally in App Development

Qlik shows no data before coming back normally

maimeo — Fri, 03 May 2024 09:12:48 GMT

Hello,

I'm having a problem with my Qlik Sense application. A part of the data doesn't appear for a few hours (from 1:00 AM to 8:00 AM) and then comes back normally (around 9:00 AM).

This happens sometimes, but there are no rules. For example, the data may appear for a whole week, then the following weeks, the problem occurs for 4/7 days.

And this only affects a part of the data, not all the data in the application. The application loader takes several separate views in SQL Server as sources. The part data affected come from the same views.

Have you ever encountered this problem? Could you please give me some ideas on the root cause and resolution of this issue?

Thank you.

Re: Qlik shows no data before coming back normally

marcus_sommer — Fri, 03 May 2024 09:19:37 GMT

I think it means that the sql-views have not the complete data to the query-time. There are various reasons possible for such a behaviour - quite common would be that certain data are locked because they are processed by any other task.

Re: Qlik shows no data before coming back normally

maimeo — Fri, 03 May 2024 12:05:52 GMT

Thank you @marcus_sommer for your reply.

In fact, the data is showing 0, not a " - " symbol. Could this be corresponding to the reason of deadlocks that you mention ?

The other part of the apps shows data normally, so I am a bit confused. If the deadlocks happens upstream of qlik loader, is this possible that the loader continue to run, taking the dead part as 0 and other normal parts of data to display in the front ?

Re: Qlik shows no data before coming back normally

marcus_sommer — Fri, 03 May 2024 10:41:18 GMT

It depends on the implemented data- and NULL-handling within the SQL as well as in Qlik if the number of records are changing and/or if field-values and/or any UI views show ZERO or NULL or NOTHING.

You may look within the SQL log-files and/or tracking the number of loading records and/or implementing any kind of data-quality checks to prevent the unwanted behaviour, for example by skipping the update if the data are incomplete and remaining by an older state.