Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Hi All,
We are currently facing challenges with the increasing size of one of our Qlik Cloud applications. Below are the key details and constraints:
Initial App Design:
The app was originally designed with data aggregated at the monthly level, resulting in an app size of approximately 1.5 GB.
Requirement Change:
A few months later, a requirement emerged to analyze data at the daily level. After this change, the app size increased to ~3 GB.
Recent Change – Ageing Buckets:
In a recent enhancement, we introduced ageing buckets in a large transactional table.
Prior to bucketing: ~5 crore (50 million) rows
After bucketing: ~13 crore (130 million) rows
Current app size: ~7 GB
Data Volume:
We are only loading 3 years of data in the app.
Insight Advisor Constraint:
The app also has Insight Advisor configured.
We are considering splitting the app into two using document chaining to manage size, but we foresee a major limitation:
If a user queries from the landing sheet using Insight Advisor, it may not support document chaining to the second app, which is a blocker from a user experience standpoint.
Kindly suggest a feasible approach or workaround to help reduce the app size while considering the above constraints. Specific input on how to retain Insight Advisor functionality in a document-chained environment would be particularly helpful.
Thanks in advance for your support.
Have you removed unused fields from the data model and AutoNumbered the keys?
-Rob
http://www.easyqlik.com
http://masterssummit.com
http://qlikviewcookbook.com
Yes, I already did and used best practices.
have you run Performance Evaluation feature ??
hope you follow below Qlik Recommended practice.
https://help.qlik.com/en-US/cloud-services/Subsystems/Hub/Content/Sense_Hub/Apps/app-performance.htm
Regards,
Prashant Sangle
Difficult to make any more suggestions without seeing the data model or a QSDA analysis. Can you post a screenshot of the data model?
-Rob
Maybe not the entire 3 years are needing a daily level. If not - the daily- and monthly-level might be mixed:
Fact Table with Mixed Granularity - Qlik Community - 1468238
It's not quite clear what's meant with the aging buckets but it seems to duplicate the data in some way. Maybe this could be prevented with a special kind of dimensional table to address overlapping areas, like:
Difficult without seeing the model, but front end may be visual heavy, or has conditional formatting
script wise , are any tables being called and scripted multiple times? Unused columns removed and as mentioned autogenerate keys.
can the data be aggregated at another level?