Thursday, March 19, 2015

QV25 How to tune a qlikview application

We know that Qlikview is data pulling and other operations are quite faster than other tools. Still we need to optimize the qlikview document.

Reload, chart loading are depend upon various factor. Good data model also leads to high speed.

Some below factor we need to take care in document:

Reduce Rows
Eliminate unneeded data volume (rows) from apps (how far back does the app really need to store data?
Reduce Columns
Find unused fields in the data model - use DocumentAnalyzer_V1.5.qvw against your app to do this, then copy the "DROP FIELD" statements that are generated for the unused fields and place these into a new tab in YOUR application, as the VERY LAST tab in the script. 
Reduce Distinctness
Reduce Distinctness - Reduce distinct values in your application - use the QlikView Optimizer.qvw against your app to find the fields that are candidates for these changes.   Select the "Symbols" in the Optimizer app to see these (timestamps, system keys, etc..)
Increase efficiency
Convert fields to numeric that are used in compare expressions, like IF(ClientType = 'Direct'….)
Reduce data model complexity
Collapse snowlfaked tables to attain a more pure star schema   
Reduce App complexity
Reduce number of tabs/objects in the application, where appropriate
Reduce App complexity
Reduce the number of open charts down to 1 on tabs where this is possible and appropriate
Best Practices
Follow checklist items above in all areas

For Charts
1. Avoid Calculated dimensions - if possible use direct column in dimensions, calculated dimension leads to decrease the performance of table chart, because calculated dimension, do calculation for every row.

No comments:

Post a Comment

web stats