Skip to main content
Announcements
Qlik Connect 2024! Seize endless possibilities! LEARN MORE
cancel
Showing results for 
Search instead for 
Did you mean: 
dmohanty
Partner - Specialist
Partner - Specialist

Understanding RAM Footprint in Document Analyzer and ways to improve the performance

Hi Experts, @rwunderlich 

I have reloaded the Document Analyzer for one of my QlikView App (QVW) which is 3GB in size. The QVW has to open up on Qlik Windows server (on google Virtual Machine instance) which has 416GB of RAM, which is taking average of 5 mins to open up on browser. Which is still high time for user and I want to reduce the time to opening the App.

I have the below summary generated for me. 
I am also able to track the Unused fields, Suggestions and do the necessary changes/drop the unused fields/convert to right data-type etc. 

But can someone please help me understand from this highlighted section -

  1. What exactly RAM Footprint numbers are denoting, are there any scope to improve this?
  2. Does this Object Calc Time number look too high, can it be reduced/what's the ideal Object Calc Time value? 

My motive is to make the 3GB App open up on 416GB RAM Server a bit faster (like 1-2 mins?)

cardinalteam1_0-1653077731957.png

 

Labels (1)
2 Replies
rwunderlich
Partner Ambassador/MVP
Partner Ambassador/MVP

Sorry for the delayed response, I just saw this.  Reducing the RAM requirement should help the opening time. You can start by:

1. Dropping the unused fields. 

2. Examine the fields in the Fields sheet and sort by size. Based on your numbers, I suspect you have several fields that have many values.  Look at those fields to determine if you can reduce the storage requirement by reducing the number of values. Do you have keys that can be autonumbered?

You do have several charts that seem to take a long time to calc - 40, 80 and 100 secs. I would focus on those to see if they can be improved. 

-Rob

 

dmohanty
Partner - Specialist
Partner - Specialist
Author

Thanking you @rwunderlich for your generous suggestions. 

I have started looking into the above pointers and look for further improvements.  Will keep you posted. 

Regards!