Skip to main content
Announcements
Accelerate Your Success: Fuel your data and AI journey with the right services, delivered by our experts. Learn More
cancel
Showing results for 
Search instead for 
Did you mean: 
g_kolkman
Contributor
Contributor

Error Binary Load Large QVW-file

Hi all,

I'm trying to load a (large, 600mb) QlikView-file into QlikSense using binary load. The binary load statement works fine with other (smaller) qvw's, so no issues there and it's not a typo or anything. I was just wondering if anyone knows about problems regarding the binary load of large qvw-files in QlikSense.

It completely crashes, most of the times not only the QlikSense desktop client but my whole pc as well. 

I have seen several errors with exactly the same script-code (only a binary load):

  • Creating Search Index -- Unknown Error
  • Internal Engine Error
  • Or just simply crashing my pc (sometimes after as little as 4/5 seconds)

Anyone knows if this behaviour is avoidable? Or if QlikSense is simply not capable of handling such large datamodels?

Labels (4)
3 Replies
jonathandienst
Partner - Champion III
Partner - Champion III

That is not an exceptionally large model.

Are you sure that this not caused by out of memory conditions? A binary load from a 600MB qvw file could require far more than 600MB or RAM, and even more if the QVW is compressed. So you could be requiring more RAM than is available.

Monitor the binary load in the task manager to see how the RAM is used.

Logic will get you from a to b. Imagination will take you everywhere. - A Einstein
g_kolkman
Contributor
Contributor
Author

Thanks for the answer, I've been experimenting a little bit more and got access to a temporary server with 32gb of RAM. There the binary load does succeed, though it takes very long (12 minutes). The same binary load takes around 1 minute 20 in QlikView.

This means QlikSense takes 9 as long to load the exact same file as QlikView. How is this possible? We are considering implementing QlikSense with our customers side by side with Qlikview. We figured doing the ETL with QlikView and loading the final datamodel in QlikSense.

I was wondering if this is common behaviour or if it would something specific to our QV-application causing the slow behaviour of QlikSense.

Greetings

marcus_sommer

Before looking in other directions take a look on the transfer-rates of the used storage/network (also including any security tools) - not that you just measures their performance.

- Marcus