Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 
obriened83
Creator
Creator

Qlikvew Server Performance Issues (Preceding Loads)

Hi

I was wondering if someone could please help me in figuring out this perfomance problem. We currently have Qlikview 10 SR4 installed on our Sever with the following hardware and settings

  • 40 Physical Core Processors (Intel(R) Xeon CPU      E7-4870 @ 2.4GHz)
  • 512 GB of RAM
  • 64 Bit Operating System
  • Hyper Threading disabled

The problem lies in the fact that my local PC (4 Cores Only) can load un-optimized QVDs faster than the Server. Some load time differences could be over 2 mins compared to 30 seconds on the local PC.

The problem seems to lie in Qlikview's Preceding Loads. For Example loading data from the Data Warehouse is almost ten times faster on the Server when you do not have a preceding load in front of the SQL query. You would expect way better performance from such a powerful Server.

I would greatly appreciate it if anyone can help me solve this performance problem. Thanks again.

Kind Regards

Eamonn

5 Replies
gonzaloruizgonz
Partner - Creator
Partner - Creator

Hi Eamonn,

I have the same problem, Have you gound any solutions for this?

BR

Gonza

obriened83
Creator
Creator
Author

Hi Gonza

     I am currently in the process of logging this issue with Qliktech. If I get any development on this I will let you know as soon as possible . Please let me know if you find anything helpful on this. Thanks very much.

Kind Regards

Eamonn

gonzaloruizgonz
Partner - Creator
Partner - Creator

I will also register the issue on the support. and i can refer to this discussion. In our case we are using the E7340, is look like an issue with the E series.

//Gonza

obriened83
Creator
Creator
Author

Hi Gonza

             Did you ever get to the solution for this problem?

Thanks very much.

Kind Regards

Eamonn

Not applicable

I too would like to hear if there is any update on this? Our server runs mud with preceding load whereas my development machine is much faster even though it is not anywhere near as beefy.

My guess is that the preceding load breaks some kind of load optimizing when using SQL, much like preceding load will brake an optimized load from a QVD.