QlikView 11 - Initial Report load is taking too much time

[#1]

Hello everyone,

We are using QlikView 11.

In our data model we have used LinkTable between 2 facts and using AutoNumberHash256 with 4 Keys.

The data volume is 50 million records in 1st fact and 10K records in 2nd fact.

When we are opening the QlikView either with "Home Page" or "Page with Charts", it is taking 13~15 secs and 20~23 secs respectively.

We need to bring down the Reprot open time to 5~6 secs.

Please help/advice.

Thanks in advance.

Regards,

Ritam Dhara

[Gysbert_Wassenaar]

Hard to tell what is causing the performance problem without looking at the document. Try reading this discussion and see if you can apply some of the guidelines and tips.

[vikasmahajan]

Hii,

Go to Task Manager and check for memory utilization how much memory is consumed by

your qlikview server and if you are using large amount of data whether incremental load is

implemeted at your end if not you could thing for incremental load.

*** Storing data into qvd is much faster

Vikas

[#4]

I have had this problem when the duplex setting on my network card was not set correclty for my network.

[#5]

Hi everyone,

Thanks for the above replies.

Currently we are loading QVW from QVDs having Link Tables and using AutoNumberHash256 on 4 key columns.

We are concerened about the QVW Application load which is taking 13~15 secs.

Can anybody suggest what kind of QVW datamodel is good w.r.t performance where there are 2 & more facts with common and unique dimensions.

Regards,

RD

[#6]

Hi everyone,

Thanks for the above replies.

Currently we are loading QVW from QVDs having Link Tables and using AutoNumberHash256 on 4 key columns.

We are concerened about the QVW Application load which is taking 13~15 secs.

Can anybody suggest what kind of QVW datamodel is good w.r.t performance where there are 2 & more facts with common and unique dimensions.

Regards,

RD

[#7]

Hi everyone,

Thanks for the above replies.

Currently we are loading QVW from QVDs having Link Tables and using AutoNumberHash256 on 4 key columns.

We are concerened about the QVW Application load which is taking 13~15 secs.

Can anybody suggest what kind of QVW datamodel is good w.r.t performance where there are 2 & more facts with common and unique dimensions.

Regards,

RD

[#8]

Hi everyone,

Thanks for the above replies.

@Vikas - Currently we are loading QVW from QVDs having Link Tables and using AutoNumberHash256 on 4 key columns.

We are concerened about the QVW Application load which is taking 13~15 secs.

Can anybody suggest what kind of QVW datamodel is good w.r.t performance where there are 2 & more facts with common and unique dimensions.

Regards,

RD

[vikasmahajan]

Dear RD,

Any Specific reasons for using autonumberhas256  ?  we are using  autonumber for generation keys

as per my knowledge it is faster with link table .

for example

	AutoNumber([Customer No_] & '_' & Date([Posting Date])
	& '_' & [Document No_] & '_' & [Global Dimension 1 Code]
	& '_' & [Global Dimension 2 Code]) AS %CLE_Key

Link Table:

  LOAD DISTINCT

  AutoNumber ...

Hope this helps you .

Vikas