Performance issue

[Anonymous]

Hi,

I have fact table which is having almost 200 millions of records and almost 12 dimensions joining with with this table. With tables prepared data model and dash board. It is taking long time to open dashboard and charts.

Is there any way to increase the performance of loading records into dash board.

And also how to increase(taking long time to open) the opening time of charts.

Please share any thoughts on this issue.

Regards,

Kumar

[Chanty4u]

1.Try to avoid  complex expressions in front end.

2)Create a flags in back end and call in front end expressions.

3)which the fileds not required in front end  or back end just  drop the fileds .

[Chanty4u]

Meanwhile look into this

Re: QV App performance

Re: Performance tuning in QlikView?

[Anonymous]

Hi Chanty,

Thanks for responding.

Here I am excluding unused columns and I don't have any long keys. It is simple joins directly with one to one.

My application size is almost 50 gb. Here is the problem

[Chanty4u]

Try to reduce ur application size..

chk with ur RAM usage and

Re: High memory usage (RAM)

RAM & CPU in the system

[Colin-Albert]

This post may be a few years old but the information in it is still relevant.

*** 6 Weeks in to QV Development, 30 Million Records QV Document and Help Needed!!! ****

Thanks to swuehl‌ for bringing the post to light!

*** 6 Weeks in to QV Development, 30 Million Records QV Document and Help Needed!!! ****

[HirisH_V7]

Thanks for info Colin_Albert‌.

[manojkulkarni]

You could try to avoid time stamps if not being used. Use only date part of it.

Try to make aggregates for dimensions,if details are not required

[engishfaque]

Dear Kumar,

First, I would suggest you don't use "Aggr Function" on UI Level.

Second, try to use N-Tier Architecture (i. Extraction Layer, ii. Transformation Layer, iii. Visualization or Presentation Layer).

Please note, 20 million records are not more to degrade your dashboard performance. I think you are doing all of the things on a same QVW and not dropping tables after use.

Kind regards,

Ishfaque Ahmed

[#10]

Hi Kumar,

Simple things you can do are to avoid showing frequencies in list box and sort based on numerical fields and using keeps instead of joins.