BO Designer - What's is the Guidelines to Build Universe with Better Performance? R Performance tuning Issues of U

GeekInterview.com > Interview Questions > Data Warehousing > BO Designer

Go To First | Previous Question | Next Question

BO Designer | Question 65 of 116

What's is the Guidelines to Build Universe with Better Performance? R Performance tuning Issues of Universes?

Total Answers and Comments: 4

Last Update: August 07, 2007 Asked by: narendra_tatineni

Sponsored Links

Best Rated Answer

—

No best answer available. Please pick the good answer available or submit your answer.

April 30, 2006 23:56:48

Ravi

RE: What's is the Guidelines to Build Universe with Be...

Guidelines are provided in the Universe guide:

Here is the Summary:

1) Modify the array fetch size in SBO file

2) Assign table weights and change the default order of the tables in a query by changing parameter in PRM file

3) Use shorcut joins to avoid unnecessary tables that may come in Query

4) Use aggregated tables

Thanks

Ravi

Is this answer useful? Yes | No

May 11, 2006 07:49:20

venkat

RE: What's is the Guidelines to Build Universe with Be...

adding to the answer is

1. see that ur desgin schema is not having any loops,cartesain products,fan traps,chasm traps etc that will give u the wrong data and the performance might be a problem is those cases.

Is this answer useful? Yes | No

February 16, 2007 00:18:42

Tushar

RE: What's is the Guidelines to Build Universe with Be...

I agreed with Ravi what he said except his 4th point. Short cut join privide shortest path and and gives result in shortest time but Business Objects itself suggest using shortcut join is not good practice.

Is this answer useful? Yes | No

August 07, 2007 07:01:13

Muthukrishnan

RE: What's is the Guidelines to Build Universe with Be...

To enhance the performance of the universe, we have to do the following:

1) Keep universe light,
2) Deselect multiple SQL statements for each measure options,
3) Try to avoid outer joins,
4) Use off-line mode if possible,
5) Remove unnecessary list of values
6) Use aggregations if possible and
7) Take off all variables and objects.

Cheers.

Is this answer useful? Yes | No

Go To Top