I have a cube which contains 15 to 20 members including the calcuated
members. I would like to know the following
1. Do I need to verify all possible combination (verifying all combinations
will be a tedious process) in cube in order to assure the cube is working
without any problem ?
2. If the answer to the question 1 is "No" then what is the best way to
verify?

Let me just re-phrase the question to make sure I understand. I think
you are asking if you need to filter by every dimension and check every
figure. If this is what you are asking, the answer is No, unless you
have custom rollups you should not need to test test too far down the
dimension hierarchies.

What I usually do is to do a select over the fact table grouping by
something that can easily be verified. For example if you were doing a
sales cube over 2 years data you should know that the fact table
contains 3000 items that were sold this year and 4500 items that were
sold last year. You should also be able to verify these figures match
exactly with the ones in the source system (or if the don't that you can
account for any discrepancies).

Then it is a simple matter of opening up the cube and viewing sales by
year and you should see the exact same figures. If you are seeing the
correct totals at the top level of the cube you can be sure that every
row from the fact table has been correctly processed into the cube.

Typical problems that occur with cubes are that there are either
duplicate entries in one or more dimensions, resulting in increased
figures. Or if you do not have referential integrity between the fact
and dimension tables, there can be facts with out matching dimension
members which results in lower figures.

--
Regards
Darren Gosbell [MCSD]
<dgosbell_at_yahoo_dot_com>
Blog: http://www.geekswithblogs.net/darrengosbell

If there are rollups and downs do i need to verify all the possible
combinations ?

"test test" wrote: