# how to decide dense & sparse dimension?

how to easily identifed dense & sparse dimension
###### Who is Participating?
I wear a lot of hats...

"The solutions and answers provided on Experts Exchange have been extremely helpful to me over the last few years. I wear a lot of hats - Developer, Database Administrator, Help Desk, etc., so I know a lot of things but not a lot about one thing. Experts Exchange gives me answers from people who do know a lot about one thing, in a easy to use platform." -Todd S.

Commented:
The idea of Dense and Sparse dimensions is quite lengthy and unfortunately there is no easy answer to the question.  As a general rule, you would start off making your accounts and time dimensions dense.  This usually makes sense because it is likely that a large number of your time periods will have data across all your other dimensions, as will your accounts (measures).  From there you have to look at your block size, which is made up of the combinations of your dense dimensions.

So for example if you have 12 stored members in your periods dim and 200 stored measures in your accounts dim then your block size would approx 19K (12 x 200 x 8 bytes)

Optimal block size for 32 bit OS is between 8k and 100k with 64 bit OS, you can go larger, but you have to test and tune.

One good way to test which are your "densest" dimensions is to make a single dimension dense, load your data and run a default calc.  Then go into database properties and look under statistics and check block density.  Write it down, then clear the database, make a different dim dense with all others sparse and repeat load, calc, and check stats, again write it down and repeat for all dimensions.  This will give you an indication of which are the sparsest and which are the densest dimensions.

What you have to realize is that even after you are armed with that information, you still have to take a number of other factors into consideration.  Just because a dimension is the densest, doesn't always mean it will be a good dense dimension.  I know it is confusing, but tuning Essbase cubes is more art than science.  There are some basic rules to start out and then you have to test, test, and test.  It also makes a difference what your goals are.  I might make certain dims dense and others sparse to speed up calc time, but I might make other dims dense and sparse to get better retreival time.  So it depends on what the requirements are and it is usually a balancing act of what works best.

As I stated in the beginning, try to start off with Time and Accounts dense and go from there.

I would also recommend a great book called "Look Smarter Than You Are with Essbase 11"
http://www.lulu.com/content/paperback-book/look-smarter-than-you-are-with-essbase-11/6402010
(full disclosure: while I am not the author of the book, I did contribute content to the publication)

Hope this helps.

0

Experts Exchange Solution brought to you by

Facing a tech roadblock? Get the help and guidance you need from experienced professionals who care. Ask your question anytime, anywhere, with no hassle.

###### It's more than this solution.Get answers and train to solve all your tech problems - anytime, anywhere.Try it for free Edge Out The Competitionfor your dream job with proven skills and certifications.Get started today Stand Outas the employee with proven skills.Start learning today for free Move Your Career Forwardwith certification training in the latest technologies.Start your trial today
DB Reporting Tools

From novice to tech pro — start learning today.