Essbase Knowledge base

Hi Guys:
Please find some of the interview questions :

The BSO architecture like

• How blocks are built


• The impact of block size on calculations


• The impact of number of blocks on retrieval

The interviewer looks for real life scenarios not the bookish answers. For example if he asks “Tell me about blocks in Essbase”:

DBAG says:
"Essbase creates a data block for each unique combination of sparse standard dimension members (providing that at least one data value exists for the sparse dimension member combination). The data block represents all the dense dimension members for its combination of sparse dimension members."

You should answer the question not like you memorized the definition of Block. You are good if answer the question like this :

“In BSO the blocks are created for every combination of sparse standard dimension member provided data existed for that combination.”

In addition you should add the following when you answer the question on blocks.

“Generally Blocks are in compressed state on disk. During calculation they are pulled in cache and expanded. That’s why Block size is very important. If the block size is huge, only few blocks are pulled onto cache at a time and if a calculation requires 10,000 blocks to finish calculation it takes forever to finish calculation and if block size is so small and the index file becomes huge; Essbase takes forever to search for the blocks it needed like needle in the stack of hay.”

This answer shows your confidence on subject and also makes interviewer to think that you have real time experience.

2.What are the different storage options in Essbase?

There are two different storage options in Essbase. Previously there used to be only one BSO. With the version 7.0 ASO came into place.

Deciding which storage option to use depends on the business requirements. If business don’t care about write back option or complex calculations like allocations, currency conversions etc and need fast aggregation then ASO is the answer.

In Contrast if the business requirements need write back functionality and complex calculations like inter company eliminations, allocation based on percentages etc BSO is the answer.

Because of the above reasons, Hyperion planning need to go with BSO option.

Test on Sparse and Dense knowledge

I had business requirements gathered and built initial dimensions. We don’t sell all products in all cost centers. Can you suggest us the dimension types for the following dimensions? (This question is intended to test the concept of dense and sparse in BSO and outline optimization)

The second table gives you answer for the question.

The outline is arranged in hour glass structure.

Reason why we should arrange outline in HourGlass structure

Hour glass structure makes index to be more search freindly. Like assume that we have largest sparse dimension as the first dimension and then next largest in that order. when the essbase calculates it has to search all the largest sparse dimension and then to next largest which take more time.

In the hour glas sstructure it takes less time to search in index.