How Aggregates work in cube?

[Former Member]

Hi All,

I have created Aggregates for a cube in SAP BI. Please explain how it works and how it enhances performance. How does it store data in it?

[Former Member · Accepted solution]

Hi,

Aggregates basically can be referred to a smaller cubes to the Main Cube. When we create an aggregate based on the fields (F1, F2, F3), it creates a small cube with the fields (F1, F2, F3). When we perform aggregation operation with option Rollup, the data from the main cube will sits into the smaller Cube (Aggregate what you have created) with the Fields (F1, F2, F3), the date will be aggregated by comparing all the fields are similar.

When we execute the report with the fields F1, F2, F3, (Query to be created with the fields F1, F2, F3), the data will get from Aggregates (Smaller Cubes). As the Aggregates are already consists of aggregated data, the report will not perform the aggregation operation at the report level. So, the Query Performance is optimized here.

Hope this helps you.

Thanks and Regards,

Mahesh

[KodandaPani_KV]

HI,

Aggregates are basically will create the imporving the performance.

it is the like mini cube or baby cubes and it is also same strucutre of info cubes.

let say example

in cube your are using the 50 info objets but in report your are using 20 info objets then only 20 info objects will create the aggregates.

select the cube -> right click -> maintain aggregates -> fill the aggregates and deactivate the aggregates it your depends on the requirments.

after performing the Roll up data move form standard cube to aggregates cube.

if not roll up done data will not move form standard cube to aggregates cube.

Roll up mandatory funciton you have created the aggregates.

after maintain aggreates -> select goto menu -> aggregate data -> now you will find the aggreagate data.

each aggregates generate table like E1007,E1008.

hope it will help.

Thanks,

Phani.

[former_member489487]

Hello,

The aggregates are multidimensional data structures with similar structure that of an Infocube. When you are using aggregates, the data volume for each query is reduced considerably, as the data is stored in a compressed form. This in turn helps in improving the performance for queries.  You can check the table RSDDAGGRDIR to check the directory of Aggregates.

Regards,

Rohit

[former_member183519]

Hi ,

Aggregate is nothing but Baby Cube. Main purpose of aggregate is to make reporting faster.

suppose your cube have 100 char & KF , but in reporting you want only 10 .

So in such case you will create aggregate based on this 10 fields.

Now the question is how data filled/store for these 10 fields???

so, to feed data to aggregate   , Rollup comes into picture.

if any new request comes to cube, to load that records in aggregate we need to rollup.

When your query executes, then for fetching data, query always hit first aggregate....

if it found aggregate then it returns data to query immediately,

if query did not found any aggregate then it will hit cube.

Regards,

Hitesh Gavande