Roll Up & Compression

[Former Member]

Hi Gurus,

Would like to know more about Roll Up and (ii) Compression. Please help.

Thanks

Sheeja.

[pruthvi_r]

Hi,

RollUp:-

It is used to update from the cube to aggregates. you load data into cube then create aggregate. after sometime again you load some more data into cube, then you do rollup to update the new data in cube to aggregates.

Compression:-

Each time you load data into cube, they are stored in different request. if you load data 10 times, 10 diff requests are created. You compress the cube data so that all request are combined into one single request, so that data access is faster.

hope this helps

regards,

Pruthvi R

[Former Member]

Sheeja,

Roll up is a delta mechanism for aggregates, once after you do roll up then the aggregates will be updated with the new data.

Compression is the mechansim where it will compress all the records based on the Primary keys... i.e, when you do compression then the request id's will become zero..

Assign Points if it helps...

Thanks,

Gattu.

[Former Member]

Hi

For an info cube there will be two tables.

1.F-Fact table----

---

Uncompressed table

2.E-Fact table----

---

Compressed table

F fact table contains all the information with respect to Request id.

Compressing means removing the request id information by aggregating the Key figure information.

Roll up :- Filling up the aggregates with data.

After loading the info cube, for updating the data in the aggregates we will do Roll up.

Compression and Roll up we will do for Improving the Query performance.

Assign points if it helps.........

[Former Member]

Hi,

Have a look at the below threads