steps to compress Cube

[Former Member]

Hi Gurus,

I need full detail steps while doing cube compression, thanks Dave

[Former Member · Accepted solution]

Hi Dave,

Basically when you create a cube 2 tables created ,they are F and E tables.

When you load the data, data loaded into F table.

Compression is a process, to compress the data in fact table. Generally when you do compression all the request ids are deleted and data moved in to E table.

There may be data with same characteristic combination and with different Req ids, these records can be compressed.

For this,

Right click on Cube -> manage -> goto Collapse tab -> provide Request id ( defaultly it will take the latest successful request) -> Click on Release.

Then all the data moved to E table. Now you cannot delete data request wise.

If you want to delete data you need to do reverse posting.

I think this will helpful.

[purvang_zinzuwadia · Accepted solution]

hi,

adding to all the above comments, compressing cube is a critical and important decision. So go for it, only when u need it. Because request ids are deleted when u compress the cube, so after that you cannot retrieve data based on request ids.

Regards,

Purvang

[Former Member]

Dave,

What do you exactly expect on compression? If I can I can provide if it is not already answered.

Alex

[Former Member]

Hi,

basically 2 types of fact tables, F fact table and E fact table,

initially when we trigger the load the data will place into the F fact table when we say compression of cube the data will move from F fact table to E fact table.

steps involved in Compression of Cube:

context menu on cube>>manage click on collapse tab give the latest request id select the check box with zero elimination and then click on relese button.

hope this will helps you..

thanks & regards

Ramesh

[Former Member]

Dave,

There should a lot of info about compression available here. just keep few things in mind.

1. After compression request id's are removed so no longer you can delete a request, only selective deletion is possible.

2. It will increase the performance as the entries will move from one table to another F->E table and it is read optimized.

3. COmpress cubes can be done in the batch mode, using process chains - process type 'compress cube'

Hope this helps. Award points if useful.

Goodluck,

Alex.

[Former Member]

The purpose of the compression process is to delete the Request IDs.

The same data record can have multiple occurrences in the same cube due to different Request IDs.

This increases the data volume and decreases the performance while executing a query.

Lemme know if this helps.

-Bikas

[Former Member]

CUBE -> MANAGE -> Go to COLLOPSE tab

Request ID -

---

F4 (Select the latest record)

With zero elimination. (Select)

Click on RELESE

Go to REQUEST tab and see ‘C’ is checked.

When we select "with zero elimination" then it deletes all records which key figure value is zero after compression.

After compression we can’t delete the data based on request number

thanks.

Wond

[Former Member]

Hi Dave,

Check these links.

For BI 7.0

http://help.sap.com/saphelp_nw70/helpdata/en/ca/aa6437e7a4080ee10000009b38f842/frameset.htm

BW 3.5

http://help.sap.com/saphelp_nw04/helpdata/en/ca/aa6437e7a4080ee10000009b38f842/frameset.htm

Please search the forum, this is a very common topic, you will get lot of good stuff.

Thanks

CK