Request Compression in BI 7.0

Your replies would be helpful to me in establishing a compression strategy.

Guys who have been actually involved in compression would have better idea. I have done compression of an infocube myself.

When I was asking about tracing information i meant how would I be able trace a request if I have already compressed my data?

If you compress your infocube right after loading you need to have some mechanism to fix errors or reload erroneous request.

How are you handling compression? Immidiately after data load? Its true that when not all data is compressed you have to read data from two database table but is there a way around?

I dont think SAP designed this to copmpress requests immidiately after loading data.

Thanks all of you! Any more information will be highly appreciable.

Ram

Hi Vaishali,

indexes are data structure sorted values containing pointer to records in table.

indexes are use to improve data reading performance ,

but decreases data writing performance

every compressed data is taken from E table, every uncompressed data is taken from F table: also for this reason you can lost in performance if you have a lot of uncompressed data because the system has to read two different set of data .

So it's gud to compress cube afeter loading data.

It improves Query perfomance .

Hope this helps.

Reg

Pra