cancel
Showing results for 
Search instead for 
Did you mean: 

Load the product of two ODSses to a new ODS (using an infoset as datasource

Former Member
0 Kudos

Hi All,

I am working on a project and can't seem to find the sollution, I hope you can help.

What we want to do is load the product of two ODSses to a new ODS. We have on ODS with customer data and one ODS with financial postings. Some customers have received a posting and some have not, and we want to see all customers with all postings in one view. I have created an infoset with a left outer join and the query works correctly. Now I want to use this data further on in the flow, so load it to a new ODS.

Do you know if there is a simple way of loading the product of two ODSses to a new ODS?

or use an infoset as a datasource for a new ODS?

I hope you can help me with this!

Kind regards,

Hermen

Accepted Solutions (1)

Accepted Solutions (1)

Former Member
0 Kudos

Well that won't really work as soon as you get into production volumes

Are you plannign to have in the new ODS a zero value row for EVERY customer for EVERY day?

At my site that will generate a ODS of 730,000,000 records per year

There are better ways of doing this - ie using a multiprovider of cube to customer infoobject (using the rowcount in the infoprovider to display the slowly moving dimension) or in version 7 using query designer to run a query of the cube with master data values and not posted records

Former Member
0 Kudos

Hi Simon,

Thank you for your reply. In our case the data set is much much MUCH smaller and on a different time basis.

Thanks again,

Hermen

Former Member
0 Kudos

Well I would still go for the delivered options (ie the multiprovider or the bex 7 option - although I prefer the multiprovider one)

If you really wanted to do it your way then yes infoset will work with a outer join - I dont like it because you end up with a fragmented data model

Or if you are trying to snapshot them then APD with the multiprovider above will work (or readign master data as well) - or you could create a generic datasourcetype function module which has inside it a infoprovider read function module to read the cube and then do what you want

(in volume situations APD falls down with severe memory problems - most of us then use the infoprovider read option)

Answers (1)

Answers (1)

Former Member
0 Kudos

If your query works correctly, you can use the query as a datasource in the Analysis Process Designer (txn RSANWB), and populate a downstream ODS.