Data Federator and Information Design tool

[Former Member]

Hi,

First, Can Data Federator be a single data source for Information Design Tool?

Second, from performance point of view, using Information Design tool to build multi-sources Universe or building Universe on Data Federator, which way is better?

Thanks

[Henry_Banks · Accepted solution]

Hi,

let me explain :

in BI 4.0, the Data Federator Query Server engine is now part of the server-side Adaptive Processing Server .

You only invoke this architecture when you select "Multi Source Enabled" - any other entry point is NOT using the Data Federation Service.

So - if you select SingleSource - that's not using DF.

However, it is entirely possible to select "Multi Source Enabled" then only use 1 connection in the MSU !!

Having said the above, Your second question doesn't make sense.

Unless you mean, using the Data Federator middleware driver in BI 4.0 IDT, to connect to an XI 3.1 Data Federator Query Server? (i.e. consuming the 'old source' in the new enviornment)

This backwards compatibility was delivered to facilitate transition of DF customers, especially since there is no standalone DF product in BI 4.0 (yet)

I trust this is of use to you.

Regards,

H

[Former Member · Accepted solution]

Hi Henry,

Thank you for your reply.

I got what you meant, let me explain a little bit more of my question:

We are going to build brand new Universes, we have multiple data sources, we have BO 4.0 environment, we also got Data Federator.

Since Information Design Tool(IDT) doesn't have full Data Federator functions, like data mapping and target tables, so we are thinking of merging the data sources in Data Federator and then as you said using the Data Federator middleware driver in BI 4.0 IDT(JDBC) to connect to an XI 3.1 Data Federator Query Server. Will the performance will be better this way? But according to what you said, this seems not a recommended way for building new Universe.

How can i achieve data mapping and target tables in new BO 4.0 environment.

Thanks a lot.

[Henry_Banks · Accepted solution]

Hi,

Sorry, I'm not aware of any performance comparisons / benchmarks between the 2 approaches. But i'm sure you will be able to ascertain this after a short test.

I'm not making any assurances about what's better / what's recommended . it's a case of "horses for courses", if you excuse the idiom.

Some of the DF features missing from BI4.0 IDT include;

- mapping rules

- XML and Web Services connections

- Table filters

- Pre-filters, post-filters

Some of those functionalities might appear in BI4.1 (end 2012) but I don't have visibility of the roadmap, and there is no commitment..

Best of luck

Regards,

H

[Former Member · Accepted solution]

Henry, i am just afraid that at one point SAP will stop support DF Universes.

That will be scary.

[Henry_Banks · Accepted solution]

i doubt that very much since functionality is being ported into the core semantic layer, and backward compability is available whilst they address the fit-gaps