SLT Replication for the same table from Multiple Source Systems

[Former Member]

Hello,

With HANA 1.0 SP03, it is now possible to connect multiple source systems to the same SLT Replication server and from there on to the same schema in SAP HANA - does this mean same table as well? Or will it be different tables?

My doubt:

Consider i am replicating the information from KNA1 from 2 Source Systems - say SourceA and SourceB.

If I have different records in SourceA.KNA1 and SourceB.KNA1, i believe the records will be added during the replication and as a result, the final table has 2 different records.

Now, if the same record appears in the KNA1 tables from both the sources, the final table should hold only 1 record.

Also, if the same Customer's record is posted in both the systems with different values, it should add the records.

How does HANA have a check in this situation?

Please throw some light on this.

[Former Member]

Hi Vishal,

The scenario you mention (multiple source to single target schema) is supported with HANA 1.0 SP03. However, I am not sure whether SLT will automatically take care of manipulating 2 tables with same name. With HANA SP03 SLT also delivers transformation capabilities that can be applied on the incoming tables from source and after modifications(in SLT) data can be replicated to the target HANA box. i.e. ABAP include programs can be written for data manipulation of incoming data.

However, I would wait for some SLT expert to comment if I am right.

Regards, Rahul

[Former Member]

Thank you for the reply Rahul.

I do understand that SLT with HANA 1.0 SP03 is equipped with various new possibilities - transformations, filters etc.

But, there has to be something in place for a condition such as above.

One thing that i can think of is: replicate the data from the same table coming from different source systems in different schemas in SLT. Then perform a transformation, store in a single table and then replicate the same in HANA in a single schema table.

I need more clarity on this though.

[Former Member]

Hi Vishal,

Definitely there should be a way in SLT to handle such scenarios (as this is one of the scenario in which replication from different sources to same schema helps)

I also expect SLT to handle the scenario you mention in the following way:

Setup replication from the 2 source to same schema in HANA. Handle the tables having similar names in SLT by doing transformation(event based/parameter based) and handle things you mention in your first post.

Note:This is my assumption as I have not tried the replication scneario you mention above.

Regards, Rahul

[Former Member]

Hi Vishal,

i suggest you to take a look to SAP HANA SPS03 Master Guide. There is a comparison table for the three replication technologies available (see page 25).

For Multi-System Support, there are these values:

- Trigger-Based Replication (SLT Replication): Multiple source systems to multiple SAP HANA instances (one source system can be connected to one SAP HANA schema only)

So i think that in your case you should consider BO Data Services (losing real-time analytics capabilities of course).

Regards

Leopoldo Capasso

[Former Member]

Hi Leopold,

Please check the "SAP HANA Installation Guide u2013 Trigger-Based Data Replication " guide located at:http://help.sap.com/hana_appliance , Page 11 which mentions:

"Mulitple source systems can be connected / configured to one HANA system using the same data base schema (N:1 relation)"

Regards, Rahul