Application vs Database Source Endpoint for Replicating SAP ECC Tables

[Abhi_719]

We are trying to some understanding on performance and overall impact of using an Application  Connection vs Underlying SQL Database source connection for Replicating SAP ECC tables. We are in process of Replicating SAP ECC tables and are figuring out the best approach for the task. We understand Replicate can utilize database logs if we connect through underlying SQL database and probably have less impact on the SAP Application system. But we do see differenced in datatypes and missing cluster/pooled tables in db compared to Application tables. 

Any inputs  or guidance is appreciated.

[OritA]

Hi,

First we would recommend to review the Replicate user guide for supported data types in SAP endpoint. If this does not provide you the answer, please open a case for this question and include in the case the exact data types that you need so we can provide you the relevant recommendation. 

Regards,

Orit

[Abhi_719]

Hello,

We will review the user guide. However, my question is more towards performance impact of extracting via SAP Application Endpoint vs underlying database connection.

[Steve_Nguyen]

in term of performance , you would have to test in your environment, as we do not have number extracting via SAP Application Endpoint vs underlying database connection.

[Abhi_719]

So it is sounding like there is no recommended best practice by Qlik for replicating the SAP tables in terms of performance, maintenance,  Change data capture, source system impact.

[Steve_Nguyen]

@Abhi_719 

performance tuning , best practice are handle by our Professional service team, best you engage with your account manager for some Professional service.

[huyhuynh]

@Steve_Nguyen When connecting to an SAP application table with Replicate...are changes to the data still using the transaction logs (SQL Server database) for change detection? Maybe it's not the right wording for the question...but just trying to understand the system resource impact of replicating an SAP application table versus the underlying database tables.

Thanks in advance!

[Bill_Steinagle · Accepted solution]

Hi,

Thank you for the post to the forums about the SAP Source Endpoint. As a best approach for the SAP ECC environment you will be using Replicate with it would be best to use the SAP Application (DB) Endpoint as this is how Replicate processes the Load and CDC features with the Product. The Application layer SAP is used and when this Endpoint is setup you also define the backend DB which is SQL Server in your setup. The process does access the underlining DB and would be the recommended connection for access to this SAP ECC environment.  Unlike the other 2 Source Endpoints SAP Application and the SAP Extractor. The SAP Application Endpoint uses RFC calls to the SAP environment with the added load on the SAP environment. The SAP Extractor also adds load to the SAP environment since it uses intermediate Tables that are temp Tables defined in SAP to process the data.  The load on the Source SAP environment is less since we are processing the data via the SQL Server DB.

Regards!

Bill