SAP Knowledge Base Article - Public

3784614 - Replication Flow from ODP Sources: Performance Analysis and Support Scope - SAP Datasphere

Symptom

One or more of the following symptoms are observed when replicating data from an ODP-based source to SAP Datasphere:

  • The initial or delta load takes longer than expected.
  • Data is available in the source system or Operational Delta Queue, but processing in the Replication Flow is delayed.
  • The Replication Flow completes successfully, but the achieved throughput does not meet the expected processing window.

Environment

  • SAP Datasphere
  • ABAP-based source system

Cause

In an ODP-based replication scenario, the total runtime consists of several processing stages, including data extraction by the ODP provider, creation and processing of ODQ requests, data transfer from the ABAP source system, and processing in SAP Datasphere.

A long overall runtime does not necessarily indicate a defect in SAP Datasphere. The bottleneck may be located in the source extractor, the ABAP application server, the Operational Delta Queue, the network connection, the Replication Flow configuration, or the available SAP Datasphere capacity.

Performance optimization across these components requires workload-specific analysis and may require iterative configuration and sizing activities.

Resolution

Performance optimization normally requires interactive analysis, testing, measurement and adjustment. This type of interactive optimization, capacity planning and landscape-specific tuning is a consulting activity. It is distinct from the correction of a reproducible produce defect through a support case.

Step 1: Identify the stage where the delay occurs

Use transaction ODQMON in the ABAP source system to verify the relevant provider, queue, subscription, and request.

Step 2: Validate the source extraction performance

Where applicable, use report RODPS_REPL_TEST to validate the ODP provider independently of the Replication Flow. If the extraction is already slow when tested directly in the source system, the source extractor and source-system resources must be analyzed first.

Step 3: Check ODP package size and partition configuration

Last but not least, SAP Support can continue the investigation when the collected evidence indicates a specific and reproducible malfunction, for example:

  • SAP standard delivery extractor extraction performance issue (which needs to be checked by BW-BCT-* component)
  • An ODQ request or subscription is not created or processed as expected.
  • The Replication Flow does not consume data that is available for the relevant subscription.
  • Processing remains blocked or fails with an error.
  • A documented configuration or product function does not behave as described.
  • A reproducible performance regression is observed after a product or software change.

For Support issues feel free to raise a Case. When creating a Case, follow the recommendation from KBA 2854764 to ensure SAP can get a solution to you in the shortest possible time.

Requests to determine the optimal custom extractor configuration, achieve a customer-specific throughput target, size the source system, or perform iterative end-to-end performance optimization require landscape-specific consulting. In this case customer can contact SAP Account Executive to request SAP consulting service, which can support this topic.

Keywords

Datasphere, Replication Flow, ODP, Operational Data Provisioning, ODQ, Operational Delta Queue, ODQMON, ODP SAPI, ODP BW Context, ODP DataSource, ABAP source system, slow replication, replication performance, low throughput, initial load, delta load, extraction performance, delayed replication, ODQ subscription, ODQ request, source extraction, RODPS_REPL_TEST, performance analysis, performance tuning, consulting , KBA , DS-DI-RF , Replication Flow Runtime , DS-DI-ABA , Datasphere - Data Flow / Replication Flow ABAP Integration , BC-BW-ODP , Operational Data Provisioning (ODP) and Delta Queue (ODQ) , Problem

Product

SAP Datasphere all versions