This blog post is a continuation of my previous blog post on Embedded analytics. Here I will explain about the Embedded BW with its architecture and use cases. There is a lot of misconceptions about the Embedded BW in S/4HANA. Many customers and non-BW consultants think that S4 Embedded BW could replace a standalone BW as an enterprise data warehouse solution. Though embedded BW supports all native functionalities of the BW system i.e. technically we can set up an enterprise data warehouse solution using Embedded BW, it is not recommended to use it as a full-blown EDW solution.
Embedded BW system was introduced in SAP ERP from NetWeaver 7.0 version to support operational reporting. So, this embedded BW can be found in all NetWeaver based systems like Solman, CRM systems. It’s one of the core components in Embedded Analytics. Below are some of the use cases of Embedded BW.
- Cross module operational reporting to support all day to day analytical needs of an organization.
- By architecture BPC need a BW managed schema to run, thus Embedded BW serves as the base for BPC Optimized for S4/HANA.
- Financial consolidation in S/4HANA is carried out through Real-Time Consolidation in older versions and Group Reporting in newer versions. These solutions run on top of Embedded BW.
- ETL Scenarios. These requests should be handled with great care since it is not recommended to use Embedded BW for data warehouse scenarios.
Reasons to not use Embedded BW for Enterprise Data Warehouse needs: (will add more as I come across)
- SAP Best practice is to not persist more than 20% of overall data volume in Embedded BW. The persisted data may include,
- Data from the planning application
- Data from own SAP applications.
- Data from external applications used for planning and consolidation
- Data from external applications for reporting purposes.
- Embedded BW is of version BW 7.5 powered by HANA and probably won’t be upgraded to BW/4HANA. Since all new innovations will be introduced in BW/4HANA only, it’s always better to maintain a separate Data Warehouse solution.
- OLTP(S/4HANA) systems are built with keeping Scale-up scenarios in mind. But the OLAP system should support scale-out to accommodate ever-growing data volume and processing needs.
- Apart from the scale-up / scale-out scenarios, there are a lot more characteristics disparities between the OLTP and OLAP systems as below.
- Resource management such as processes, background jobs, data tiering, etc
- Authorization concepts
- Backup and restore
- Frequency of upgrades and patches
- Maintenance activities
- Clustering mechanisms to support high availability and disaster recovery
- Embedded BW component is shipped along with the S/4HANA application with no additional license cost. The BI contents can be activated like the standard BW activation.
- However, the additional components like BPC Optimized for S/4HANA, RTC, Group reporting which uses Embedded BW needs a separate license.
Client settings for Embedded BW
Embedded BW can either be set up in the same client as S/4HANA or a separate client. This can be maintained in the RSADMINA table. Choosing between these options is based on the use cases of Embedded BW. When BPC / RTC / Consolidation is to be used, it is better to maintain the BW in separate client to avoid performance issues. Same way for ETL scenarios where the data is persisted in BW it is advisable to maintain BW in a different client. If the embedded BW is to be used only for the standard use case of Embedded analytics (operational reporting) without much data persistence, it can be maintained in the same client as S/4HANA.
Embedded BW Architecture
In general, even though we are provided with a powerful data warehouse framework to build an EDW solution it is not advisable to do so. It is always beneficial to maintain a standalone data warehouse for enterprise analytics needs. In this way, we can house kind of future requirements without affecting the transaction system.
Kindly correct me if my understanding of these concepts is wrong and please add your points in the comments section.