Offloading cold and warm (archive) data from your HANA based SAP BW to the Cloudera Enterprise Data Hub (EDH) is one of the low-hanging fruits when it comes to cost savings.
SAP HANA is one of the most powerful in-memory data stores on the market. It is exceptionally well suited for the operational demands of ERP, CRM and SAP’s own Business Warehouse (SAP BW). In particular the size of SAP BW can grow tremendously even when organisations only keep a reasonable history of data points over the years. Of course, this history is intentional and serves to generate value from your data. In certain cases, regulatory requirements further determine how long specific types of data have to be stored.
SAP HANA’s performance however comes at a cost; a cost that is directly related to the amount of data it stores. So wouldn’t it be nice if there was a way to keep HANA storage needs level by shifting some of its older, less frequently used data to a low-priced alternative in a way that’s completely transparent to the SAP user?
There is a way to do exactly that! SAP BW has a functionality called NLS (Near-Line Storage) that enables offloading data to external databases. Cloudera partner Datavard provides Datavard OutBoard DataTiering which utilizes this functionality to securely move cold and warm data to Cloudera EDH in a manner that is fully SAP and Cloudera certified. This offload happens with full transparency to the SAP end-user and frees storage capacity in HANA for data that really requires the power of an in-memory store. Doing this can reduce the TCO for SAP BW substantially. By defining the borders between hot, warm and cold data (data that is frequently used, less frequently used and rarely used), the total amount of savings (i.e. data that is moved to the Cloudera EDH) can be adjusted as you go along.
Once the data is moved to the Cloudera platform, it is available to users and connected SAP BW tools as it was before and without change. Since the cold and warm data is accessed less frequently than the hot BW data (remaining on the HANA platform), there is no noticeable performance degradation.
Needless to say that this concept works for all Cloudera clusters irrespective of how they are deployed: On premises, in public or private cloud environments. Cloudera security, high availability and disaster recovery capabilities match the high requirements usually seen with existing SAP BW deployments.
In addition to the pure IT cost savings this provides, now that you have moved a substantial part of the archive data to the Cloudera EDH, the power of the Cloudera platform can be brought to bear on new use cases. Cloudera EDH is the perfect environment for preparing and combining data from the SAP systems (ERP, CRM, BW) with other data-sources. The result can then be used for advanced analytics and machine learning using the very same Cloudera deployment that is used for HANA offloading.
Some of the possible use cases that can be realized by combining data from SAP HANA with other data sources are:
Sensor data combined with data from the SAP ERP and BW systems can help optimizing production processes or increase quality, even in real-time (e.g. predictive maintenance).
Machine Learning algorithms can help identify and reduce churn within the workforce or help automate the recruitment processes.
- Customer 360
Data from the CRM system can be combined with external datasets or social media information to improve customer satisfaction.
- Price optimization
Combined with public data sources (market data, social media), machine learning model predictions can help find the optimal price for a product – also during initial product placement.
Which use case will you be tackling first?
About the Author
Guido Oswald is a Sales Engineer at Cloudera Switzerland. As a TOGAF certified computer scientist and MBA graduate, he enjoys architecting and implementing use cases around Machine Learning, Analytics and Big Data.