Capacity management and performance optimization are complex challenges that still involve costly and time-consuming activities. As capacity planning processes are barely automated, any changing requirements result in heavy delays, media disruptions, and manual tuning efforts. To face these challenges, the Fujitsu Lab Magdeburg designed a knowledge base for capacity planning which is powered by SAP HANA utilizing various machine learning techniques. In a recently published scientific paper, the Fujitsu Lab Magdeburg presented the underlying concept of a domain-specific application performance management (APM) knowledge base evaluating the feasibility and highlighting the benefits of this approach. The main findings drawn from this research can be summarized as follows:

On the basis of SAP HANA, measurement data of various SAP landscapes for model training and extended analytics can be integrated in a central knowledge base addressing capacity management challenges in real-time. In the course of this, simulations can be carried out on the fly and allow to quickly evaluate “what-if” scenarios providing an on demand data-driven decision support. Storing performance-related data of various SAP landscapes in-memory allows not only to identify industry standards but also to provide benchmarking features across SAP systems and customer landscapes. This service is technically feasible thanks to a seamless integration of SAP HANA with, e.g., open source components such as an R server, and Java web applications. On this basis, a modern and highly responsive SAP Fiori-based user interface can be applied to retrieve various analytics and simulations as self-service tasks.

An in-depth discussion and evaluation of the approach can be viewed here.