The example shows a monthly overview, with a peak of memory used that is not normal for daily average memory consumption of HOT HANA memory.
A further drill down to evaluate which table / schema caused that peak. As shown in next figure, reveals that the SAP table RSMONMESS caused the memory peak on January 22nd.
The insights provided by this and other views created by our Team of HANA experts from our project experience and hands on analysis can help you to detect abnormal behavior e.g. after system or DB patches. The information available and leveraged to you through calculation views can unveil unwanted tables, such as PSA based data, not included in dynamic tiering or early unload settings.
Other Tips:
You can manually also spot large objects not wanted in HOT memory, or causing memory bottlenecks, using the Non-Active Data Monitor in the SAP HANA Database in the RSA1 – Administrator Workbench:
Administration -> Monitors -> Active/Non-Active
You will find further tips in SAP Service Marketplace Note:
1776749 – Activating the Handling of Non-Active Data During Main Memory Bottlenecks