Hi all,
I just wanted to question the lack of memory management in HANA. Please correct me if the premise of moving data onto some other storage itself is wrong/infeasible.
Consider a SAP HANA instance where a lot of data is entered continually out of which only the recent data is of use. For the older data (that also might be required for instance to perform Predictive Analysis) there is not much processing and it can be safely moved to some other data store that is not as fast as hana instance. I specifically wanted to ask this due to emergence of hadoop based M-R processing that can be applied over the database to speed up things, but would require more memory (to account for multiple reads/writes) due to the fact that performing PAL over big data on Hana quickly is not feasible.
Kindly inform me if such compacting of memory held by Hana is possible?