1. Why do we need to use the Program RSDRI_CONVERT_CUBE_TO_INMEMORY ( to convert a standard Cube to HANA ) when, by “default" the Cube is “in memory ? ( I mean, when we create a new cube in HANA or even when we transport from BI 7.0 to BW 7.3, I think its in Memory)
2. Can also please let me know if I need to Migrate Std DSO’s to HANA" In memory DSO’s". or are they not in “In memory” by default again? I mean do I still need to convert them? if so, whats the procedure?
When migration of info cubes into in-memory, they go back to a classical star schema model as opposed to an extented star schema. The master data SIDs in the fact table still connect to their corresponding master data tables (info object) like a primary key - foreign key relationship except that it doesnt have to go via the dimension tables.
You can get more detail here;
Yes, if you want to take advantage of the performance increases that HANA offers, you will need to convert them to HANA optimized cubes or dsos. After the db migration, yes, all objects are in memory, but the structural changes (eliminate E fact table, elimination of the dimension tables, etc) that take place during the conversion are the root of the performance increases which come with BW powered by HANA.
After you've done the DB migration to HANA you could simply leave cubes and dsos as they were in the RDBMS environment and everything will work fine. But if you do this, you will not get the performance that HANA is built to deliver. Additionally, in a BW powered by HANA environment, each new cube or dso created will be created the HANA optimized way.
SAP Customer Solution Adoption