The fix? Rebuild your CompositeProvider as a HANA Calculation View directly in the HANA Studio (or XSA). Then consume it in BW via an External View.
Page 28 would show you the dark art of the 窶� specifically, how to convert your cube to "cube merge" mode and enable INMEMORY_AGGREGATION . sap bw 7.4 practical guide pdf 28
That is the legacy of page 28. It wasn't just a guide. It was a warning: Respect the database, or the database will humble you. The "28" in your search isn't a version number. It窶冱 a reminder that the deepest knowledge is always hidden in the appendices and the troubleshooting sections窶馬ot the glossy introduction. The fix
Why? Because HANA窶冱 optimizer relies on fresh statistics. If your stats were from the last system copy three months ago, HANA would generate a brilliant execution plan for a dataset that no longer existed. You窶囘 see a query take 12 seconds that should take 200 milliseconds. Page 28 would show you the dark art
In BW 3.5 and 7.0, your fact tables (F-fact tables and E-fact tables) were designed to minimize disk I/O for row-based databases like Oracle or DB6. But on HANA, row storage is poison. It destroys parallelization.