Hi Lars,
If the result set is cached, it most likely be in aggregated form vs. the base data in memory being more granular. I'm assuming the cache would avoid the need to aggregate the granular data and utlimately resulting in faster query execution. So wouldn't this free up workspace memory during query execution as the "aggregaton" work wouldn't have to be executed each time? If many users are running the same query, couldn't the workspace consumption for aggregation be large?
Also, can the OLAP cache on the BW application server also provide the same cache functionality?
Regards,
Dae Jin