Forcing Hana queries to use more processes and cores

I have found a few instances where Hana decides to usea  single core to process queries, even though they take quite some time. I’ve noticed this most with queries involving grouping.

I think it might be to do with the engine selected by the optimiser, but I’ve also found that adding the following hint to SQL can radicall reduce run times of some of these queries (and use many more cores to do so):

      with hint(OLAP_PARALLEL_AGGREGATION)

There is some limited mention of this in documentation, and also a SAPNote to apply this to queries created by BI on Hana. Hopefully the optimizer will get smarter as time goes on, and this kind of thing won;t be needed!

Jon