1974423 - SAP HANA SPS 07 Database Revision 71
CAUTION:
- Due to an error in the Calculation Engine optimizer, you might get not correctly aggregated data in some cases of in-memory planning scenarios.
As a workaround set the following parameter within indexserver.ini (section calculation engine):
opimize_filter_remove_attributes = 0 - Wrong data result are returned under the following conditions:
- There is an equality condition between a grouping column and a constant (e.g. sclas = 'new').
Note that this should be over a grouping column.
- There is an outer join. And this column belongs to:
1) the right-side table for the left outer join or
2) the left-side table for the right outer join or
3) any side for the full outer join.
Example of the query pattern where this occurs:
"select a.objid, a.otype, k.sclas
from a left outer join
(select otype, sclas
from t2
where sclas = 'new'
group by otype, sclas
) k on a.otype = k.otype;"
Due to those two issue it is not recommended to apply SAP HANA Revision 71 in productive systems. Nevertheless using it for testing and proof-of-concepts is not critical.
The fixes for the mentioned issues are planned to be released in CW 08/2014.
==---===---===
What does it mean?