I am working with a ~100 GB datawarehouse. Our largest segment is about 5GB (excluding indexes) and we are having problems running out of temporary tablespace with our batch queries and transformations. Our temporary tablespace is sized at 10GB, and our queries often run very close using all this.

Does anyone have any guidelines for Temporary tablespace sizings? Does 10GB seem to little?

Regards and thanks for any help,
Brendan.