I recently saw a factor 5 improvement on a grouped aggregate + join by moving from Windows to Linux.
On 2 Apr 2014 09:15, =?ISO-8859-1?Q?Hannes_M=FChleisen?= <Hannes.Muehleisen@cwi.nl> wrote:
On 04/02/2014 02:38 AM, Dino Veritas wrote:
> Any thoughts on how I can achieve better performance?
I would also suggest testing on Linux. We have seen large differences in
performance, in particular as the database size gets close to the amount
of RAM.