On 08/28/2014 01:50 AM, gmb wrote:
> Can somebody please confirm whether aggregate functions such as GROUP BY
> should use indexes ?
Sometimes. In your case, the index has one more column than the GROUP
BY, which makes it less likely that Postgres will use it (since
depending on the cardinality of ddet_id, it might actually be slower to
use the index).
In addition, other folks on this thread have already pointed out the
memory settings issues to you.
--
Josh Berkus
PostgreSQL Experts Inc.
http://pgexperts.com