Igor,
Thank you for the hint, I read about the planner, added "vacuum analyze " to
my procedures.
There is no join in my query but GROUP BY that is taking all the time and I
don't know how to tune it.
It gets executed by the procedure, the execution time requirement is < 4
sec,
but it takes 8-11 sec against 3 partitions , 9 mln rec each, it goes to 22
sec for 5 partitions.
I've been testing PostgreSQL performance for the last 2 months, comparing it
whith MySQL,
PostgreSQL performance with 5+ mln records on the table with 14 columns is
worse.
Is 14 columns is a big table for Postgres or 5mln rec is a big table?
The whole picture is that there are 2 databases : OLTP & "OLAP" that use to
be on different machines and on different databases.
The new project requires to put it on one database & machine.
I preferred Postgres ( poorly designed oltp would not suffer even more on
mysql) and now I'm trying to tune OLAP db.
thank you.
Helen
--
View this message in context:
http://postgresql.1045698.n5.nabble.com/How-to-see-memory-usage-using-explain-analyze-tp4694962p4709415.html
Sent from the PostgreSQL - performance mailing list archive at Nabble.com.