Now the execution time for my query is down to ~10 - 13 seconds, which is already a big step ahead. Thanks! Are there any other settings that might be necessary to tweak on windows in order to reduce execution time even a little bit more? One thing i don't understand very well though is that if I execute the query on table 1 with some conditions for the first time it is still slow, but when i execute it more often with changing the conditions it gets faster. Even when i query table 1 then query table 3 (with the same table definition) and then query table 1 again, the query on table 1 gets faster again. Christian Rengstl M.A. Klinik und Poliklinik für Innere Medizin II Kardiologie - Forschung Universitätsklinikum Regensburg B3 1.388 Franz-Josef-Strauss-Allee 11 93053 Regensburg Tel.: +49-941-944-7230 >>> On Tue, Oct 30, 2007 at 8:21 PM, in message <47278421.6010906@xxxxxxxxxxxx>, Guillaume Lelarge <guillaume@xxxxxxxxxxxx> wrote: > Christian Rengstl a écrit : >> My OS is Windows 2003 with 4GB Ram and Xeon Duo with 3.2 GHz; >> shared_buffers is set to 32MB (as I read it should be fairly low on >> Windows) and work_mem is set to 2500MB, but nevertheless the query takes >> about 38 seconds to finish. The table "table1" contains approx. 3 >> million tuples and table2 approx. 500.000 tuples. If anyone could give >> an advice on either how to optimize the settings in postgresql.conf or >> anything else to make this query run faster, I really would appreciate. >> > > 32MB for shared_buffers seems really low to me but 2500MB for work_mem > seems awfully high. The highest I've seen for work_mem was something > like 128MB. I think the first thing you have to do is to really lower > work_mem. Something like 64MB seems a better bet at first. > > Regards. > > > -- > Guillaume. > http://www.postgresqlfr.org > http://dalibo.com ---------------------------(end of broadcast)--------------------------- TIP 6: explain analyze is your friend
