Re: Query Analyzing
От | Manfred Koizar |
---|---|
Тема | Re: Query Analyzing |
Дата | |
Msg-id | rrcajug3jo80tohqbm1g0f8fh6e1qcasqf@4ax.com обсуждение исходный текст |
Ответ на | Re: Query Analyzing ("Booth, Robert" <Robert_Booth@intuit.com>) |
Список | pgsql-general |
On Mon, 8 Jul 2002 11:16:34 -0700, "Booth, Robert" <Robert_Booth@intuit.com> wrote: >[about performance problems with a query containing two inner and > eight outer joins] Rob, sorry for being quiet for so long, the printout of your mail got lost in a pile of paper. It wouldn`t have helped much, however, if I responded earlier. I could not find a hot spot in your query, you are losing time constantly, not much with each outer join, but it sums up. Rewording your SQL won't help, AFAICS. I thought of converting the outer joins to inner joins (you would have to provide default entries in your lookup tables), writing the joins in this form: FROM a, b, c ... WHERE a.b_id=b.id AND a.c_id=c.id ... and letting the optimizer do its work, but I fear that you would again end up with lots of hash joins. Maybe caching the lookup tables in the frontend might help? Servus Manfred
В списке pgsql-general по дате отправления: