| From: | Robert Haas <robertmhaas(at)gmail(dot)com> |
|---|---|
| To: | Claudio Freire <klaussfreire(at)gmail(dot)com> |
| Cc: | Mark <Marek(dot)Balgar(at)seznam(dot)cz>, pgsql-performance(at)postgresql(dot)org |
| Subject: | Re: Query improvement |
| Date: | 2011-05-14 02:28:51 |
| Message-ID: | BANLkTimV2Y8uT6SjsvZhR1kZ5Pf8+HPtYg@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
On Mon, May 2, 2011 at 3:58 AM, Claudio Freire <klaussfreire(at)gmail(dot)com> wrote:
> Hash joins are very inefficient if they require big temporary files.
Hmm, that's not been my experience. What have you seen?
I've seen a 64-batch hash join beat out a
nested-loop-with-inner-indexscan, which I never woulda believed,
but...
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Stefan Keller | 2011-05-14 10:10:32 | KVP table vs. hstore - hstore performance (Was: Postgres NoSQL emulation) |
| Previous Message | Robert Haas | 2011-05-14 02:20:17 | Re: reducing random_page_cost from 4 to 2 to force index scan |