| From: | Robert Haas <robertmhaas(at)gmail(dot)com> |
|---|---|
| To: | Claudio Freire <klaussfreire(at)gmail(dot)com> |
| Cc: | Kevin Grittner <Kevin(dot)Grittner(at)wicourts(dot)gov>, srobertjames <srobertjames(at)gmail(dot)com>, pgsql-performance <pgsql-performance(at)postgresql(dot)org> |
| Subject: | Re: Understanding Hash Join performance |
| Date: | 2011-06-03 03:55:35 |
| Message-ID: | BANLkTinN7Q_RXvs8qSuzdih0+9v7fe432g@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
On Thu, Jun 2, 2011 at 2:56 PM, Claudio Freire <klaussfreire(at)gmail(dot)com> wrote:
> On Thu, Jun 2, 2011 at 4:57 PM, Kevin Grittner
> <Kevin(dot)Grittner(at)wicourts(dot)gov> wrote:
>> And the
>> planner does take the size of work_mem and the expected data set
>> into consideration when estimating the cost of the hash join.
>
> And shouldn't it?
>
> In a gross mode, when hash joins go to disk, they perform very poorly.
> Maybe the planner should take that into account.
It does.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
| From | Date | Subject | |
|---|---|---|---|
| Next Message | pasman pasmański | 2011-06-05 15:25:39 | Re: Why we don't want hints Was: Slow count(*) again... |
| Previous Message | Kevin Grittner | 2011-06-02 19:57:03 | Re: Problem query |