| From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
|---|---|
| To: | Robert Haas <robertmhaas(at)gmail(dot)com> |
| Cc: | Kevin Grittner <Kevin(dot)Grittner(at)wicourts(dot)gov>, Mladen Gogala <mladen(dot)gogala(at)vmsinfo(dot)com>, "pgsql-performance(at)postgresql(dot)org" <pgsql-performance(at)postgresql(dot)org> |
| Subject: | Re: anti-join chosen even when slower than old plan |
| Date: | 2010-11-11 19:17:00 |
| Message-ID: | 19475.1289503020@sss.pgh.pa.us |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
Robert Haas <robertmhaas(at)gmail(dot)com> writes:
> On Thu, Nov 11, 2010 at 1:23 PM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
>> I do think that something based around a settable-per-table caching
>> percentage might be a reasonable way to proceed. But the devil is in
>> the details, and we don't have those yet.
> I think one of the larger devils in the details is deciding how to
> estimate the assumed caching percentage when the user hasn't specified
> one.
I was imagining something very similar to the handling of seq_page_cost,
ie, there's a GUC controlling the system-wide default and then you can
override that per-table. But the real question is whether per-table
is a useful granularity to control it at. See my later message.
regards, tom lane
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Kevin Grittner | 2010-11-11 19:22:59 | Re: anti-join chosen even when slower than old plan |
| Previous Message | Kevin Grittner | 2010-11-11 19:15:55 | Re: anti-join chosen even when slower than old plan |