| From: | Hannu Krosing <hannu(at)skype(dot)net> |
|---|---|
| To: | Alvaro Herrera <alvherre(at)commandprompt(dot)com> |
| Cc: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, "Jim C(dot) Nasby" <jnasby(at)pervasive(dot)com>, Heikki Linnakangas <hlinnaka(at)iki(dot)fi>, pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: Dead Space Map |
| Date: | 2006-02-28 16:12:30 |
| Message-ID: | 1141143150.3864.3.camel@localhost.localdomain |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Ühel kenal päeval, T, 2006-02-28 kell 10:16, kirjutas Alvaro Herrera:
> Hannu Krosing wrote:
>
> > In the "big table with hotspots" case, I doubt that the win from
> > btbulkdelete will outweight having to scan 100000 index pages to get to
> > the 30 or 50 that can be bulkdeleted.
>
> Remember that WAL traffic is fsync'ed, so it can be _much_ slower than
> anything else.
But it should need to be fsynced only once, at commit time, so if you
can clean up more than one page, you end up with many wal records, but
just one fsync.
-------------
Hannu
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Jim C. Nasby | 2006-02-28 16:14:01 | Re: Dead Space Map |
| Previous Message | Hannu Krosing | 2006-02-28 16:10:05 | Re: Dead Space Map |