| From: | Greg Smith <greg(at)2ndquadrant(dot)com> |
|---|---|
| To: | Benjamin Krajmalnik <kraj(at)servoyant(dot)com> |
| Cc: | pgsql-performance(at)postgresql(dot)org |
| Subject: | Re: Configuration for a new server. |
| Date: | 2011-02-02 18:22:55 |
| Message-ID: | 4D49A0FF.4010104@2ndquadrant.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-performance |
Benjamin Krajmalnik wrote:
> So, if I understand correctly, I should strive for a relative increase in buffers_clean to buffers_backend
>
Right. Buffers written by a backend are the least efficient way to get
data out of the buffer cache, because the client running into that is
stuck waiting for a write call before it can use the resuling free
block. You want to avoid those in favor of checkpoint and
background-writer cleaner writes instead.
--
Greg Smith 2ndQuadrant US greg(at)2ndQuadrant(dot)com Baltimore, MD
PostgreSQL Training, Services, and 24x7 Support www.2ndQuadrant.us
"PostgreSQL 9.0 High Performance": http://www.2ndQuadrant.com/books
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Kenneth Marshall | 2011-02-02 18:32:17 | Re: Server Configuration |
| Previous Message | Robert Haas | 2011-02-02 18:20:59 | Re: [HACKERS] Slow count(*) again... |