Re: LDC - Load Distributed Checkpoints with PG8.3b2 on Solaris

From: Greg Smith <gsmith(at)gregsmith(dot)com>
To: "Jignesh K(dot) Shah" <J(dot)K(dot)Shah(at)Sun(dot)COM>
Cc: pgsql-hackers(at)postgresql(dot)org
Subject: Re: LDC - Load Distributed Checkpoints with PG8.3b2 on Solaris
Date: 2007-11-13 19:19:25
Message-ID: Pine.GSO.4.64.0711131414230.15021@westnet.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Tue, 13 Nov 2007, Jignesh K. Shah wrote:

> I have tried increasing the bg_lru_multiplier from 2 to 5 from default but I
> dont see any more writes by bgwriter happening than my previous test which
> used the default.

The multiplier only impacts writes being done by the LRU eviction
mechanism; it doesn't change anything related to the checkpoint writes.
In fact, since those LRU writes are going on at the same time as the
checkpoint ones, increasing the multiplier too much can make the
checkpoint I/O spike worse. It's unlikely that higher values will
decrease the spike.

--
* Greg Smith gsmith(at)gregsmith(dot)com http://www.gregsmith.com Baltimore, MD

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Greg Smith 2007-11-13 19:36:14 Re: How to keep a table in memory?
Previous Message Josh Berkus 2007-11-13 19:13:31 Re: How to keep a table in memory?