Received: from maia.hub.org (unknown [200.46.208.211])
	by mail.postgresql.org (Postfix) with ESMTP id B630863249B
	for <pgsql-admin-postgresql.org@mail.postgresql.org>;
	Mon,  8 Feb 2010 21:36:16 -0400 (AST)
Received: from mail.postgresql.org ([200.46.204.86])
	by maia.hub.org (mx1.hub.org [200.46.208.211]) (amavisd-maia,
	port 10024) with ESMTP id 20741-03 for
	<pgsql-admin-postgresql.org@mail.postgresql.org>;
	Tue,  9 Feb 2010 01:35:32 +0000 (UTC)
X-Greylist: from auto-whitelisted by SQLgrey-1.7.6
Received: from sss.pgh.pa.us (sss.pgh.pa.us [66.207.139.130])
	by mail.postgresql.org (Postfix) with ESMTP id D7DC763242D
	for <pgsql-admin@postgresql.org>; Mon,  8 Feb 2010 21:36:05 -0400 (AST)
Received: from sss2.sss.pgh.pa.us (tgl@localhost [127.0.0.1])
	by sss.pgh.pa.us (8.14.2/8.14.2) with ESMTP id o191a1YD004539;
	Mon, 8 Feb 2010 20:36:01 -0500 (EST)
To: Greg Stark <gsstark@mit.edu>
cc: Kevin Kempter <kevink@consistentstate.com>, pgsql-admin@postgresql.org
Subject: Re: Re: "effective_io_concurrency" cannot be changed error when
	starting 8.4.2 instance
In-reply-to: <407d949e1002081711h4b13e10cy4f571034813567c5@mail.gmail.com> 
References: <201002081800.26118.kevink@consistentstate.com>
	<407d949e1002081711h4b13e10cy4f571034813567c5@mail.gmail.com>
Comments: In-reply-to Greg Stark <gsstark@mit.edu>
	message dated "Tue, 09 Feb 2010 01:11:08 +0000"
Date: Mon, 08 Feb 2010 20:36:01 -0500
Message-ID: <4538.1265679361@sss.pgh.pa.us>
From: Tom Lane <tgl@sss.pgh.pa.us>
X-Virus-Scanned: Maia Mailguard 1.0.1
X-Spam-Status: No, hits=-2.465 tagged_above=-10 required=5 tests=AWL=0.134,
	BAYES_00=-2.599
X-Spam-Level: 
X-Archive-Number: 201002/64
X-Sequence-Number: 33667

Greg Stark <gsstark@mit.edu> writes:
> The error means that your system does not support the system interface
> we need to support prefetching. In that case effective_io_concurrency
> will be a read-only parameter set to 0 and any attempt to set it will
> result in an error.

I wonder whether it would be better to have it still be a read-write
parameter, but with min/max values equal to zero.  All the other cases
where we have read-only parameters, they're *always* read-only.  It
seems a bit odd that this one is platform-specific not only as to the
allowed range but whether it can be assigned at all.

			regards, tom lane