From owner-freebsd-questions@FreeBSD.ORG Wed Nov 22 16:53:04 2006 Return-Path: X-Original-To: freebsd-questions@freebsd.org Delivered-To: freebsd-questions@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id 037C316A417 for ; Wed, 22 Nov 2006 16:53:04 +0000 (UTC) (envelope-from kris@obsecurity.org) Received: from elvis.mu.org (elvis.mu.org [192.203.228.196]) by mx1.FreeBSD.org (Postfix) with ESMTP id B36B143D7B for ; Wed, 22 Nov 2006 16:52:23 +0000 (GMT) (envelope-from kris@obsecurity.org) Received: from obsecurity.dyndns.org (elvis.mu.org [192.203.228.196]) by elvis.mu.org (Postfix) with ESMTP id 26C131A4D89; Wed, 22 Nov 2006 08:52:53 -0800 (PST) Received: by obsecurity.dyndns.org (Postfix, from userid 1000) id 9C95E515B5; Wed, 22 Nov 2006 11:52:38 -0500 (EST) Date: Wed, 22 Nov 2006 11:52:38 -0500 From: Kris Kennaway To: Dieter Message-ID: <20061122165238.GA37819@xor.obsecurity.org> References: <20061120111952.4213dacb.wmoran@collaborativefusion.com> <200611220712.HAA18428@sopwith.solgatos.com> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="jI8keyz6grp/JLjh" Content-Disposition: inline In-Reply-To: <200611220712.HAA18428@sopwith.solgatos.com> User-Agent: Mutt/1.4.2.2i Cc: freebsd-questions@freebsd.org Subject: Re: processes not getting fair share of available disk I/O (was: Re: TCP parameters and interpreting tcpdump output ) X-BeenThere: freebsd-questions@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: User questions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 22 Nov 2006 16:53:04 -0000 --jI8keyz6grp/JLjh Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Tue, Nov 21, 2006 at 11:12:38PM +0000, Dieter wrote: > > I'm surprised that you're seeing that much of a "hang". Even if the di= sks > > are busy, the system should slow down all disk processes equally, so no > > one process "blocks", but they're all a little slower. >=20 > I collected a bit of data: >=20 > While copying a large file from disk1 to disk2, >=20 > time ls on a small directory on disk3 (not cached in memory) >=20 > real 0m0.032s > user 0m0.000s > sys 0m0.003s >=20 > time ls on a small directory on disk2 >=20 > real 4m51.911s > user 0m0.000s > sys 0m0.002s >=20 > I expect access to a busy disk to take longer, but 5 minutes is > a bit much. And that's the root directory of the filesystem, > it didn't have to follow a long chain of directories to get there. >=20 > Sometimes I see long delays when accessing disk3, but it is > behaving at the moment. ls still has to acquire a number of locks in order to be sure that the contents of the directory aren't changing. If there are lots of other processes all competing for these locks, it will be slow. It looks like that's the case on your system, although details of your workload have been trimmed from your email. This should be better in a future version of FreeBSD (when it's safe to use shared locks instead of exlusive for vfs lookups) but today there may not be much you can do apart from lowering the load. Kris --jI8keyz6grp/JLjh Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (FreeBSD) iD8DBQFFZIBWWry0BWjoQKURArN6AKD+u0/+FDVSWrPLw5bXk0JAJZ8ecgCfXeMi bPcH8jWbjO9eHlZgATLwY40= =txuM -----END PGP SIGNATURE----- --jI8keyz6grp/JLjh--