From owner-freebsd-stable@FreeBSD.ORG Sun Apr 6 09:01:22 2008 Return-Path: Delivered-To: freebsd-stable@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id A5CE21065670 for ; Sun, 6 Apr 2008 09:01:22 +0000 (UTC) (envelope-from guido@gvr.org) Received: from gvr.gvr.org (gvr-gw.gvr.org [82.95.154.195]) by mx1.freebsd.org (Postfix) with ESMTP id 5EBA48FC13 for ; Sun, 6 Apr 2008 09:01:22 +0000 (UTC) (envelope-from guido@gvr.org) Received: by gvr.gvr.org (Postfix, from userid 657) id 8D15342D952; Sun, 6 Apr 2008 11:01:21 +0200 (CEST) Date: Sun, 6 Apr 2008 11:01:21 +0200 From: Guido van Rooij To: Peter Jeremy Message-ID: <20080406090121.GA36387@gvr.gvr.org> References: <20080404174605.GA41929@gvr.gvr.org> <20080404221532.GT49813@server.vk2pj.dyndns.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20080404221532.GT49813@server.vk2pj.dyndns.org> Cc: freebsd-stable@freebsd.org Subject: Re: NTP on amd74 + powernow issues X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 06 Apr 2008 09:01:22 -0000 On Sat, Apr 05, 2008 at 09:15:32AM +1100, Peter Jeremy wrote: > On Fri, Apr 04, 2008 at 07:46:05PM +0200, Guido van Rooij wrote: > >I've got a amd64 system where I;m using powernow. I am using th i8254 > >timecounter. > > I'm using ACPI-fast on amd64 without problems. i8254 should not be > affected by PowerNow! but ACPI-fast should be listed as preferable. > Do you have a particular reason for using the i8254? > > >Especially the largers resets (+10 seconds) are annoying as e.g. dovecot > >bails out. > > > >Any clue on how this could be fixed? > > Any time resets are abnormal. There's nothing obviously wrong with > your NTP servers. I have had problems in the past with the ntpd PLL > saturating and demonstrating symptoms similar to what you are seeing. > > What does your ntp.drift contain? My guess is that it's either +500 > or -500. If this is true, I suggest you: gvr# cat /var/db/ntpd.drift -288.731 > - stop ntpd > - delete ntp.drift > - enable loopstats collection in your ntp.conf (optional) > - run "ntptime -f 0" > - run "/etc/rc.d/ntpdate start" > - run "/etc/rc.d/ntpd start" > > This will force ntpd to re-calibrate and, hopefully, it will stabilise, > though this can take a day or so. Didn't work :-( -Guido