From owner-freebsd-questions@FreeBSD.ORG Mon Oct 30 19:00:44 2006 Return-Path: X-Original-To: freebsd-questions@freebsd.org Delivered-To: freebsd-questions@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id C615A16A4A0 for ; Mon, 30 Oct 2006 19:00:44 +0000 (UTC) (envelope-from th.lacoste@wanadoo.fr) Received: from ultra1.univ-paris12.fr (ultra1.univ-paris12.fr [193.51.100.100]) by mx1.FreeBSD.org (Postfix) with ESMTP id 106ED43DF1 for ; Mon, 30 Oct 2006 18:59:44 +0000 (GMT) (envelope-from th.lacoste@wanadoo.fr) Received: from st-simon.miage.univ-paris12.fr (localhost [127.0.0.1]) by ultra1.univ-paris12.fr (8.13.8/8.13.8) with ESMTP id k9UIxfTD008934; Mon, 30 Oct 2006 19:59:42 +0100 (MET) Received: from sirius.miage.univ-paris12.fr (sirius.miage.univ-paris12.fr [194.214.13.28]) by st-simon.miage.univ-paris12.fr (Postfix) with ESMTP id 5CD5F47281A; Mon, 30 Oct 2006 19:59:41 +0100 (CET) From: Thierry Lacoste To: Chuck Swiger Date: Mon, 30 Oct 2006 19:59:37 +0100 User-Agent: KMail/1.9.4 References: <200610291354.58097.th.lacoste@wanadoo.fr> <4544BEB6.80606@mac.com> In-Reply-To: <4544BEB6.80606@mac.com> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200610301959.38092.th.lacoste@wanadoo.fr> Cc: freebsd-questions@freebsd.org Subject: Re: clock running too fast 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: Mon, 30 Oct 2006 19:00:44 -0000 Thank you. I tried TSC, ACPI-fast and i8254 but I still have the same problem. Best regards, Thierry. On Sunday 29 October 2006 15:46, Chuck Swiger wrote: > Thierry Lacoste wrote: > > On one of my servers running 6.1-RELEASE-p10 I cannot keep the clok > > synchronized using ntpd. AFAICS this is certainly because the clock > > is running way too fast (about one second per minute). > > > > After I run ntpdate then ntpd the clock is drifting and /var/db/ntp.drift > > contains 0.00. > > > > Is there a way to slow down the system clock (something like tickadj > > under some linux distributions) ? > > Take a look at "sysctl kern.timecounter", and choose another clock from the > list of choices (by setting kern.timecounter.hardware to something else in > the list of choices). > > If you are using TSC now, especially on a dual-CPU system, try using > ACPI-safe or i8254 instead. If you are using the ACPI timecounter, try > looking for a BIOS update for your hardware; perhaps that might fix the > bogus clock.