From owner-freebsd-geom@FreeBSD.ORG Mon Mar 29 14:37:00 2010 Return-Path: Delivered-To: freebsd-geom@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 085181065678 for ; Mon, 29 Mar 2010 14:37:00 +0000 (UTC) (envelope-from nonsolosoft@diff.org) Received: from smtpi4.ngi.it (smtpi4.ngi.it [88.149.128.34]) by mx1.freebsd.org (Postfix) with ESMTP id BFC9E8FC18 for ; Mon, 29 Mar 2010 14:36:59 +0000 (UTC) Received: from lap.diff.org (81-174-26-135.static.ngi.it [81.174.26.135]) by smtpi4.ngi.it (Postfix) with ESMTP id F3CBE421E1 for ; Mon, 29 Mar 2010 16:36:57 +0200 (CEST) Message-ID: <4BB0BB09.4000000@diff.org> Date: Mon, 29 Mar 2010 16:36:57 +0200 From: Ferruccio Zamuner Organization: NonSoLoSoft User-Agent: Thunderbird 2.0.0.23 (X11/20100121) MIME-Version: 1.0 To: freebsd-geom@freebsd.org Content-Type: multipart/mixed; boundary="------------040102050903050308080606" X-Content-Filtered-By: Mailman/MimeDel 2.1.5 Subject: gmirror and disk failure X-BeenThere: freebsd-geom@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: GEOM-specific discussions and implementations List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 29 Mar 2010 14:37:00 -0000 This is a multi-part message in MIME format. --------------040102050903050308080606 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit kernel: ad6: TIMEOUT - FLUSHCACHE48 retrying (1 retry left) kernel: ad6: TIMEOUT - WRITE_DMA retrying (1 retry left) 246989210 kernel: ad6: TIMEOUT - WRITE_DMA retrying (1 retry left) 247031443 kernel: ad6: TIMEOUT - READ_DMA retrying (1 retry left) 50334911 Why such errors make freebsd 8.0 to reboot? I suppose that gmirror has to mark ad6 broken and bring the mirror in degraded mode. Isn't it? After reboot: ad6: FAILURE - READ_DMA us=51 error=40 LBA=25293375 GEOM_MIRROR: request failed (error=5) . ad6s1[READ(offset=12950175744, lenght 131072)] GEOM_MIRROR: Synchronization request failed (error=5). mirror/gm0[READ(offset=12950175744, lenght 131072)] /sbin/gmirror status Name Status Components mirror/gm0 DEGRADED ad4s1 (26%) ad6s1 If the error is on ad6s1, why ad4s1 has to sycronize with the fault one? --------------040102050903050308080606--