From owner-freebsd-geom@FreeBSD.ORG Mon Aug 28 16:22:51 2006 Return-Path: X-Original-To: freebsd-geom@freebsd.org Delivered-To: freebsd-geom@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 215DD16A4E1 for ; Mon, 28 Aug 2006 16:22:51 +0000 (UTC) (envelope-from arne_woerner@yahoo.com) Received: from web30314.mail.mud.yahoo.com (web30314.mail.mud.yahoo.com [209.191.69.76]) by mx1.FreeBSD.org (Postfix) with SMTP id 5ECC243D77 for ; Mon, 28 Aug 2006 16:22:50 +0000 (GMT) (envelope-from arne_woerner@yahoo.com) Received: (qmail 28912 invoked by uid 60001); 28 Aug 2006 16:22:49 -0000 DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=Message-ID:Received:Date:From:Subject:To:Cc:In-Reply-To:MIME-Version:Content-Type:Content-Transfer-Encoding; b=xSFT0YNrUOhE3njQKKxwwEKxcD0lQx0LNR+JgkQSlqehmvXzzaJR30oUingRe6GvOmlPnoh/hQRiIfC9jXRDIQ3iuwHTBTIpxAYUyzt05FLPRqXdMrO0yvJZPto697xBtXAPCPuSSUHDDk1DBfqD4byrjfhc81O670SBg/pK/pM= ; Message-ID: <20060828162249.28910.qmail@web30314.mail.mud.yahoo.com> Received: from [213.54.67.25] by web30314.mail.mud.yahoo.com via HTTP; Mon, 28 Aug 2006 09:22:49 PDT Date: Mon, 28 Aug 2006 09:22:49 -0700 (PDT) From: "R. B. Riddick" To: Pawel Jakub Dawidek , Alastair Watts In-Reply-To: <20060828153716.GB64750@garage.freebsd.pl> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Cc: freebsd-geom@freebsd.org Subject: Re: gvinum behaviour on 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, 28 Aug 2006 16:22:51 -0000 Hmm... I just did some experiments with gnop and my graid5 class: Somehow a gnop device with 100% failure probability is not removed from the graid5, although there were read accesses which failed with EIO... So I added some new lines, that retry up to 50 times (via my good old ENOMEM trick: that is issuing a ENOMEM instead of EIO or so...) and after 10 retries it removes the failing disk... Should we do that for the other geom classes, too? For maximum reliability... Or do I miss something? Or is my graid5 class somehow ill in this point? -Arne __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com