From owner-freebsd-questions@FreeBSD.ORG Fri Dec 19 20:57:23 2008 Return-Path: Delivered-To: freebsd-questions@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 06AC21065675 for ; Fri, 19 Dec 2008 20:57:23 +0000 (UTC) (envelope-from fsb@thefsb.org) Received: from smtp154.iad.emailsrvr.com (smtp154.iad.emailsrvr.com [207.97.245.154]) by mx1.freebsd.org (Postfix) with ESMTP id D1F518FC0C for ; Fri, 19 Dec 2008 20:57:22 +0000 (UTC) (envelope-from fsb@thefsb.org) Received: from relay5.relay.iad.emailsrvr.com (localhost [127.0.0.1]) by relay5.relay.iad.emailsrvr.com (SMTP Server) with ESMTP id 2465B731654; Fri, 19 Dec 2008 15:57:22 -0500 (EST) Received: by relay5.relay.iad.emailsrvr.com (Authenticated sender: fsb-AT-thefsb.org) with ESMTPSA id E308D72EB57; Fri, 19 Dec 2008 15:57:20 -0500 (EST) User-Agent: Microsoft-Entourage/12.10.0.080409 Date: Fri, 19 Dec 2008 15:57:20 -0500 From: Tom Worster To: Wojciech Puchar Message-ID: Thread-Topic: Clearing SMART errors I don't care about? Thread-Index: AcliHGIG8i8iH/ERIkC9CI8t82VNqw== In-Reply-To: <20081219205105.F2813@wojtek.tensor.gdynia.pl> Mime-version: 1.0 Content-type: text/plain; charset="ISO-8859-1" Content-transfer-encoding: quoted-printable Cc: FreeBSD Questions ML Subject: Re: Clearing SMART errors I don't care about? 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: Fri, 19 Dec 2008 20:57:23 -0000 On 12/19/08 2:51 PM, "Wojciech Puchar" wrote: > which means that in 2 cases of 3 you ARE warned. yes. but what do i do with a smart warning? the google paper indicates that even they haven't figured it out yet, although they express some hope. "Despite those strong correlations, we =DEnd that failure prediction models based on SMART parameters alone are likely to be severely limited in their prediction accuracy, given that a large fraction of our failed drives have shown no SMART error signals whatsoever. This result suggests that SMART models are more useful in predicting trends for large aggregate populations than for individual components." in managing my servers, it's the failure of individual components that count. considering also 1) false positives and 2) replacement comes with a finite chance of causing service outage, it's not clear to me that replacing a dis= k on the basis of a smart warning is prudent. i'm not proposing any firm conclusions should be drawn. i started this by saying "i'm not convinced that smart monitoring is of much value..."=20