From clamav-users-bounces@lists.clamav.net Wed Apr 11 18:26:23 2007 Return-Path: X-Original-To: list@tad.clamav.net Delivered-To: list@tad.clamav.net X-Virus-Scanned: Debian amavisd-new at tad.clamav.net Received: from tad.clamav.net ([127.0.0.1]) by localhost (tad.clamav.net [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 2FvKXE4EXFoU; Wed, 11 Apr 2007 18:26:23 +0200 (CEST) Received: from tad.clamav.net (localhost.localdomain [127.0.0.1]) by tad.clamav.net (Postfix) with ESMTP id 35A714140C9; Wed, 11 Apr 2007 18:26:13 +0200 (CEST) X-Original-To: clamav-users@tad.clamav.net Delivered-To: clamav-users@tad.clamav.net X-Virus-Scanned: Debian amavisd-new at tad.clamav.net Received: from tad.clamav.net ([127.0.0.1]) by localhost (tad.clamav.net [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id qdigC2Aqw3C4 for ; Wed, 11 Apr 2007 18:26:11 +0200 (CEST) Received: from mailgate.vbhcs.org (mailgate.vbhcs.org [65.83.58.34]) by tad.clamav.net (Postfix) with ESMTP id D6FEB4140A2 for ; Wed, 11 Apr 2007 18:26:10 +0200 (CEST) X-Mailer: QUALCOMM Windows Eudora Version 7.1.0.9 Date: Wed, 11 Apr 2007 11:26:09 -0500 To: ClamAV users ML From: Noel Jones In-Reply-To: <461D01DD.7040808@gmx.net> References: <461CBC94.3040607@gmx.net> <20070411142229.C18E679785A@mgate2.vbhcs.org> <461D01DD.7040808@gmx.net> Mime-Version: 1.0 Message-Id: <20070411162609.CF76E797838@mgate2.vbhcs.org> Subject: Re: [Clamav-users] error stops clamd X-BeenThere: clamav-users@lists.clamav.net X-Mailman-Version: 2.1.9 Precedence: list Reply-To: ClamAV users ML List-Id: ClamAV users ML List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: clamav-users-bounces@lists.clamav.net Errors-To: clamav-users-bounces@lists.clamav.net At 10:42 AM 4/11/2007, jacusy wrote: > > Yes, clamd and the whole clamav structure should be more resistant to > > failure. Your choices are to either work with it as is, or wait > > until clamav is more mature. > > >It should not be too hard to rewrite freshclam so it downloads the >update to a temporary file first, and then processes the database. I'm quite sure that freshclam does that already, but I don't think that any testing is done on the database other than verifying the digital signature is good. I don't know if it's practical to do more extensive testing on the downloaded database before using it, but that might help. It would also help if all of clamav was more robust when presented with unexpected database results. I'm not sure why there were so many failures last night, but I suspect it has something to do with failing or maybe taking too long while trying to convert the main.cvd file to the main.inc directory structure for incremental updates, but that's just a guess. Freshclam download failures are not terribly uncommon and don't usually cause clamd to croak, so this problem was more complicated than a simple freshclam failure. Hopefully the developers will be able to solve this. -- Noel Jones _______________________________________________ Help us build a comprehensive ClamAV guide: visit http://wiki.clamav.net http://lurker.clamav.net/list/clamav-users.html