From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Wed, 27 Jun 2001 03:52:26 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Wed, 27 Jun 2001 03:52:17 -0400 Received: from delrom.ro ([193.231.234.28]:47377 "HELO delrom.ro") by vger.kernel.org with SMTP id ; Wed, 27 Jun 2001 03:52:03 -0400 Date: Wed, 27 Jun 2001 10:52:56 +0300 From: Silviu Marin-Caea To: linux-kernel@vger.kernel.org Subject: Realtek 8139 driver or sucky hardware? Message-Id: <20010627105256.2e75fdca.silviu@delrom.ro> Organization: Delta Romania X-Mailer: Sylpheed version 0.4.99cvs3 (GTK+ 1.2.9; i586-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit X-AntiVirus: OK (checked by AntiVir Version 6.7.0.1) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org I have a server that had a Realtek 8139 card that worked nicely under normal circumstances. But I made a mistake in a crontab and I had 60 instances of a backup script starting one per minute, all of them wishing to create the same .tar.gz into a samba mounted share. This crazy situation had the server freeze solid, with only cold boot as remedy. No matter what stupid things I do on it, I shouldn't be able to take the kernel down, right? After I replaced the Realtek with a 3com, I could see all of the 60 instances fighting like worms in shit, but the server survived. Kernel 2.4.5 compiled with Red Hat gcc 2.96-81. With kgcc, it was acting the same way. Please CC me. Thank you. -- Systems and Network Administrator - Delta Romania Phone +4093-267961