From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Mon, 27 Jan 2003 12:08:42 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Mon, 27 Jan 2003 12:08:42 -0500 Received: from mx2.elte.hu ([157.181.151.9]:42394 "HELO mx2.elte.hu") by vger.kernel.org with SMTP id ; Mon, 27 Jan 2003 12:08:41 -0500 To: linux-kernel@vger.kernel.org Subject: /proc problem From: =?iso-8859-2?q?Burj=E1n_G=E1bor?= Date: Mon, 27 Jan 2003 18:16:42 +0100 Message-ID: MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii User-Agent: Emacs Gnus Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi, I've a mysterious problem with Linux kernels on a production system. The box is a web frontend machine and the problem is that after a couple of days of running one of httpd process "sticks" in. Commands w, ps and their friends usually hang after this case. I cannot stat the files under /proc//, so I think this should be a kernel-related problem. Only the reboot helps. I tried out a lot of kernel versions: 2.4.10, 2.4.19, 2.4.19rmap14b (the last one with the new procfs code), but these didn't help. Any idea? Buga