From owner-freebsd-questions@FreeBSD.ORG Thu Oct 8 19:37:36 2009 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 E13491065679 for ; Thu, 8 Oct 2009 19:37:36 +0000 (UTC) (envelope-from pldrouin@pldrouin.net) Received: from smtp.cyberfingers.net (smtp.cyberfingers.net [198.177.254.227]) by mx1.freebsd.org (Postfix) with ESMTP id BE7388FC13 for ; Thu, 8 Oct 2009 19:37:35 +0000 (UTC) Received: from mdaemon.pldrouin.net (pldrouinlap2-pc.physics.carleton.ca [134.117.23.34]) by smtp.cyberfingers.net (Postfix) with ESMTP id 273E2AB6C0F for ; Thu, 8 Oct 2009 15:34:44 -0400 (EDT) Message-ID: <4ACE3F95.7030005@pldrouin.net> Date: Thu, 08 Oct 2009 15:37:57 -0400 From: Pierre-Luc Drouin User-Agent: Thunderbird 2.0.0.23 (X11/20090824) MIME-Version: 1.0 To: freebsd-questions@freebsd.org References: <4ACE293B.8050201@pldrouin.net> In-Reply-To: <4ACE293B.8050201@pldrouin.net> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Subject: STOP state in top + segmentation fault and core dump? 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: Thu, 08 Oct 2009 19:37:37 -0000 Hi, When I run many heavily multi-threaded (1000+ threads/process) and memory intensive (1800 MB+/process) processes on FreeBSD 8.0 amd64, some of them get in STOP state in top and then seg fault and core dump. Are stop states in top caused by seg faults or it is the OS that stops the processes because they exceed some limit? I do not see any message about any limit being exceeded. The machine has 12GB of memory so it should be able to sustain 4 processes using 1800MB each of memory I guess (I have 12 GB of swap space anyway but it stays unused). Everything runs fine if I run a single process and I have never encountered such problems when running these on Linux. Anyone has an idea? I am using the default limits right now: >limit cputime unlimited filesize unlimited datasize 33554432 kbytes stacksize 524288 kbytes coredumpsize unlimited memoryuse unlimited vmemoryuse unlimited descriptors 11095 memorylocked unlimited maxproc 5547 sbsize unlimited swapsize unlimited Thank you!