From owner-freebsd-stable@FreeBSD.ORG Sat Jan 24 13:42:19 2004 Return-Path: Delivered-To: freebsd-stable@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 7013C16A4CE for ; Sat, 24 Jan 2004 13:42:19 -0800 (PST) Received: from ganymede.hub.org (u46n208.hfx.eastlink.ca [24.222.46.208]) by mx1.FreeBSD.org (Postfix) with ESMTP id 304F543D53 for ; Sat, 24 Jan 2004 13:42:18 -0800 (PST) (envelope-from scrappy@hub.org) Received: by ganymede.hub.org (Postfix, from userid 1000) id 1F5943789F; Sat, 24 Jan 2004 17:38:10 -0400 (AST) Received: from localhost (localhost [127.0.0.1]) by ganymede.hub.org (Postfix) with ESMTP id 1495C344A9; Sat, 24 Jan 2004 17:38:10 -0400 (AST) Date: Sat, 24 Jan 2004 17:38:10 -0400 (AST) From: "Marc G. Fournier" To: freebsd-stable@freebsd.org Message-ID: <20040124172849.L96629@ganymede.hub.org> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII cc: mab@red-bean.com Subject: MultiThreaded AOLServer under FreeBSD 4.x X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 24 Jan 2004 21:42:19 -0000 The OS is a bit old right now, from Sept 20th -STABLE sources, but ... I'm getting a weird circumstance with clients running AOLServer and OpenACS 3.2.5 ... after a few days of running, the process no longer responds to HTTP requests, and the logs fill up with a whack of 'out of file descriptor' messages ... limit on the machine shows I have "descriptors 11095" available ... and using lsof on one of the processes that as restarted on Thursday shows that ~424 files opened, out of which there are 388 KQUEUE: nsd8x 65198 nobody 46u KQUEUE 0x92599b00 count=0, state=0 nsd8x 65198 nobody 47u KQUEUE 0x915d4400 count=0, state=0 nsd8x 65198 nobody 48u KQUEUE 0x8fb8b100 count=0, state=0 Checking another one that has been running longer: mars# lsof -p 98527 | wc -l 1046 mars# lsof -p 98527 | grep KQUEUE | wc -l 1009 What are these KQUEUE files that are open? Is there a 'leak' maybe, where they aren't being closed properly? Checking another server, with -STABLE current to Jan 6th, there still shows >90% of files opened being these KQUEUEs: neptune# lsof -p 84915 | wc -l 456 neptune# lsof -p 84915 | grep KQUEUE | wc -l 420 Thoughts? ---- Marc G. Fournier Hub.Org Networking Services (http://www.hub.org) Email: scrappy@hub.org Yahoo!: yscrappy ICQ: 7615664