From owner-freebsd-current Sat Oct 12 21:13: 0 2002 Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 6DB3237B401 for ; Sat, 12 Oct 2002 21:12:59 -0700 (PDT) Received: from mail.chesapeake.net (chesapeake.net [205.130.220.14]) by mx1.FreeBSD.org (Postfix) with ESMTP id 9D67743E8A for ; Sat, 12 Oct 2002 21:12:57 -0700 (PDT) (envelope-from jroberson@chesapeake.net) Received: from localhost (jroberson@localhost) by mail.chesapeake.net (8.11.6/8.11.6) with ESMTP id g9D4Chu36573; Sun, 13 Oct 2002 00:12:43 -0400 (EDT) (envelope-from jroberson@chesapeake.net) Date: Sun, 13 Oct 2002 00:12:43 -0400 (EDT) From: Jeff Roberson To: Kris Kennaway Cc: current@FreeBSD.ORG Subject: Re: current unable to exec processes under load In-Reply-To: <20021012213213.GC93425@xor.obsecurity.org> Message-ID: <20021013001013.N30714-100000@mail.chesapeake.net> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On Sat, 12 Oct 2002, Kris Kennaway wrote: > I'm having a strange problem with my -current box: I am running the > bfbtester port on system utilities, which basically forks 100 copies > of the binary at once to test different options. The problem is that > sometimes the -current system becomes unable to fork *any* new > processes, and even suspending or trying to kill the bfbtester process > does not work. The rest of the system performance is fine (I can use > existing processes with no noticeable CPU latency): so it does not > appear to be a resource starvation condition. > > rot13# ls > [hangs] > load: 0.00 cmd: csh 22888 [vmmapw] 0.00u 0.00s 0% 1160k > > This process now cannot be interrupted by signals. > > => /usr/sbin/sendmail > * Single argument testing > ^Z > Suspended > [hang] > load: 0.00 cmd: tcsh 8403 [inode] 0.01u 0.00s 0% 1076k > > Any ideas? Can you back out my scheduler changes just to be sure? I can not foresee any way that they could cause this, but I'd like to be certain. Blocking on inode usually indicates a vfs deadlock. Can you break into ddb and type 'show lockedvnods' and paste me that output? Thanks! Jeff To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message