From owner-freebsd-bugs@FreeBSD.ORG Sat Aug 19 20:15:21 2006 Return-Path: X-Original-To: freebsd-bugs@FreeBSD.org Delivered-To: freebsd-bugs@FreeBSD.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id C99B916A4E0 for ; Sat, 19 Aug 2006 20:15:21 +0000 (UTC) (envelope-from kris@obsecurity.org) Received: from elvis.mu.org (elvis.mu.org [192.203.228.196]) by mx1.FreeBSD.org (Postfix) with ESMTP id 0184243D55 for ; Sat, 19 Aug 2006 20:15:21 +0000 (GMT) (envelope-from kris@obsecurity.org) Received: from obsecurity.dyndns.org (elvis.mu.org [192.203.228.196]) by elvis.mu.org (Postfix) with ESMTP id 9BC561A4DD8; Sat, 19 Aug 2006 13:15:20 -0700 (PDT) Received: by obsecurity.dyndns.org (Postfix, from userid 1000) id 5A14D540E1; Sat, 19 Aug 2006 16:15:19 -0400 (EDT) Date: Sat, 19 Aug 2006 16:15:19 -0400 From: Kris Kennaway To: Geoffroy Desvernay Message-ID: <20060819201519.GD36605@xor.obsecurity.org> References: <200607311700.k6VH0fqV087593@freefall.freebsd.org> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="5QAgd0e35j3NYeGe" Content-Disposition: inline In-Reply-To: <200607311700.k6VH0fqV087593@freefall.freebsd.org> User-Agent: Mutt/1.4.2.2i Cc: freebsd-bugs@FreeBSD.org Subject: Re: kern/97686: [nfs] massive NFS server performance break under "options QUOTA" X-BeenThere: freebsd-bugs@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Bug reports List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sat, 19 Aug 2006 20:15:21 -0000 --5QAgd0e35j3NYeGe Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Mon, Jul 31, 2006 at 05:00:41PM +0000, Geoffroy Desvernay wrote: > We now have a stable machine without QUOTA *and* INET6. > (only without QUOTA the machine died 24hours later). We left rpc.lockd > running. > =20 > Here is a simple zsh script to make it crash when with QUOTA-enabled > kernel (use $(seq 1 50) or ? with another shell). launch it on a client > and see 'top' on the server... I'm not kernel (nor nfs) hacker, but I > may try some patches if it can help. > =20 > mount options: > /dev/da0s1g on /u (ufs, NFS exported, local, with quotas, soft-updates, > acls, fsid e0bf4c447f5cf8a5) > =20 > ---------------- crashlocks.zsh > #!/usr/local/bin/zsh -x > for i in {1..50} ; do > lockfile truc$i & > done > ---------------- > =20 > Hope this can help... Please provide the details of the panic, per the chapter on kernel debugging in the developers handbook. If it's not panicking but deadlocking, update to 6.1-STABLE where this bug is believed to be fixed. Thanks, Kris --5QAgd0e35j3NYeGe Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (FreeBSD) iD8DBQFE53FWWry0BWjoQKURArRSAJ9i1Cez068i5W/cAygPZ04rH+bLPACg0KTm AM8CS5a9PBnAHsuZMZ55r3M= =rF8Z -----END PGP SIGNATURE----- --5QAgd0e35j3NYeGe--