From owner-freebsd-bugs@FreeBSD.ORG Fri May 8 01:30:06 2009 Return-Path: Delivered-To: freebsd-bugs@hub.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 615B6106566C for ; Fri, 8 May 2009 01:30:06 +0000 (UTC) (envelope-from gnats@FreeBSD.org) Received: from freefall.freebsd.org (freefall.freebsd.org [IPv6:2001:4f8:fff6::28]) by mx1.freebsd.org (Postfix) with ESMTP id 4F33E8FC12 for ; Fri, 8 May 2009 01:30:06 +0000 (UTC) (envelope-from gnats@FreeBSD.org) Received: from freefall.freebsd.org (gnats@localhost [127.0.0.1]) by freefall.freebsd.org (8.14.3/8.14.3) with ESMTP id n481U6Iq000138 for ; Fri, 8 May 2009 01:30:06 GMT (envelope-from gnats@freefall.freebsd.org) Received: (from gnats@localhost) by freefall.freebsd.org (8.14.3/8.14.3/Submit) id n481U6pm000135; Fri, 8 May 2009 01:30:06 GMT (envelope-from gnats) Date: Fri, 8 May 2009 01:30:06 GMT Message-Id: <200905080130.n481U6pm000135@freefall.freebsd.org> To: freebsd-bugs@FreeBSD.org From: John Kilburg Cc: Subject: Re: kern/133980: panic: ffs_valloc: dup alloc X-BeenThere: freebsd-bugs@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: John Kilburg List-Id: Bug reports List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 08 May 2009 01:30:06 -0000 The following reply was made to PR kern/133980; it has been noted by GNATS. From: John Kilburg To: bug-followup@freebsd.org Cc: Subject: Re: kern/133980: panic: ffs_valloc: dup alloc Date: Thu, 07 May 2009 18:28:38 -0700 I am guessing now that this is a problem that I've seen mentioned in passing as a possibility in a few postings about the maximum number of inodes for UFS2. Using the default settings the filesystem ended up with more than 2^31 inodes. I decreased the inode density to about half the default (~1.4M inodes) and now things seem to be working perfectly. I rsync'd with 3 sessions, did a make buildworld and otherwise loaded the thing with no problems at all. It would be nice if newfs (or something) warned people about the effective 2^31 inode limit for ufs2 until the problem is spotted. I guess the real fix is to switch to a filesystem that doesn't have static inode allocation. :) -john