From owner-freebsd-questions@FreeBSD.ORG Fri Feb 9 07:41:34 2007 Return-Path: X-Original-To: freebsd-questions@freebsd.org Delivered-To: freebsd-questions@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id 8925916A401 for ; Fri, 9 Feb 2007 07:41:34 +0000 (UTC) (envelope-from dking@ketralnis.com) Received: from ketralnis.com (melchoir.ketralnis.com [68.183.67.83]) by mx1.freebsd.org (Postfix) with ESMTP id 6448013C46B for ; Fri, 9 Feb 2007 07:41:34 +0000 (UTC) (envelope-from dking@ketralnis.com) Received: from [10.0.1.239] ([10.0.1.239]) (authenticated bits=0) by ketralnis.com (8.13.8/8.13.6) with ESMTP id l1975FdI049309 (version=TLSv1/SSLv3 cipher=AES128-SHA bits=128 verify=NO) for ; Thu, 8 Feb 2007 23:05:15 -0800 (PST) (envelope-from dking@ketralnis.com) Mime-Version: 1.0 (Apple Message framework v752.3) Content-Transfer-Encoding: 7bit Message-Id: <9B5571FE-BB31-49CE-A5C8-5C27BC6CB5DB@ketralnis.com> Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed To: freebsd-questions@freebsd.org From: David King Date: Thu, 8 Feb 2007 23:05:04 -0800 X-Mailer: Apple Mail (2.752.3) Subject: restore: no memory to extend symbol table 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: Fri, 09 Feb 2007 07:41:34 -0000 When moving from one disk to another, I did a dump of a large filesystem, like this: mkdir ~/tmo && cd ~/tmp && dump -0f - /dev/devicenode | split So now I have 99G in 100 files named xaa through xdv However, when I try to restore it: mount /mountpoint && cat ~/tmp/* | (cd /mountpoint && restore -rf -) after cat spits out 2.18 GB (in the same spot every time), after restore creates a number of directories (but no files), I get the error: no memory to extend symbol table abort? [yn] I've seen two other posts about similar errors: I'm sure that I have enough space on /tmp (since I symlinked it to a directory on a large disk), which is one proposed fix. Another proposed fix is to allow restore to get more than 512MB of RAM, but I'm not sure where to set that (since I'm not familiar with vm and I have no such limit in /etc/login.conf, and can find nothing in sysctl). It's not a RAM limit, because I have 1GB of RAM and 2.5 GB of swap. The other proposed fix is to use interactive mode and restore less at a time, but if I do that hardlinks between directories may be broken and I need to retain those (unless they wouldn't be broken between successive restores, in which case I can do that). Any ideas? I'm running 6.2-RELEASE -- David