From owner-freebsd-hackers Tue Oct 23 7:30:10 2001 Delivered-To: freebsd-hackers@freebsd.org Received: from bogslab.ucdavis.edu (bogslab.ucdavis.edu [169.237.68.34]) by hub.freebsd.org (Postfix) with ESMTP id 1621F37B405 for ; Tue, 23 Oct 2001 07:30:04 -0700 (PDT) Received: from thistle.bogs.org (thistle.bogs.org [198.137.203.61]) by bogslab.ucdavis.edu (8.9.3/8.9.3) with ESMTP id HAA80430 for ; Tue, 23 Oct 2001 07:29:57 -0700 (PDT) (envelope-from greg@bogslab.ucdavis.edu) Received: from thistle.bogs.org (localhost [127.0.0.1]) by thistle.bogs.org (8.11.3/8.11.3) with ESMTP id f9NER0t46839 for ; Tue, 23 Oct 2001 07:27:01 -0700 (PDT) (envelope-from greg@thistle.bogs.org) Message-Id: <200110231427.f9NER0t46839@thistle.bogs.org> To: hackers@FreeBSD.ORG X-To: "PSI, Mike Smith" X-Sender: owner-freebsd-hackers@FreeBSD.ORG Subject: Re: Duping a hard disk In-reply-to: Your message of "Tue, 23 Oct 2001 08:35:05 EDT." <3BD563F9.299FE3C@mitre.org> Reply-To: gkshenaut@ucdavis.edu Date: Tue, 23 Oct 2001 07:26:59 -0700 From: Greg Shenaut Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG In message <3BD563F9.299FE3C@mitre.org>, "PSI, Mike Smith" cleopede: >I am running a lab with 43 FreeBDS machines and will be adding about 20 >more in the near future. ALL these machines are absolutely identical >except for IP address and machine name. To speed up the adding of new >machines, I envision making a duplication station, where I would add a >"new" disk as a slave and then dup the master disk to the slave disk. >Then I would only have to change IP and machine name. > >But alas, I cannot find any procedures for doing this. Does anyone know >how to duplicate a master disk to a "new" slave disk??? It would REALLY >make my life much easier. > >BTW, We are running 3.2. Yea, I know it's ancient but we have added >significant kernel hacks to support specialized ATN and X.25 protocols >and don't envision upgrading until we get our modifications completed. I have done this two ways. One way was to set up, as you say, a duplication station where I would connect the target disk and dd the source disk to it. The other way was to create a compressed tape image of the source disk and carry an external tapedrive around, along with a floppy with enough of a system on it to boot, read the tape, and write the disk. I have been thinking about this lately, because both of those methods are somewhat unwieldy. My current thinking is to configure all the disks to have a second, minimal root partition on them that doesn't depend upon the actual system in any way. The idea is to boot the systems to that partition and then copy in the new configuration across the LAN, leaving the copyin partition untouched. While the copyin system is running, you can mount & tweek the newly installed system. If you want to update the copyin partition, you can do it later once the main system is running. A second idea I've been pondering is to buy some of those kits that convert a drive into a removeable, and go back to the duplication station concept. For this to work well, you need some extra hard disks--you build the new system on a basketful of spares, and then walk around the lab swapping them in. Build the new system on the ones you just swapped out, and repeat until everyone is updated. This would imply that user data (if any) would be somewhere else, such as a second internal disk. Also, as someone else has suggested, use rdist for small updates. But I think that for a major upgrade it is worth setting everything up on a testbed system, and then doing a reinstall on all the other machines. BTW, if you are running 3.2, make sure you have updated or disabled telnetd to avoid the exploit. The system in our lab that was hacked a while back was running 3.2, so the exploit definitely works on it. Greg Shenaut To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message