Date: Sun, 8 Jul 2007 02:23:12 +0200 From: Bernd Walter <ticso@cicely12.cicely.de> To: Volker <volker@vwsoft.com> Cc: current@freebsd.org Subject: Re: moving a zfs pool Message-ID: <20070708002312.GC82285@cicely12.cicely.de> In-Reply-To: <469014E9.5050906@vwsoft.com> References: <469014E9.5050906@vwsoft.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, Jul 08, 2007 at 12:34:17AM +0200, Volker wrote: > Hi, > > this is more likely currently a theoretical question: How does one > move a zfs pool from one machine to another? on old machine: zpool export foo attach disks to new machine on new machine: zpool import foo You can also do zpool import without parameter to list pools available to import. You need to "zpool import -f foo" on the new machine to force the import in case you hadn't exported it on the old machine. > I've done that in the past a lot with ufs filesystems by piping a dump > through an ssh session and restoring on another machine just to move > data to another system. This should also work on a per filesystem > basis when using zfs but isn't there a way to move the whole pool to > another machine (w/o physically moving the hard disk)? > > This would be the case when migrating a machine or migrating hard > disks. The zfs export feature doesn't look promising for this or do I > misread some bits? Please don't tell me, 42 is the answer... ;) You misread :) > I'm missing something I would call 'zpool dump' and 'zpool restore'. There is none :( But you can zfs send/receive at filesystem/volume level. It seems that you can trust the zfs list order on snapshots when copying multiple/all snapshots. Unfortunately things become more complex if you want to save clone data - especially to keep common stored blocks - as well. -- B.Walter http://www.bwct.de http://www.fizon.de bernd@bwct.de info@bwct.de support@fizon.de
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20070708002312.GC82285>