Date: Wed, 11 May 2016 15:27:12 -0700 From: Mel Pilgrim <list_freebsd@bluerosetech.com> To: Graham Allan <allan@physics.umn.edu>, freebsd-questions@freebsd.org Subject: Re: FreeBSD-10.3 root ZFS redux Message-ID: <3b96335d-d1cd-4a88-bbe2-457c4379d9df@bluerosetech.com> In-Reply-To: <573377D4.7080909@physics.umn.edu> References: <0409d2208ccce8e2f9e0ce6526bf6756.squirrel@webmail.harte-lyne.ca> <5845c9cfc3bed4171df9b4b386573782.squirrel@webmail.harte-lyne.ca> <573377D4.7080909@physics.umn.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
On 5/11/2016 11:20 AM, Graham Allan wrote: > On 5/11/2016 1:13 PM, James B. Byrne wrote: >> >> I managed to get this reversed in my head as I composed. The question >> is what happens when a nominally 3TB HDD is put in that has fewer >> accessible sectors than a similar 3TB drive already present? > > My understanding is that it would just be rejected (I see some older > solaris examples mention a "device too small" error). > > Best practice is probably to create a gpt partition of some standard > size on the drives before adding that to the pool, to give a little > adjustment room, though I confess I have made all my pools to date using > the raw drives (other than a geom label); and simply try to replace > drives with the same model. You need to do this anyway for root on ZFS without other storage in the system because of the freebsd-boot partition.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3b96335d-d1cd-4a88-bbe2-457c4379d9df>