Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 14 Dec 2011 19:16:40 -0600
From:      CyberLeo Kitsana <cyberleo@cyberleo.net>
To:        lists@midsummerdream.org
Cc:        ryallsd@gmail.com, perryh@pluto.rain.com, freebsd-questions@freebsd.org
Subject:   Re: AHCI driver and static device names
Message-ID:  <4EE94A78.1030506@cyberleo.net>
In-Reply-To: <4EE91308.8050300@midsummerdream.org>
References:  <4ED98E9F.9010401@midsummerdream.org> <CAN3mi_2u%2BHwFf3m%2BxvsNncfNpj_rFp94xjAv%2Bf0eFT7c4a%2B8Tg@mail.gmail.com> <4EDA489B.9060503@midsummerdream.org> <4EDA56A3.6090108@cyberleo.net> <4edb4b48.LvOhZvvsP1inJeps%perryh@pluto.rain.com> <4EE91308.8050300@midsummerdream.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On 12/14/2011 03:20 PM, Rob wrote:
> Can glabels, gpt, and zfs all work together?  I have a system where I
> have disks with 4 gpt partitions. Partitions 2 and 3 are part of gmirror
> arrays, and partition 4 is part of a zfs pool.  glabel says it writes to
> the end of the partition, which I believe zfs also writes to doesn't it?

Yup. However, all nestable geoms protect their metadata (when it exists)
by providing a device that is smaller, so any nested consumer never even
sees the provider's metadata. The end of the glabel device to which zfs
writes its metadata in your implied example is actually several sectors
prior to the end of the device or partition to which glabel writes its
metadata.

Explicit glabels are not strictly necessary with the GPT partitioning
scheme, since the glabel module can peek into the GPT data structure,
extract label names from there, and automatically create appropriate
/dev/gpt/ entries for those labels.

-- 
Fuzzy love,
-CyberLeo
Technical Administrator
CyberLeo.Net Webhosting
http://www.CyberLeo.Net
<CyberLeo@CyberLeo.Net>

Furry Peace! - http://wwww.fur.com/peace/



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4EE94A78.1030506>