Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 16 Jun 1999 10:58:20 +0930
From:      Greg Lehey <grog@lemis.com>
To:        Christopher Michaels <ChrisMic@clientlogic.com>
Cc:        FreeBSD Questions <questions@FreeBSD.org>
Subject:   Re: What would happen if a vinum drive failed?
Message-ID:  <19990616105820.D521@freebie.lemis.com>
In-Reply-To: <6C37EE640B78D2118D2F00A0C90FCB440110598F@site2s1>; from Christopher Michaels on Tue, Jun 15, 1999 at 04:46:35PM -0400
References:  <6C37EE640B78D2118D2F00A0C90FCB440110598F@site2s1>

next in thread | previous in thread | raw e-mail | index | archive | help
On Tuesday, 15 June 1999 at 16:46:35 -0400, Christopher Michaels wrote:
> Hello,
> I have a simple question... I have a vinum volume that consists of 3
> concatenated drives.  If say.. the 3rd drive were to fail, is the whole
> volume trashed? or just the information on the 3rd drive?

Is this your configuration?

drive 1 device /dev/da1h
drive 2 device /dev/da2h
drive 3 device /dev/da3h
volume foo
  plex org concat
    sd size 4g drive 1
    sd size 4g drive 2
    sd size 4g drive 3

And you're asking what happens if drive 3 (subdisk foo.p0.s2) dies?

In this case, your drive stays up, your plex (foo.p0) is degraded, and
your subdisk is obviously down.  If you're very lucky, you can access
some data on the drive, but effectively this is not a good way to do
things, since you can't control the layout of files in a ufs file
system.

If you want protection against drive failure, you have two choices:
RAID-1 or RAID-5.  Corresponding configurations would be:

(RAID-1)

drive 1 device /dev/da1h
drive 2 device /dev/da2h
drive 3 device /dev/da3h
volume foo
  plex org concat
    sd size 12g drive 1
  plex org concat
    sd size 12g drive 2

(RAID-5)
drive 1 device /dev/da1h
drive 2 device /dev/da2h
drive 3 device /dev/da3h
drive 4 device /dev/da3h
volume foo
  plex org raid5 512k
    sd size 4g drive 1
    sd size 4g drive 2
    sd size 4g drive 3
    sd size 4g drive 4

Each of these examples will give you a 12 GB volume, and the failure
of any one drive will not affect availability (though it will affect
performance).  The original version uses 12 GB of disk, the RAID-1
version uses 24 GB, and the RAID-5 version uses 16 GB.

Greg
--
See complete headers for address, home page and phone numbers
finger grog@lemis.com for PGP public key


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-questions" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19990616105820.D521>