All of lore.kernel.org
 help / color / mirror / Atom feed
From: Gavin McCullagh <gmccullagh@gmail.com>
To: Rich <rich@pcfusion.co.uk>
Cc: linux-raid@vger.kernel.org
Subject: Re: LINEAR RAID, little help
Date: Sat, 7 Apr 2007 21:34:35 +0100	[thread overview]
Message-ID: <20070407203435.GA32499@gmail.com> (raw)
In-Reply-To: <4617E5E2.2090609@pcfusion.co.uk>

Hi,

On Sat, 07 Apr 2007, Rich wrote:

> Er, I went with Linear as reading around people seemed to recommend this 
> for odd sized drives (my old drives are 80's, 120 and 320's) also a read 
> somewhere that data on the other drives is more recoverable that most of 
> the other RAID's.

You just want to make a large filesystem out of odd sized disks.  I guess
that's fair enough.  The only reservation I'd have is that with N disks
your likelihood of failure is multiplied by N as any disk failure takes the
array down.  Personally, I'd be more inclined to try and put a separate
filesystem on each disk and use symlinks to pull them together into one
tree.

But now I know why linear raid can be more useful than raid-0, thanks.

> Well it contains stuff like TV and films a lot of which my friends have, 
> so loosing it is not that big of a deal just inconvenient. As I said 
> before I read linear was a better choice for odd sized drives and had 
> better data recovery that the other RAID's

I guess you're using it similarly to LVM then. 

http://en.wikipedia.org/wiki/Logical_Volume_Manager_(Linux)
http://tldp.org/HOWTO/LVM-HOWTO/

It sounds reasonable that recovery would be more feasible than RAID-0 but
neither would be nice to have to do.  RAID1, 5, etc. would be less likely
to require recovery as they have redundancy so a single disk failure
shouldn't take down the array at all.

> I'm sure I read somewhere on the mdadm page in one of the posts about 
> adding support for linear growing, and that was dated 2004 so it might 
> be in there? Just wasn't to sure how to do it...

This looks like your page.

http://cgi.cse.unsw.edu.au/~neilb/SoftRaid

and the mdadm changelog has it:

Changes Prior to 1.7.0 release
   -   Support "--grow --add" to add a device to a linear array, if the
       kernel supports it.  Not documented yet.

I've only grown raid5 arrays myself.  I imagine you must do --add to add
the extra disk partition to the array and then do "mdadm --grow
--size=max", same as with raid5.  

Gavin


  reply	other threads:[~2007-04-07 20:34 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-04-07 11:17 LINEAR RAID, little help Rich
2007-04-07 17:33 ` Gavin McCullagh
2007-04-07 18:41   ` Rich
2007-04-07 20:34     ` Gavin McCullagh [this message]
2007-04-10 21:49       ` Henrik Holst
2007-04-10 23:12     ` Neil Brown
     [not found]       ` <461DE4CA.9020406@pcfusion.co.uk>
2007-04-12  7:58         ` Rich
2007-04-12  8:08           ` Neil Brown
2007-04-12 11:19             ` matt s.
2007-04-12 14:55             ` Bill Davidsen

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20070407203435.GA32499@gmail.com \
    --to=gmccullagh@gmail.com \
    --cc=linux-raid@vger.kernel.org \
    --cc=rich@pcfusion.co.uk \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.