From: Roger Binns <rogerb@rogerbinns.com>
To: linux-btrfs@vger.kernel.org
Subject: Re: [PATCH 0/5] [RFC] RAID-level terminology change
Date: Sun, 10 Mar 2013 04:48:18 -0700 [thread overview]
Message-ID: <khhru0$jug$1@ger.gmane.org> (raw)
In-Reply-To: <CAFWF=amO=G+BtqJxBNj8h_5Y7d8T=38xK4qnu=LZbUQyn3Zvew@mail.gmail.com>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On 09/03/13 22:37, Harald Glatt wrote:
> I have to add something to my own message: Even the notion of thinking
> in 'how many devices do I want to give away for redundancy' is
> outdated...
Devices are the only real tangible thing that you can actually point at.
They are the final physical unit and what gets added or replaced, and can
be put in another machine for recovery.
> What it really comes down to is how much space am I willing to
> sacrifice so that my reliablilty is increasing.
The answer as a whole is simple - you want all unused space for
reliability and performance. If you put 5GB of data on 10GB of space then
50% is the value.
> Rather than addressing that at a per-drive level with a futuristic fs
> like btrfs I think setting a percent value of total space would be
> best.
The problem is that drives are what fail, can be added or removed. A
percent value makes that considerably harder to deal with especially
changes after the initial setup.
It would be nice to be able to indicate that same data/files/directories
are more important than others (eg cache/spool/trash directories are
unimportant, my documents and photos are very important).
> This is pretty much what it would be like in a perfect world, in my
> opinion :)
Conceptually I want to point at the drives and say "do the right thing"
without any further configuration, monitoring or micro-managing. In the
short/medium term I'd even be happy to run a btrfs cron job that digs
around, rearranges things and gives me a final report with a colour coding
homeland security style. When it hits yellow, it is time to delete files
or add more storage.
Roger
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
iEYEARECAAYFAlE8cwIACgkQmOOfHg372QSjUACfcDn1lB/wHXZH9E5a5elUlAT3
Y+QAoL9cNtjMpdZtqYH+t6QXTcOYwYBy
=UMKJ
-----END PGP SIGNATURE-----
next prev parent reply other threads:[~2013-03-10 11:48 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-03-09 20:31 [PATCH 0/5] [RFC] RAID-level terminology change Hugo Mills
2013-03-09 20:31 ` [PATCH 1/5] Use nCmSpP format for mkfs Hugo Mills
2013-03-09 20:31 ` [PATCH 2/5] Move parse_profile to utils.c Hugo Mills
2013-03-09 20:31 ` [PATCH 3/5] Convert balance filter parser to use common nCmSpP replication-level parser Hugo Mills
2013-03-09 20:31 ` [PATCH 4/5] Change output of btrfs fi df to report new (or old) RAID names Hugo Mills
2013-03-09 20:31 ` [PATCH 5/5] Add man page description for nCmSpP replication levels Hugo Mills
2013-03-10 14:01 ` Goffredo Baroncelli
2013-03-10 17:20 ` Hugo Mills
2013-03-10 17:52 ` Goffredo Baroncelli
2013-03-09 21:38 ` [PATCH 0/5] [RFC] RAID-level terminology change Harald Glatt
[not found] ` <CAFWF=am4ki529Zez4123gYk3BD+Z9RONRpAK7NZe=skHzcdMiw@mail.gmail.com>
2013-03-09 21:46 ` Hugo Mills
2013-03-09 22:25 ` Roger Binns
2013-03-10 1:44 ` Hugo Mills
2013-03-10 5:41 ` Roger Binns
2013-03-10 6:29 ` Harald Glatt
2013-03-10 6:37 ` Harald Glatt
2013-03-10 11:31 ` Martin Steigerwald
2013-03-10 11:48 ` Roger Binns [this message]
2013-03-10 22:04 ` Hugo Mills
2013-03-11 0:21 ` Roger Binns
2013-03-27 4:27 ` Brendan Hide
2013-03-27 5:24 ` Roger Binns
2013-03-10 11:23 ` Martin Steigerwald
2013-03-10 14:11 ` Goffredo Baroncelli
2013-03-10 21:36 ` Hugo Mills
2013-03-10 21:45 ` Harald Glatt
2013-03-10 22:59 ` Goffredo Baroncelli
2013-03-10 22:06 ` Harald Glatt
2013-03-10 23:06 ` Diego Calleja
2013-03-10 15:43 ` Goffredo Baroncelli
2013-03-10 22:24 ` Hugo Mills
2013-03-10 22:42 ` Harald Glatt
2013-03-10 23:40 ` sam tygier
2013-03-10 23:49 ` Hugo Mills
2013-03-11 14:14 ` David Sterba
2013-03-10 23:55 ` sam tygier
2013-03-11 8:56 ` Hugo Mills
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='khhru0$jug$1@ger.gmane.org' \
--to=rogerb@rogerbinns.com \
--cc=linux-btrfs@vger.kernel.org \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox