LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Andi Kleen <andi@firstfloor.org>
To: Juan Piernas Canovas <piernas@ditec.um.es>
Cc: Jan Engelhardt <jengelh@linux01.gwdg.de>,
sfaibish <sfaibish@emc.com>,
kernel list <linux-kernel@vger.kernel.org>
Subject: Re: [ANNOUNCE] DualFS: File System with Meta-data and Data Separation
Date: 15 Feb 2007 21:38:18 +0100 [thread overview]
Message-ID: <p73sld7f96t.fsf@bingen.suse.de> (raw)
In-Reply-To: <Pine.LNX.4.61.0702151910370.19998@ditec.inf.um.es>
Juan Piernas Canovas <piernas@ditec.um.es> writes:
[playing devil's advocate here]
> If the data and meta-data devices of DualFS can be on different disks,
> DualFS is able to READ and WRITE data and meta-data blocks in
> PARALLEL.
XFS can do this too using its real time volumes (which don't contain
any metadata). It can also have a separate log.
Also many storage subsystems have some internal parallelism
in writing (e.g. a RAID can write on different disks in parallel for
a single partition) so i'm not sure your distinction is that useful.
If you stripe two disks with a standard fs versus use one of them
as metadata volume and the other as data volume with dualfs i would
expect the striped variant usually be faster because it will give
parallelism not only to data versus metadata, but also to all data
versus other data.
Also I would expect your design to be slow for metadata read intensive
workloads. E.g. have you tried to boot a root partition with dual fs?
That's a very important IO benchmark for desktop Linux systems.
-Andi
next prev parent reply other threads:[~2007-02-15 19:38 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <op.tnkdlbgsrwwil4@brcsmondepl2c.corp.emc.com>
2007-02-14 21:10 ` sfaibish
2007-02-14 21:57 ` Jan Engelhardt
2007-02-15 18:38 ` Juan Piernas Canovas
2007-02-15 20:09 ` Jörn Engel
2007-02-15 22:59 ` Juan Piernas Canovas
2007-02-16 9:13 ` Jörn Engel
2007-02-16 11:05 ` Benny Amorsen
2007-02-16 23:47 ` Bill Davidsen
2007-02-17 15:11 ` Jörn Engel
2007-02-17 18:10 ` Bill Davidsen
2007-02-17 18:36 ` Jörn Engel
2007-02-17 20:47 ` Sorin Faibish
2007-02-18 5:59 ` Jörn Engel
2007-02-18 12:46 ` Jörn Engel
2007-02-19 23:57 ` Juan Piernas Canovas
2007-02-20 0:10 ` Bron Gondwana
2007-02-20 0:30 ` Jörn Engel
2007-02-21 4:36 ` Juan Piernas Canovas
2007-02-21 12:37 ` Jörn Engel
2007-02-21 18:31 ` Juan Piernas Canovas
2007-02-21 19:25 ` Jörn Engel
2007-02-22 4:30 ` Juan Piernas Canovas
2007-02-22 16:25 ` Jörn Engel
2007-02-22 19:57 ` Juan Piernas Canovas
2007-02-23 13:26 ` Jörn Engel
2007-02-24 22:35 ` Sorin Faibish
2007-02-25 2:41 ` Juan Piernas Canovas
2007-02-25 12:01 ` Jörn Engel
2007-02-26 3:48 ` Juan Piernas Canovas
2007-02-20 20:43 ` Bill Davidsen
2007-02-15 20:38 ` Andi Kleen [this message]
2007-02-15 19:46 ` Jan Engelhardt
2007-02-16 1:43 ` sfaibish
2007-02-15 21:09 ` Juan Piernas Canovas
2007-02-15 23:57 ` Andi Kleen
2007-02-16 4:57 ` Juan Piernas Canovas
2007-02-26 11:49 ` Yakov Lerner
2007-02-26 13:08 ` Matthias Schniedermeyer
2007-02-26 13:24 ` Sorin Faibish
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=p73sld7f96t.fsf@bingen.suse.de \
--to=andi@firstfloor.org \
--cc=jengelh@linux01.gwdg.de \
--cc=linux-kernel@vger.kernel.org \
--cc=piernas@ditec.um.es \
--cc=sfaibish@emc.com \
--subject='Re: [ANNOUNCE] DualFS: File System with Meta-data and Data Separation' \
/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
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).