From owner-freebsd-questions  Fri Jun  2  4:58:56 2000
Delivered-To: freebsd-questions@freebsd.org
Received: from ibb0021.ibb.uu.nl (ibb0021.ibb.uu.nl [131.211.124.21])
	by hub.freebsd.org (Postfix) with ESMTP id 60C4637BEA9
	for <freebsd-questions@FreeBSD.ORG>; Fri,  2 Jun 2000 04:58:48 -0700 (PDT)
	(envelope-from mipam@ibb.net)
Received: by ibb0021.ibb.uu.nl (Postfix)
	id DF0ABF38; Fri,  2 Jun 2000 13:34:09 +0200 (CEST)
Date: Fri, 2 Jun 2000 13:34:09 +0200
From: Mipam <mipam@ibb.net>
To: Greg Lehey <grog@lemis.com>
Cc: William Freeman <wdf@picusnet.com>, freebsd-questions@FreeBSD.ORG
Subject: Re: xfs and freebsd?
Message-ID: <20000602133409.D377@ibb0021.ibb.uu.nl>
Reply-To: mipam@ibb.net
References: <20000403022421.A13027@ibb0021.ibb.uu.nl> <20000402193140.A43016@gwernache.picusnet.com> <20000403022421.A13027@ibb0021.ibb.uu.nl> <20000403101925.H42140@freebie.lemis.com>
Mime-Version: 1.0
Content-Type: multipart/signed; micalg=pgp-md5;
	protocol="application/pgp-signature"; boundary="opJtzjQTFsWo+cga"
Content-Disposition: inline
User-Agent: Mutt/1.2i
In-Reply-To: <20000403101925.H42140@freebie.lemis.com>; from grog@lemis.com on Mon, Apr 03, 2000 at 10:19:25AM +0930
X-Obviously: All email clients suck. Only Mutt sucks less!
X-Editor: Vi
X-Operating-System: BSD
Sender: owner-freebsd-questions@FreeBSD.ORG
Precedence: bulk
X-Loop: FreeBSD.ORG


--opJtzjQTFsWo+cga
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

> > Now xfs is comming into linux, what about the thought of xfs in
> > freebsd?
>=20
> We're investigating it.
>=20
> > and contains no inodes=20
>=20
> XFS uses inodes.
>

Okay, okay..... let me phrase it differently.
XFS uses a space manager to allocate disk space for the file system=20
and control the inodes.
Inodes are created as needed and are not restricted to a particular=20
area on a disk partition. XFS tries to position the
inodes close to the files and directories they reference.=20
Very small files, such as symbolic links and some directories, are
stored as part of the inode, to increase performance and save space.=20
Large directories use B-tree indexing within the
directory file to speed up directory searches, additions and deletions.
So in a way inodes are created dynamicly.

> > which is good for scalibility.
>=20
> Why?
>=20
Suppose you have an disk with 3 terabyte on a partition and you just created
an fs on it with the default options.
If you put millions of small files in many directories,
you'll run out of inodes and the rest
of the diskspace cant be used anymore cause of that.
And just changing the amount of bytes per inode isnt possible without
newfs the partition. Xfs isnt bothered by this limit and is therefor
scalable.
And of course the jornalling is cool and therefore the possibility
to recover in case of disaster is better then that of ufs.
So when i am very rude i say: xfs =3D ufs - limitations.
Of course, performance wise i cant say it.
But for a journalling fs the xfs performs well.
Bye,

Mipam.

--opJtzjQTFsWo+cga
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Comment: this is the gnupg Mipam key

iD8DBQE5N5uxsdVeYIaXThgRAgB3AJ48Yrx6ZTRE7InDWnVAePTEHBjL5QCgjTNS
C8xfPaX5s1+BBetI9Fv7Rfg=
=+daC
-----END PGP SIGNATURE-----

--opJtzjQTFsWo+cga--


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