Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 2 Feb 2004 17:52:02 +1100
From:      Peter Jeremy <PeterJeremy@optushome.com.au>
To:        Oliver Brandmueller <ob@e-Gitt.NET>
Cc:        current@freebsd.org
Subject:   Re: max blockdevice/filesystem size?
Message-ID:  <20040202065202.GQ908@cirb503493.alcatel.com.au>
In-Reply-To: <20040131130337.GH774@cicely12.cicely.de>
References:  <20040130115025.GB96881@e-Gitt.NET> <200401310143.51186.freebsd-current@webteckies.org> <20040131114548.GA2632@e-Gitt.NET> <20040131130337.GH774@cicely12.cicely.de>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Jan 31, 2004 at 02:03:38PM +0100, Bernd Walter wrote:
>One point for a 10T UFS2 filesystem is a limit of 2G inodes.

It shouldn't - ino_t is unsigned.

>Someone mentioned a problem in another thread with more inodes.

It's not clear that his problem was definitely caused by having
more than 2G inodes - though it is possible.

>Another point is the fsck memory footprint when checking such a
>filesystem - that should very much depend on your newfs args and
>number of files.

Not to mention the disk space eaten by the inodes and time to
actually perform an fsck.

>Otherwise it should just work.

It's probably a good idea to adjust the blocksize, fragsize, cylinders
per group and bytes per inode to suit the expected number of files.  A
larger blocksize (up to a certain point) will improve I/O performance
and increase the maximum cylinder size.  Fewer cylinders and fewer
inodes will speed up fsck.

Peter



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040202065202.GQ908>