*BSD News Article 48072


Return to BSD News archive

Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!simtel!zombie.ncsc.mil!news.mathworks.com!gatech!howland.reston.ans.net!xlink.net!rz.uni-karlsruhe.de!news.uni-stuttgart.de!informatik.uni-koblenz.de!ralf
From: ralf@informatik.uni-koblenz.de (Ralf Baechle)
Newsgroups: comp.unix.bsd.netbsd.misc
Subject: Re: NetBSD Filesystems
Date: 29 Jul 1995 21:35:48 GMT
Organization: Uni Koblenz, Germany.
Lines: 50
Distribution: world
Message-ID: <3ve9jk$11b8@info4.rus.uni-stuttgart.de>
References: <1995Jul26.123455.28242@lssec.bt.co.uk> <MICHAELV.95Jul29005359@MindBender.HeadCandy.com>
Reply-To: ralf@waldorf-gmbh.de
NNTP-Posting-Host: ozzy.uni-koblenz.de

In article <MICHAELV.95Jul29005359@MindBender.HeadCandy.com>, michaelv@MindBender.HeadCandy.com (Michael L. VanLoon) writes:
|> In article <1995Jul26.123455.28242@lssec.bt.co.uk> bwheeler@lssec.bt.co.uk (Ben/Jammin Wheeler) writes:
|> 
|>    Here are my questions:
|>    Does NetBSD only have its own ``proprietary'' filesystem?
|> 

[...]
 
|>    Is there any support for other filesystems, specifically MS-DOG FAT, Minix, Linux
|>    Ext2, Linux Xiafs?

You don't really want XiaFS?  I don't see any real argument for it except
really paranoid compatibility.

|> NetBSD will let you mount a DOS partition read/write for import/export
|> purposes, but I don't believe you can actually run the system off a
|> DOS partition.  It does not understand Linux "proprietary"
|> filesystems.  Though someone some day may get the urge to write the
|> code to make that possible.  Most likely it would be for compatibility
|> import/export purposes also, since there is nothing inherently better
|> about the Linux filesystems vs. BSD 4.4 FFS (nor inherently worse).
|> Some say the Linux filesystem isn't quite as robust or well-tested as
|> FFS, but the Linux people deny that vehemently.

Personal experience about FFS reliability: never more than just minor damage
even on heavily loaded machines.  The same applies to Linux' ext2fs though
it is in theory a bit more vulnerable to corruption as long as not mounted
with -o sync.  The performance specially for handling lots of small files
is everything else but not comparable.  I found ext2fs MUCH faster due to it's
asynchronous updates of the fs meta information for that case.

|> NetBSD only "supports its own little filesystem" in the same way Linux
|> only "supports its own little filesystem".  The difference being that
|> the NetBSD "proprietary" fs is a derivative of one of the most
|> standard filesystems in unix.  The Linux folks, on the other hand,
|> have taken a religious position that FFS is not good enough for them,
|> written their own proprietary fs, and refuse to write a true FFS for
|> Linux, last I heard.  For all intents and purposes, both filesystems
|> are pretty much functionally equivalent.

There is a Linux kernel patch to add a FFS filesystem to the kernel. Since
however since quite some time no new version of this patch is available I'm
not shure if development still goes on.  Since I didn't try the patch myself
I can't say very much else about except that it only supports read access.
To sad - even if FFS shouldn't be the best fs for a UNIX system it is
definately the best native UNIX fs that is spread across lots of kernel
variants.

  Ralf