*BSD News Article 61401


Return to BSD News archive

Path: euryale.cc.adfa.oz.au!newshost.anu.edu.au!harbinger.cc.monash.edu.au!news.mel.connect.com.au!munnari.OZ.AU!metro!metro!inferno.mpx.com.au!news.mel.aone.net.au!imci4!newsfeed.internetmci.com!howland.reston.ans.net!math.ohio-state.edu!magnus.acs.ohio-state.edu!csn!ub!newsstand.cit.cornell.edu!news.tc.cornell.edu!newsserver.sdsc.edu!news.cerf.net!proton.llumc.edu!galaxy.ucr.edu!corsa!grif
From: grif@corsa.ucr.edu (Michael Griffith)
Newsgroups: comp.unix.bsd.freebsd.misc,comp.os.linux.development.system
Subject: Re: The better (more suitable)Unix?? FreeBSD or Linux
Date: 14 Feb 1996 01:30:03 GMT
Organization: UC Riverside, Dept. of Computer Science
Lines: 67
Message-ID: <4frdur$hq@galaxy.ucr.edu>
References: <4er9hp$5ng@orb.direct.ca> <311C5EB4.2F1CF0FB@freebsd.org> <4fjodg$o8k@venger.snds.com> <4fo1tu$n31@news.jf.intel.com>
NNTP-Posting-Host: cs.ucr.edu
Xref: euryale.cc.adfa.oz.au comp.unix.bsd.freebsd.misc:13762 comp.os.linux.development.system:17306

In article <4fo1tu$n31@news.jf.intel.com>,
Mike Haertel <haertel@ichips.intel.com> wrote:
|In article <4fjodg$o8k@venger.snds.com>,
|Michael Griffith <grif@hill.ucr.edu> wrote:
|>	Writing metadata synchronously and data asynchronously
|>	can put a filesystem in a state that has undetectable 
|>	errors.
|
|For some reason, in this sync-vs-async metadata discussion,
|someone always brings up the claim that synchronous metadata
|updates can leave you in a situation where fsck reports no
|errors, yet you can have bogus data.
|
|True.

Ahh.

|Then they claim that async metadata update is superior,
|because it doesn't have this problem.
|
|FALSE!

You are quite correct.  If I was misleading in this regard, I
apologize.  The real intent of the discussion was to show that async
was no worse than sync metadata.  However, if you add ordered writes,
you eliminate the problem.

|It could perfectly well happen that your async metadata
|update might purely by chance choose a block ordering
|which would leave the file system structure consistent,
|yet leave data blocks out of date.
|
|So this claim is bogus.
|
|
|As far as I know, the only practical difference between
|sync and async metadata update is that async results in
|faster write performance, especially for small files.

The performance implications are quite substantial AND sync
metadata doesn't really gain you anything in terms of reliability
(it may actually hurt a bit, because you are more likely to have
unordered writes.)  Given this, why bother with sync metadata?

|There may be reliability differences, but I have yet to
|see any adequate empirical results to convince me.  My gut
|feeling is that async requires a more sophisticated fsck,
|and that even so the resulting fsck may be more easily
|confused.  However, this is just a gut feeling.  It would
|be really neat if someone would do the reliability study
|that Jordon suggested...

Having inconsistent filesystem structures really isn't the issue.  A
hard failure where you know you have to restore from backups because
fsck can't figure things out is a lot better than silently corrupting
data.

The study may be worthwhile, but I am holding out for ordered writes.
A comparison between properly ordered writes and the current situation
would me much more interesting.  Anybody have good ideas for the
setup of the experiments?

-- 
Michael A. Griffith (grif@cs.ucr.edu) | http://www.cs.ucr.edu/~grif/
Department of Computer Science        | PGP public key available.
University of California, Riverside   | "My freedom of speech implies
(909) 787-3803                        |  your freedom to be offended."