From owner-freebsd-questions  Tue Oct 17  7:28:16 2000
Delivered-To: freebsd-questions@freebsd.org
Received: from boat.mail.pipex.net (our.mail.pipex.net [158.43.128.75])
	by hub.freebsd.org (Postfix) with SMTP id 699E137B4E5
	for <freebsd-questions@FreeBSD.ORG>; Tue, 17 Oct 2000 07:28:12 -0700 (PDT)
Received: (qmail 21210 invoked from network); 17 Oct 2000 14:28:11 -0000
Received: from mailhost.puck.pipex.net (HELO mailhost.uk.internal) (194.130.147.54)
  by our.mail.pipex.net with SMTP; 17 Oct 2000 14:28:11 -0000
Received: (qmail 8388 invoked from network); 17 Oct 2000 14:28:10 -0000
Received: from camgate2.cam.uk.internal (172.31.6.21)
  by mailhost.uk.internal with SMTP; 17 Oct 2000 14:28:10 -0000
Received: by camgate2.cam.uk.internal with Internet Mail Service (5.5.2650.21)
	id <48RBLPQS>; Tue, 17 Oct 2000 15:27:10 +0100
Message-ID: <FB7CAC781DB6D311BEE800805FE6FADA2F4CB4@camexch4.cam.uk.internal>
From: Daniel Bye <Daniel.Bye@uk.uu.net>
To: 'Micke Josefsson' <mj@isy.liu.se>
Cc: FBSD-Q <freebsd-questions@FreeBSD.ORG>,
	Odhiambo Washington <wash@iconnect.co.ke>
Subject: RE: Defragmentation
Date: Tue, 17 Oct 2000 15:21:56 +0100
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2650.21)
Content-Type: text/plain;
	charset="iso-8859-1"
Sender: owner-freebsd-questions@FreeBSD.ORG
Precedence: bulk
X-Loop: FreeBSD.ORG

Agreed.  I have never seen anything above 3%, on a disk that has been
hammered for months.  I would only start to worry if disk usage were above,
say, 80% AND I saw high fragmentation (above 5-6%).  The work involved in
taking a copy of the FS and rebuuilding disk setups (or manually copying and
deleting files :o\  ) is too much to contemplate!

I didn't know about tunefs, though, so thanks for that gem!

Dan

> -----Original Message-----
> From: Micke Josefsson [mailto:mj@isy.liu.se]
> Sent: Tuesday, October 17, 2000 3:20 PM
> To: Daniel Bye
> Cc: FBSD-Q; FBSD-Q; Odhiambo Washington
> Subject: RE: Defragmentation
> 
> 
> 
> On 17-Oct-00 Daniel Bye wrote:
> > As far as I know, there are no defragmentation agents for 
> FreeBSD (or most
> > other *NICES).  Therefore, you would have to rebuild your 
> file systems.
> > Take a tape backup of the entire system (this is sequential, so disk
> > fragmentation is not an issue), rebuild your disk 
> slice/partition setup, and
> > then restore the backed up file systems from the tape.
> 
> 
> ffs tries to minimise fragmentation when writing files so 
> simply *copying* files
> around should reallocate them with less framentation (then 
> delete the original
> of course:). Moving files is not the same, as the actual 
> inodes must be moved
> around for the system to get a chance of optimizing it.
> 
> How many percent fragmentation is there? I have never seen 
> more than a couple of
> percents worth. If the file system is full ffs will have a 
> harder time doing a
> good job, in which case perhaps using tunefs to set size 
> optimization is a good
> thing?
> 
> > However, you shouldn't need to do this as often as under MS 
> systems, as the
> > various UNIX file systems aggressively try to minimise data 
> fragmentation.
> > 
> > If anyone knows of a defragmentation agent, I'd like to know :o)
> > 
> 
> Never heard of it either. I don't think it is necessary.
> 
> 
> ----------------------------------
> Michael Josefsson, MSEE
> mj@isy.liu.se
> 
> This message was sent by XFMail
> running on FreeBSD 3.5-STABLE
> ----------------------------------
> 


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