*BSD News Article 31420


Return to BSD News archive

Path: sserve!newshost.anu.edu.au!munnari.oz.au!constellation!convex!convex!cs.utexas.edu!usc!math.ohio-state.edu!jussieu.fr!univ-lyon1.fr!itesec!keltia.frmug.fr.net!not-for-mail
From: roberto@keltia.frmug.fr.net (Ollivier Robert)
Newsgroups: comp.os.386bsd.questions
Subject: Re: Linux or FreeBSD?
Date: 4 Jun 1994 01:56:24 +0200
Organization: A Happy FreeBSD-current Usenet Site
Lines: 19
Message-ID: <2sofvm$jnj@keltia.frmug.fr.net>
References: <2seo29$5c5@vishnu.jussieu.fr> <FRwaNc2w165w@oasys.pc.my>
Reply-To: roberto@hsc.fr.net (Ollivier Robert)
NNTP-Posting-Host: keltia.frmug.fr.net
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit

In article <FRwaNc2w165w@oasys.pc.my>, Othman Ahmad <othman@oasys.pc.my> wrote:
>card@masi.ibp.fr (Remy CARD) writes:
>The unique thing about BSD FFS(Fast File System) is the presense of 2 block
>sizes. A small chunk size but large block sizes. 16K block size with chunk
>size of 1 K is the default.

Nope, the default is the very  classical 8 KB  / 1KB. Strangely, I saw many
systems with default value of  4 KB / 512B which  is less efficient in term
of speed.

>Has this been implemented under ext2fs? From benchmark tests, it is very
>clear that linux FS a year ago still has not implemented it.

ASAIK Remy has done it or plan to add fragments to ext2fs (I am almost sure
he did it).
-- 
Ollivier ROBERT                                           roberto@hsc.fr.net
Hervé Schauer Consultants                Ollivier.Robert@keltia.frmug.fr.net
PERL / MIME / PGP 2.6ui     FreeBSD keltia 1.1.0(Current) CURRENT_092#0 i386