From owner-freebsd-fs Fri Feb 14 7:59:41 2003 Delivered-To: freebsd-fs@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 0E35237B401 for ; Fri, 14 Feb 2003 07:59:40 -0800 (PST) Received: from sluis.cs.vu.nl (sluis.cs.vu.nl [130.37.30.210]) by mx1.FreeBSD.org (Postfix) with ESMTP id 2435243FDD for ; Fri, 14 Feb 2003 07:59:39 -0800 (PST) (envelope-from philip@cs.vu.nl) Received: from centaur.cs.vu.nl (centaur.cs.vu.nl [130.37.30.211]) by sluis.cs.vu.nl with esmtp (Smail #86) id m18jiFV-000OXKC; Fri, 14 Feb 2003 16:59 +0100 Received: from cs.vu.nl (localhost [127.0.0.1]) by centaur.cs.vu.nl with esmtp (Smail #86) id m18jiFU-001cz2C; Fri, 14 Feb 2003 16:59 +0100 Message-Id: To: freebsd-fs@FreeBSD.ORG Subject: Re: Why is there no JFS? References: <3E4A5B77.5080103@ant.uni-bremen.de> <3E4A863E.2030801@potentialtech.com> <3E4A8EF5.1070308@ant.uni-bremen.de> <3E4A9712.8030609@potentialtech.com> <3E4AA331.5040701@ant.uni-bremen.de> <3E4AA734.5040102@potentialtech.com> <045401c2d2db$f9d45c30$0a0aa8c0@dweebsoft.com> <15946.52429.222082.74590@guru.mired.org> Date: Fri, 14 Feb 2003 16:59:36 +0100 From: Philip Homburg Sender: owner-freebsd-fs@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org In your letter dated Wed, 12 Feb 2003 16:38:05 -0600 you wrote: >In <045401c2d2db$f9d45c30$0a0aa8c0@dweebsoft.com>, Daxbert soft.com> typed: >> Is anybody currently working on or does there exist >> a JFS for FreeBSD? > >To the best of my knowledge, there is no JFS, and nobody is working on >one. I'm working on one. Don't expect results soon. >> Is there not a JFS for FreeBSD becuase, Softupdates >> do the job just fine and nobody has the >> time or interest to work on this? > >Softupdates with a background fsck solve the problem of wanting to >come back up quickly after a crash, which is the most common reason >people ask for a JFS. The main feature that I like from the system I'm working on is that (write) system calls can be made atomic. If you write 1MB and crash, either the entire write will be recovered, or you won't see any effects of the write. It is possible to improve on softupdates in the area of block allocation and write scheduling, but I'm not sure it is worth the trouble. I'd like to be able to group a collection of related write system calls in a transaction, but I'm not sure what the (system call) interface should look like. Philip Homburg To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-fs" in the body of the message