From mboxrd@z Thu Jan  1 00:00:00 1970
From: engler@csl.stanford.edu (Dawson Engler)
Subject: Re: Fix(es) for ext2 fsync bug
Date: Tue, 20 Feb 2007 13:47:55 -0800 (PST)
Message-ID: <20070220214756.8200519D60254@csl.stanford.edu>
References: <20070220213927.GG5264@nifty>
Reply-To: engler@csl.stanford.edu
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Cc: junfeng@gmail.com (Junfeng Yang),
	shaggy@linux.vnet.ibm.com (Dave Kleikamp),
	tytso@mit.edu (Theodore Tso), linux-fsdevel@vger.kernel.org,
	csar@stanford.edu (Can Sar)
To: val_henson@linux.intel.com (Valerie Henson)
Return-path: <linux-fsdevel-owner@vger.kernel.org>
Received: from agp.Stanford.EDU ([171.67.73.10]:42301 "EHLO agp.stanford.edu"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1030463AbXBTVsU (ORCPT <rfc822;linux-fsdevel@vger.kernel.org>);
	Tue, 20 Feb 2007 16:48:20 -0500
In-Reply-To: <20070220213927.GG5264@nifty> from "Valerie Henson" at Feb 20, 2007 09:39:29 PM
Sender: linux-fsdevel-owner@vger.kernel.org
List-Id: linux-fsdevel.vger.kernel.org

> On Tue, Feb 20, 2007 at 01:30:25PM -0800, Junfeng Yang wrote:
> > On 2/20/07, Valerie Henson <val_henson@linux.intel.com> wrote:
> > >
> > >Google. (GoogleFS runs on top of ext2.)
> >
> > It's surprising to know that... I guess they reply on GoogleFS's own
> > replication and checksumming for consistency.
> 
> Yep, they just want a local file system with ultrafast on-line
> performance.  They don't care about recovery time particularly because
> of the GoogleFS replication (although I heard rumors they have some
> fast fsck scheme, maybe resembling the dirty bit stuff I did last
> year).

Actually, according to the GFS paper (which may be out of date), for
the chunkservers that is true, but for their "master" they really want
fast recovery as a way to reduce mean-time-to-repair (and thus increase
availability).

Though, given that they have shadow masters perhaps everyone is happy as
long as master recovery usually fast.

Dawson