From owner-freebsd-current@FreeBSD.ORG Mon Jan 7 20:46:33 2008 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id D838816A419 for ; Mon, 7 Jan 2008 20:46:33 +0000 (UTC) (envelope-from thompsa@FreeBSD.org) Received: from heff.fud.org.nz (203-109-251-39.static.bliink.ihug.co.nz [203.109.251.39]) by mx1.freebsd.org (Postfix) with ESMTP id 30A5D13C44B for ; Mon, 7 Jan 2008 20:46:32 +0000 (UTC) (envelope-from thompsa@FreeBSD.org) Received: by heff.fud.org.nz (Postfix, from userid 1001) id 7CCB27305; Tue, 8 Jan 2008 09:46:31 +1300 (NZDT) Date: Tue, 8 Jan 2008 09:46:31 +1300 From: Andrew Thompson To: Ivan Voras Message-ID: <20080107204631.GG34409@heff.fud.org.nz> References: <86y7b168ay.fsf@ds4.des.no> <9bbcef730801070637v6db39e97o34b89b10dad75617@mail.gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <9bbcef730801070637v6db39e97o34b89b10dad75617@mail.gmail.com> User-Agent: Mutt/1.5.16 (2007-06-09) Cc: freebsd-current@freebsd.org Subject: Re: When will ZFS become stable? X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 07 Jan 2008 20:46:33 -0000 On Mon, Jan 07, 2008 at 03:37:26PM +0100, Ivan Voras wrote: > On 07/01/2008, Dag-Erling Sm??rgrav wrote: > > > Your question is based on the premise that ZFS in FreeBSD 7 is unstable. > > That premise is false. > > At most, we'll have to agree to disagree. A "tuning" of the system (at > least from my experience) is about system performance, not whether the > system will crash or not. You may define the word to mean something > else but that's your thing. > > The reason I'm aggressively discussing this is that labeling the > problem as "tuning" will, for any non-trivial task which has some > growth in system load, result in a server that needs constant tuning > just to survive another day. What is tuned today may as well result in > a crash tomorrow if the load rises. Web servers are notorious for this > (though other types have of course similar behaviour) - a > "slashdotting" of a "properly tuned" FreeBSD system with ZFS will not > result in a slowdown - it will result in the system crashing. This is > not acceptable, and therefore dismissing it as "just tuning" is > counterproductive and bad engineering. ZFS is clearly marked as experimental so its reasonable to require tuning to avoid crashes. If its still the case when the experimental status is lifted then you can have this argument all over again. cheers, Andrew