From owner-svn-src-stable-8@FreeBSD.ORG Mon Dec 13 20:04:57 2010 Return-Path: Delivered-To: svn-src-stable-8@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 62DB8106567A; Mon, 13 Dec 2010 20:04:57 +0000 (UTC) (envelope-from rwatson@FreeBSD.org) Received: from cyrus.watson.org (cyrus.watson.org [65.122.17.42]) by mx1.freebsd.org (Postfix) with ESMTP id 282C98FC17; Mon, 13 Dec 2010 20:04:57 +0000 (UTC) Received: from [192.168.2.102] (host86-163-246-141.range86-163.btcentralplus.com [86.163.246.141]) by cyrus.watson.org (Postfix) with ESMTPSA id 5955946B2E; Mon, 13 Dec 2010 15:04:54 -0500 (EST) References: <201012131234.oBDCYZvl082510@svn.freebsd.org> <20101213160021.60401b79@it.buh.tecnik93.com> <20101213202830.75fbce4c@it.buh.tecnik93.com> In-Reply-To: <20101213202830.75fbce4c@it.buh.tecnik93.com> Mime-Version: 1.0 (iPad Mail 8C148) Message-Id: X-Mailer: iPad Mail (8C148) From: "Robert N. M. Watson" Date: Mon, 13 Dec 2010 20:07:05 +0000 To: Ion-Mihai Tetcu Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: quoted-printable X-Content-Filtered-By: Mailman/MimeDel 2.1.5 Cc: "svn-src-stable@freebsd.org" , "svn-src-all@freebsd.org" , "src-committers@freebsd.org" , "Robert N. M. Watson" , "svn-src-stable-8@freebsd.org" Subject: Re: svn commit: r216406 - stable/8/sys/amd64/conf X-BeenThere: svn-src-stable-8@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: SVN commit messages for only the 8-stable src tree List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 13 Dec 2010 20:04:57 -0000 On 13 Dec 2010, at 18:28, Ion-Mihai Tetcu wrote: > I see. Thanks for the time you took to explain things, I'm very new to > XEN. > Do we have any docs somewhere (well, except the sources) about our XEN > support and various options / optimizations? I only know about > AdrianChadd's wiki pags and http://wiki.freebsd.org/FreeBSD/Xen (which > are nice, if you happen to know about XEN already and especially if you > want to work towards a FreebSD as XEN host). > I guess there are more interested people taht want just to run FreeBSD > as guest OS (my case). >>=20 The current state of our Xen documentation is ... minimalist. Having basical= ly come to it for the first time last week, when we decided we needed to dep= loy FreeBSD on one of our Xen server pools, it's proven pretty hard-going. I= have some more technical problems to resolve but high on my todo list for t= he next couple of days is documenting what has worked for me on the Wiki (wi= th the intent of migrating that to the Handbook once I've confirmed my instr= uctions are reproducible!). I also want to drop in at least some stub man pa= ges for our PV drivers. However, more fundamentally, we appear to have some serious functional gaps:= - No ability to use PV drivers under HVM on i386. - No ability to use a full PV kernel for amd64 (I don't consider this a prob= lem for my use though -- HVM on amd64 appears to work well). - No (obvious) port of Xen guest command line tools to steer ballooning, etc= . - Likely under-testing in certain PV drivers, such as the balloon driver. - A possible issue with VM migration on PV i386 (we saw panics in local test= ing, but now wonder I it was an unrelated issue relating to different CPU ty= pes -- more investigation due). - No event mechanism for blkfront to propagate block device expansion up the= geom stack, in support of live file system resizing. Right now a destructiv= e orphan/retaste is required. - No live file system resizing support in UFS (does work in ZFS). - No reference "appliance" images for VM starting points (really requires li= ve file system resizing to be useful). This was a top request o our admin fo= lks. All this said, and after quite a lot of head-scratching, FreeBSD/amd64 seems= to work well on Xen here a Cambridge, both with and without Xen PV drivers c= ompiled in. We need a lot of sanding and polishing but the foundations have b= een usefully laid. And I would agree that documentation is key. Robert=