From owner-freebsd-current@FreeBSD.ORG Mon May 21 18:47:44 2012 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [69.147.83.52]) by hub.freebsd.org (Postfix) with ESMTP id 8E1AA1065764 for ; Mon, 21 May 2012 18:47:44 +0000 (UTC) (envelope-from jamie@geniegate.com) Received: from web.podro.com (unknown [IPv6:2607:f7d0:300:1::100]) by mx1.freebsd.org (Postfix) with ESMTP id 46BEB8FC1B for ; Mon, 21 May 2012 18:47:44 +0000 (UTC) Received: from web.podro.com (web.podro.com [65.18.192.166]) by web.podro.com (8.14.4/8.14.4) with ESMTP id q4LIlggc012914; Mon, 21 May 2012 13:47:42 -0500 (CDT) (envelope-from jamie@geniegate.com) Received: from ns1.podro.com (ns1.podro.com [65.18.192.210]) by web.podro.com (8.14.4/8.14.4/Submit) id q4LIle1W012911; Mon, 21 May 2012 13:47:40 -0500 (CDT) (envelope-from jamie@geniegate.com) X-Authentication-Warning: web.podro.com: joe set sender to jamie@geniegate.com using -f Date: Mon, 21 May 2012 13:47:39 -0500 From: Jamie To: Dag-Erling Sm??rgrav Message-ID: <20120521184739.GA22790@mule.podro.com> References: <20120520170702.GY22790@mule.podro.com> <1120936952.633013.1337546532026.JavaMail.root@erie.cs.uoguelph.ca> <20120521035323.GZ22790@mule.podro.com> <868vglud1e.fsf@ds4.des.no> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <868vglud1e.fsf@ds4.des.no> User-Agent: Mutt/1.5.21 (2010-09-15) X-Mailman-Approved-At: Mon, 21 May 2012 19:33:56 +0000 Cc: Jamie , Rick Macklem , Vincent Hoffman , freebsd-current@freebsd.org, freebsd-chat@freebsd.org, Vance Siemens Subject: Re: FreeBSD 10 prognostication... 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, 21 May 2012 18:47:44 -0000 On Mon, May 21, 2012 at 10:57:33AM +0200, Dag-Erling Sm??rgrav wrote: > No, they're not. VMWare, RHEV (KVM-based) etc. provide features such as > seamless migration of virtual machines from one physical machine to > another, automatic restart on a different physical server if one fails > etc. that simply aren't possible with jails; and there are certain > things you still can't run reliably / safely in jails - anything that > relies on SysV IPC, for instance, such as PostgreSQL. True about the SysV, and I mostly agree about automatic failover. But I think the FreeBSD jail system is still the better model for how I see these things being used (certainly the better *potential*). But yea, not "quite" cloud. When coupled with something like rsync, they *almost* do the job. And for a lot of the current "VPS" applications, they do the job. But lets suppose you want proper redundancy and partitioned environments, so, you put FreeBSD on a cloud, but partition your environments into jails. Now you have a cheap, low overhead way of doing logical partitioning and you still have a "cloud" with redundancy. Linux KVM has serious network issues, and overhead. (I use KVM a fair amount, it's OK for testing, but when under load, KVM-based linux hosting sucks for serious use) If the goal is to get FreeBSD uber popular, then turn the alphas on to the jail system. (or the pre-emptive swapping, or.. or.. or...) I don't believe it's about glitter as much as people say it is, at least, not anymore. I threw jails out there because I personally consider them to be the coolest part of FreeBSD. Jamie