From owner-freebsd-cluster@FreeBSD.ORG Tue Jun 17 12:05:52 2003 Return-Path: Delivered-To: freebsd-cluster@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 3080537B401 for ; Tue, 17 Jun 2003 12:05:52 -0700 (PDT) Received: from walnut.he.net (walnut.he.net [64.71.137.114]) by mx1.FreeBSD.org (Postfix) with ESMTP id B328643FB1 for ; Tue, 17 Jun 2003 12:05:51 -0700 (PDT) (envelope-from kmacy@fsmware.com) Received: from localhost (kmacy@localhost) by walnut.he.net (8.8.6p2003-03-31/8.8.2) with ESMTP id MAA02915; Tue, 17 Jun 2003 12:05:53 -0700 Date: Tue, 17 Jun 2003 12:05:53 -0700 (PDT) From: Kip Macy X-Sender: kmacy@walnut.he.net To: Michael Grant In-Reply-To: <20030617184938.GA1078@grant.org> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII cc: freebsd-cluster@freebsd.org Subject: Re: iSCSI and clustering with FreeBSD X-BeenThere: freebsd-cluster@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Clustering FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 17 Jun 2003 19:05:52 -0000 > OpenGFS and Lustre were mentioned. Has anyone actually tried to > simply compile OpenGFS for freebsd just fixing the compilation > problems, or is it really a porting job with lots of rewrite? Would > it solve my problems? (see below) I track the OpenGFS development list and it is clear that it has a ways to go with respect to robustness of lock management. I don't think that porting would be that major an undertaking. However, It almost seems that adding clustering to an existing journalled file system would be easier than stabilizing OpenGFS. Lustre is designed for extremely large local clusters. > Furthermore, I'm looking at clustering where the other boxes in the > cluster are not at the same ISP. This adds additional headache for > sure. If anyone has any ideas along these lines, let me know. I want > to do this for reliability and load sharing, not for supercomputing. > However, it sure would be cool to be able to migrate a process to a > different box at a different location... This is not at all what OpenGFS is designed for. That is something that AFS would be perfect for. However, Transarc botched it with respect to the market. At least in the medium term NFSv4 is probably the closest you're going to get. > > It looks like no matter what I do, I need a second box next to the > first one to redirect packets to the other box if one of the boxes > goes down. I'd probably do this with NAT or an ip tunnel. This > second box almost makes it seem not worthwhile to put the other boxes > in different ISPs. Anyone have better ideas? I'm sure some of the ASPs must have solved that problem.