From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from [140.186.70.92] (port=57949 helo=eggs.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1PAJm7-0006m0-OT for qemu-devel@nongnu.org; Mon, 25 Oct 2010 05:55:33 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1PAJd7-0007ET-JB for qemu-devel@nongnu.org; Mon, 25 Oct 2010 05:46:14 -0400 Received: from mx1.redhat.com ([209.132.183.28]:27797) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1PAJd7-0007EG-CY for qemu-devel@nongnu.org; Mon, 25 Oct 2010 05:46:13 -0400 Message-ID: <4CC551DF.70201@redhat.com> Date: Mon, 25 Oct 2010 11:46:07 +0200 From: Andrew Beekhof MIME-Version: 1.0 References: <1287773165-24855-1-git-send-email-mdroth@linux.vnet.ibm.com> <4CC2C79D.2010504@redhat.com> <4CC2F435.1020909@linux.vnet.ibm.com> In-Reply-To: <4CC2F435.1020909@linux.vnet.ibm.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Subject: [Qemu-devel] Re: [RFC][PATCH 00/10] virtagent: host/guest RPC communication agent List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Michael Roth Cc: aliguori@linux.vnet.ibm.com, ryanh@us.ibm.com, agl@linux.vnet.ibm.com, qemu-devel@nongnu.org On 10/23/2010 04:41 PM, Michael Roth wrote: > On 10/23/2010 06:31 AM, Andrew Beekhof wrote: >> On 10/22/2010 08:45 PM, Michael Roth wrote: >>> the ability to deploy to guests which may not support virtio-serial, >>> which currently rules Matahari out. >> >> Possibly there has been some miscommunication, Matahari does not require >> virtio-serial. >> >> It will be possible to configure Matahari to use virtio-serial if it >> exists, but like Virtproxy it can also be used over a regular IP network >> (yes, its transparent to clients too). >> > > Sorry, I'm aware of this but wasn't very clear here. I was speaking in > the context of our cloud, or other environments where there isn't any > network access between the host and the guest. This essentially leaves > virtio-serial and isa-serial, and a large number of our guests don't > support the former. Well again, Matahari has no concept of what transport is being used. That's completely hidden from us by Qpid behind a layer of abstraction. So what we're really talking about is the difficulty in adding a new transport to Qpid... and if the interface is anything like virtio-serial or a standard serial port, we're talking about a few hundred lines and "a couple of days". Not exactly a major barrier considering the pay-off would be a single set of agents.