From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail203.messagelabs.com (mail203.messagelabs.com [216.82.254.243]) by kanga.kvack.org (Postfix) with ESMTP id 1E7216B004D for ; Wed, 30 Sep 2009 13:45:26 -0400 (EDT) Received: from d06nrmr1707.portsmouth.uk.ibm.com (d06nrmr1707.portsmouth.uk.ibm.com [9.149.39.225]) by mtagate4.uk.ibm.com (8.14.3/8.13.8) with ESMTP id n8UI1P0x026376 for ; Wed, 30 Sep 2009 18:01:30 GMT Received: from d06av03.portsmouth.uk.ibm.com (d06av03.portsmouth.uk.ibm.com [9.149.37.213]) by d06nrmr1707.portsmouth.uk.ibm.com (8.13.8/8.13.8/NCO v10.0) with ESMTP id n8UI19CR2588794 for ; Wed, 30 Sep 2009 19:01:15 +0100 Received: from d06av03.portsmouth.uk.ibm.com (loopback [127.0.0.1]) by d06av03.portsmouth.uk.ibm.com (8.12.11.20060308/8.13.3) with ESMTP id n8UI182r030578 for ; Wed, 30 Sep 2009 19:01:09 +0100 Message-ID: <4AC39CE5.9080908@free.fr> Date: Wed, 30 Sep 2009 20:01:09 +0200 From: Daniel Lezcano MIME-Version: 1.0 Subject: Re: [PATCH 00/80] Kernel based checkpoint/restart [v18] References: <1253749920-18673-1-git-send-email-orenl@librato.com> <20090924154139.2a7dd5ec.akpm@linux-foundation.org> <20090928163704.GA3327@us.ibm.com> <4AC20BB8.4070509@free.fr> <87iqf0o5sf.fsf@caffeine.danplanet.com> <4AC38477.4070007@free.fr> <87eipoo0po.fsf@caffeine.danplanet.com> In-Reply-To: <87eipoo0po.fsf@caffeine.danplanet.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org To: Dan Smith Cc: "Serge E. Hallyn" , linux-api@vger.kernel.org, containers@lists.linux-foundation.org, linux-kernel@vger.kernel.org, linux-mm@kvack.org, Andrew Morton , torvalds@linux-foundation.org, mingo@elte.hu, xemul@openvz.org List-ID: Dan Smith wrote: > DL> If the checkpoint is done from the kernel, why the restart > DL> wouldn't be in the kernel too ? > > I think thus far we have taken the approach of "if it can be done > reasonably in userspace, then do it there" right? Well I am a little lost :) The tty CR can be "reasonably" done in userspace I think. But it was done in the kernel, no ? > Setup of the > network devices is easy to do in userspace, allows more flexibility > from a policy standpoint, and ensures that all existing security > checks are performed. Yep, I agree. But you didn't answer the question, what are the network resources you plan to checkpoint / restart ? eg. you let the container to setup your network, will you restore netdev statistics ? the mac address ? ipv4 ? ipv6 ? Is it possible to do a detailed list of network resources you plan to CR with the different items you will address from userspace and kernel space ? > Also, migration may be easier if the userspace > bits can call custom hooks allowing for routing changes and other > infrastructure-specific operations. > You may have some problems with the connected sockets you will restore in this case. > DL> Is there any documentation about the statefile format I can use if > DL> I want to implement myself an userspace CR solution based on this > DL> kernel patchset ? > > See linux-cr/include/linux/checkpoint_hdr.h and user-cr/restart.c. > Argh ! I was hoping there was something else than the source code :) Thanks -- Daniel -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org