From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263990AbUD2JmU (ORCPT ); Thu, 29 Apr 2004 05:42:20 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S264002AbUD2JmU (ORCPT ); Thu, 29 Apr 2004 05:42:20 -0400 Received: from jive.SoftHome.net ([66.54.152.27]:14029 "HELO jive.SoftHome.net") by vger.kernel.org with SMTP id S263990AbUD2JmR (ORCPT ); Thu, 29 Apr 2004 05:42:17 -0400 Message-ID: <4090CDF5.5030503@softhome.net> Date: Thu, 29 Apr 2004 11:42:13 +0200 From: "Ihar 'Philips' Filipau" User-Agent: Mozilla Thunderbird 0.5+ (Macintosh/20040414) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Neal Becker CC: Linux Kernel ML Subject: Re: State of linux checkpointing? References: <1Q3nZ-2hv-3@gated-at.bofh.it> <1Q40J-2Mx-9@gated-at.bofh.it> <1Q6F8-53t-1@gated-at.bofh.it> <1Q8xl-6x2-17@gated-at.bofh.it> In-Reply-To: <1Q8xl-6x2-17@gated-at.bofh.it> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Neal Becker wrote: > > I want checkpointing for: > > 1) Protect against job interruption due to system crash, operator error, > power loss, whatever > > 2) Job mygration. Even manual job mygration would be nice. > Several months ago some guys from Ukrain were announcing special patch for kernel which allows you to save task to file, and then later to load it from this file. Saving task to disk say every 5 seconds - is some sort of checkpointing, so later you will be able to go back in time :-) You might want to check l-k archives - I cannot recall correctly the name of this stuff.