From mboxrd@z Thu Jan 1 00:00:00 1970 From: ebiederm@xmission.com (Eric W. Biederman) Subject: Re: [RFC] network namespaces Date: Wed, 06 Sep 2006 14:40:13 -0600 Message-ID: References: <20060815182029.A1685@castle.nmd.msu.ru> <20060816115313.GC31810@sergelap.austin.ibm.com> <44FD7CF0.4030009@fr.ibm.com> <44FF2EA6.2030303@fr.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Daniel Lezcano , netdev@vger.kernel.org, "Serge E. Hallyn" , Andrey Savochkin , haveblue@us.ibm.com, herbert@13thfloor.at, sam@vilain.net, Andrew Morton , dev@sw.ru, devel@openvz.org, alexey@sw.ru, Linux Containers Return-path: Received: from ebiederm.dsl.xmission.com ([166.70.28.69]:30688 "EHLO ebiederm.dsl.xmission.com") by vger.kernel.org with ESMTP id S1751188AbWIFUl2 (ORCPT ); Wed, 6 Sep 2006 16:41:28 -0400 To: Cedric Le Goater In-Reply-To: <44FF2EA6.2030303@fr.ibm.com> (Cedric Le Goater's message of "Wed, 06 Sep 2006 22:25:10 +0200") Sender: netdev-owner@vger.kernel.org List-Id: netdev.vger.kernel.org Cedric Le Goater writes: > Eric W. Biederman wrote: > > hmm ? What about an MPI application ? > > I would expect each MPI task to be run in its container on different nodes > or on the same node. These individual tasks _communicate_ between each > other through the MPI layer (not only TCP btw) to complete a large calculation. All parts of the MPI application are part of the same job. Communication between processes on multiple machines that are part of the job is fine. At least that is how I used job in HPC computer context. Eric