From owner-linux-cluster@nl.linux.org Tue Sep 18 16:06:41 2001 Received: (root@humbolt.nl.linux.org) by humbolt.nl.linux.org id ; Tue, 18 Sep 2001 16:06:26 +0200 Received: from jalon.able.es ([212.97.163.2]:4017 "EHLO jalon.able.es") by humbolt.nl.linux.org with ESMTP id ; Tue, 18 Sep 2001 16:06:12 +0200 Received: from werewolf.able.es ([212.97.172.227]) by jalon.able.es (Netscape Messaging Server 4.15) with ESMTP id GJV38K00.U17 for ; Tue, 18 Sep 2001 16:07:32 +0200 Date: Tue, 18 Sep 2001 16:05:58 +0200 From: "J . A . Magallon" To: Lista Linux-Cluster Subject: MPI for bproc Message-ID: <20010918160558.A2117@werewolf.able.es> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit X-Mailer: Balsa 1.2.pre3 Content-Length: 555 Lines: 16 Sender: owner-linux-cluster@nl.linux.org Precedence: bulk Return-Path: X-Envelope-To: <"| /bin/marchive -a -m -f /home/majordomo/public_html/linux-cluster/folders/linux-cluster"> (uid 0) X-Orcpt: rfc822;linux-cluster-list Hi, everyone. I am looking for an MPI version that runs over bproc. Any help ? I have seen nothing specific in MPICH or LAM. BTW, I tried to apply bproc 3.0.1 on a 2.4.9-ac tree. I got several rejects (not many, I thought I would be a disaster). I will try to correct ofssets and so on, but, is anybody doing it yet ? Thanks. -- J.A. Magallon # Let the source be with you... mailto:jamagallon@able.es Mandrake Linux release 8.1 (Cooker) for i586 Linux werewolf 2.4.9-ac11 #1 SMP Mon Sep 17 00:00:56 CEST 2001 i686 Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From owner-linux-cluster@nl.linux.org Tue Sep 18 16:15:37 2001 Received: (root@humbolt.nl.linux.org) by humbolt.nl.linux.org id ; Tue, 18 Sep 2001 16:15:30 +0200 Received: from jalon.able.es ([212.97.163.2]:3762 "EHLO jalon.able.es") by humbolt.nl.linux.org with ESMTP id ; Tue, 18 Sep 2001 16:15:09 +0200 Received: from werewolf.able.es ([212.97.172.227]) by jalon.able.es (Netscape Messaging Server 4.15) with ESMTP id GJV3NE00.E00 for ; Tue, 18 Sep 2001 16:16:26 +0200 Date: Tue, 18 Sep 2001 16:14:52 +0200 From: "J . A . Magallon" To: Lista Linux-Cluster Subject: Hardware for a cluster Message-ID: <20010918161452.C2117@werewolf.able.es> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit X-Mailer: Balsa 1.2.pre3 Content-Length: 978 Lines: 27 Sender: owner-linux-cluster@nl.linux.org Precedence: bulk Return-Path: X-Envelope-To: <"| /bin/marchive -a -m -f /home/majordomo/public_html/linux-cluster/folders/linux-cluster"> (uid 0) X-Orcpt: rfc822;linux-cluster-list Hi everyone. I ask for the expert's opinion. We are going to build a cluster. Some questions: - Dual mobos or single ones ? Some people here say that with dual systems you fill half the space. Is it really an advantage to use dual processor boards ? - Is PentiumIV so much better that a PIII ? Will nowadays gcc (3.0.1) generate code for PIV ? - we are planning to spent not too much money on the network just now, we will see if Ether-100Mb is enough for our communication requierementes (and wait for next year grants...). But I think that an Starfire with 4 ports will be good for internal network, with channel bonding. Or perhaps we can bind NFS to eth0 and rest of comms to eth1, etc. Any known issues ? Thanks in advance for your replies. -- J.A. Magallon # Let the source be with you... mailto:jamagallon@able.es Mandrake Linux release 8.1 (Cooker) for i586 Linux werewolf 2.4.9-ac11 #1 SMP Mon Sep 17 00:00:56 CEST 2001 i686 Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From owner-linux-cluster@nl.linux.org Tue Sep 18 17:26:42 2001 Received: (root@humbolt.nl.linux.org) by humbolt.nl.linux.org id ; Tue, 18 Sep 2001 17:26:36 +0200 Received: from dfw-gate3.raytheon.com ([199.46.199.232]:10159 "EHLO dfw-gate3.raytheon.com") by humbolt.nl.linux.org with ESMTP id ; Tue, 18 Sep 2001 17:26:11 +0200 Received: from ds02c00.directory.ray.com (ds02c00.rsc.raytheon.com [147.25.138.118]) by dfw-gate3.raytheon.com (8.11.0.Beta3/8.11.0.Beta3) with ESMTP id f8IFQ0i09918; Tue, 18 Sep 2001 10:26:06 -0500 (CDT) Received: from rtshou-ds01.hou.us.ray.com (localhost [127.0.0.1]) by ds02c00.directory.ray.com (8.9.3/8.9.3) with ESMTP id KAA17915; Tue, 18 Sep 2001 10:24:33 -0500 (CDT) Subject: Re: Hardware for a cluster To: "J . A . Magallon" Cc: Lista Linux-Cluster X-Mailer: Lotus Notes Release 5.0.4 June 8, 2000 Message-ID: From: Mark_H_Johnson@Raytheon.com Date: Tue, 18 Sep 2001 10:00:22 -0500 X-MIMETrack: Serialize by Router on RTSHOU-DS01/RTS/Raytheon/US(Release 5.0.6a |January 17, 2001) at 09/18/2001 10:26:02 AM MIME-Version: 1.0 Content-type: text/plain; charset=us-ascii Sender: owner-linux-cluster@nl.linux.org Precedence: bulk Return-Path: X-Envelope-To: <"| /bin/marchive -a -m -f /home/majordomo/public_html/linux-cluster/folders/linux-cluster"> (uid 0) X-Orcpt: rfc822;linux-cluster-list It depends (on budget, performance needs, etc.) Last year, we put together a small cluster for evaluation. Five nodes (dual 600 Mhz Pentium III's w/ 256M each), dual 100 base T ethernet w/ switches, rack, serial console connections, etc. It cost about $20,000 US. Best measured performance was about 140 Mbit/sec on the network (2x 70 Mbit). That system worked well for us, but it has been disassembled to support other testing, and provide general network services. This year we are deploying four clusters for our main application (real time simulation of the International Space Station) with... - 12 nodes (dual 1000 Mhz Pentium III's, 512M each, Serverworks HE motherboards) - Scalable Coherent Interface (SCI) - measured >20-200 Mbyte/sec in DMA tests, 4-133 Mbyte/sec in PIO tests - SCI connections to four existing VME racks using PMC adapters - Dual gigabit ethernet from the head node to the LAN - switched 100 Base T on a private LAN to all nodes - Keyboard, video, mouse switch - Racks, power distribution, timer cards, etc. Total cost about $100,000 US per cluster. We are integrating our application right now with delivery to NASA in a few months. Dual CPU support costs about $300 - $500 per node (2nd CPU, better motherboard). That is a lot less than a second machine & allows the two CPU's to share memory and improve performance. When we did equipment selection the Pentium IV was not dual CPU capable and much more expensive than the Pentium III. We also looked at four CPU machines - way too expensive when we looked [$20,000 and up]. Eight CPU and higher machines are even worse. [ymmv] SCI (or Myrinet) is a lot more expensive than Ethernet but the performance is about 10 times higher. SCI is also optimized for "writes" (the creator sends the data to the receiver) - that accounts for the range in performance listed above. SCI also allows general shared memory accesses w/ care in your coding. Your choice here depends on your budget & how much data you must pass around. You may find it cheaper to repeat calculations & send less data to get better performance if you use Ethernet. We were looking at prices about $1500 per node for the connections vs. a per node price of about $2000 (even with dual CPU's). Going with single CPU's would add about $20,000 to the system price and add latency - not something I would recommend. We have had good support from both our cluster vendor (PSSC Labs) and Dolphin (SCI vendor). I would try to stay away from adding too much hardware to any node. On a separate system we had... - 4 port ethernet - Matrox G200/MMS (we use 3 of 4 displays) - 3 Win TV tuners for video input and could not get that equipment to run reliably. Our major problem is IRQ sharing and the BIOS spreading each card (network, display) on all four IRQ's dedicated for the PCI. Bad behaving drivers then caused the problems - generally a system lock up. We split that hardware to two machines & it works OK. --Mark H Johnson "J . A . Magallon" To: Lista Linux-Cluster Sent by: cc: owner-linux-cluster@nl Subject: Hardware for a cluster .linux.org 09/18/01 09:14 AM Hi everyone. I ask for the expert's opinion. We are going to build a cluster. Some questions: - Dual mobos or single ones ? Some people here say that with dual systems you fill half the space. Is it really an advantage to use dual processor boards ? - Is PentiumIV so much better that a PIII ? Will nowadays gcc (3.0.1) generate code for PIV ? - we are planning to spent not too much money on the network just now, we will see if Ether-100Mb is enough for our communication requierementes (and wait for next year grants...). But I think that an Starfire with 4 ports will be good for internal network, with channel bonding. Or perhaps we can bind NFS to eth0 and rest of comms to eth1, etc. Any known issues ? Thanks in advance for your replies. -- J.A. Magallon # Let the source be with you... mailto:jamagallon@able.es Mandrake Linux release 8.1 (Cooker) for i586 Linux werewolf 2.4.9-ac11 #1 SMP Mon Sep 17 00:00:56 CEST 2001 i686 Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From owner-linux-cluster@nl.linux.org Tue Sep 18 17:29:42 2001 Received: (root@humbolt.nl.linux.org) by humbolt.nl.linux.org id ; Tue, 18 Sep 2001 17:29:26 +0200 Received: from woody.fsl.noaa.gov ([137.75.132.225]:64384 "EHLO woody.fsl.noaa.gov") by humbolt.nl.linux.org with ESMTP id ; Tue, 18 Sep 2001 17:29:11 +0200 Received: (from tierney@localhost) by woody.fsl.noaa.gov (8.9.3/8.9.3) id JAA01474; Tue, 18 Sep 2001 09:32:50 -0600 Date: Tue, 18 Sep 2001 09:32:50 -0600 From: Craig Tierney To: "J . A . Magallon" Cc: Lista Linux-Cluster Subject: Re: Hardware for a cluster Message-ID: <20010918093250.A1399@hpti.com> References: <20010918161452.C2117@werewolf.able.es> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.3.6i In-Reply-To: <20010918161452.C2117@werewolf.able.es>; from jamagallon@able.es on Tue, Sep 18, 2001 at 04:14:52PM +0200 Sender: owner-linux-cluster@nl.linux.org Precedence: bulk Return-Path: X-Envelope-To: <"| /bin/marchive -a -m -f /home/majordomo/public_html/linux-cluster/folders/linux-cluster"> (uid 0) X-Orcpt: rfc822;linux-cluster-list On Tue, Sep 18, 2001 at 04:14:52PM +0200, J . A . Magallon wrote: > Hi everyone. > > I ask for the expert's opinion. > > We are going to build a cluster. Some questions: > > - Dual mobos or single ones ? Some people here say that with dual systems > you fill half the space. Is it really an advantage to use dual processor > boards ? It depends on your codes. At the Forecast Systems Lab we chose to go with Dual Alpha nodes for our upgrade. It gave us more bang for the buck when considering the savings of not having to buy twice as many cases, power supplies, cables, racks, and most importantly, myrinet cards and switches. Some of the codes are able to use as much as 90% of the second CPU. However, there are some codes that are only able to use 70% of the second CPU. It really depends on your codes. > > - Is PentiumIV so much better that a PIII ? Will nowadays gcc (3.0.1) generate > code for PIV ? Look into the Intel Compilers for Linux (assuming that is the OS you want). If you only have to by copies where you compile it isn't that expensive ~ $500US per copy. Craig Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From owner-linux-cluster@nl.linux.org Tue Sep 18 18:36:25 2001 Received: (root@humbolt.nl.linux.org) by humbolt.nl.linux.org id ; Tue, 18 Sep 2001 18:36:09 +0200 Received: from 85-MADR-X33.libre.retevision.es ([62.83.4.85]:61961 "EHLO mioooldpc") by humbolt.nl.linux.org with ESMTP id ; Tue, 18 Sep 2001 18:35:53 +0200 Received: from mioooldpc (unknown [192.168.0.2]) by mioooldpc (Postfix) with SMTP id F18A02F43B; Tue, 18 Sep 2001 18:46:00 +0200 (CEST) Content-Type: text/plain; charset="iso-8859-1" From: jordi To: "J . A . Magallon" Subject: Re: Hardware for a cluster Date: Tue, 18 Sep 2001 18:45:59 +0200 X-Mailer: KMail [version 1.2] References: <20010918161452.C2117@werewolf.able.es> In-Reply-To: <20010918161452.C2117@werewolf.able.es> Cc: MIME-Version: 1.0 Message-Id: <01091818455901.03731@mioooldpc> Content-Transfer-Encoding: 8bit Sender: owner-linux-cluster@nl.linux.org Precedence: bulk Return-Path: X-Envelope-To: <"| /bin/marchive -a -m -f /home/majordomo/public_html/linux-cluster/folders/linux-cluster"> (uid 0) X-Orcpt: rfc822;linux-cluster-list > - Dual mobos or single ones ? Some people here say that with dual systems > you fill half the space. Is it really an advantage to use dual processor > boards ? It depends in the apps but surely if you are going to build a cluster your apps will take real advantage of this. Even so the second processor won't just double the performance so it's a matter of money. > - Is PentiumIV so much better that a PIII ? Will nowadays gcc (3.0.1) > generate code for PIV ? No it doesn't. they are working in it. You can pay the Intel's Linux compiler. If you won't generate special code for PIV , PIV don't get very good results and any other option will be better. It's a expensive option also. Anyway if what you want is FPU power then a Athlon it's much better , you don't need to recompiler and rocks. If you really are low in money duron is the option. I don't know if PIV dual exist. > - we are planning to spent not too much money on the network just now, we > will see if Ether-100Mb is enough for our communication requierementes (and > wait for next year grants...). But I think > that an Starfire with 4 ports will be good for internal network, with > channel bonding. Or perhaps we can bind NFS to eth0 and rest of comms to > eth1, etc. Any known issues ? If you don't want to spend money i think that approach is the best. Just use 1 eth 100 network (really cheap) . You'll surely will upgrade soon (adding other network). Depending in what aplications are running in the cluster you can split them between your eth networks in a intelligent way. It mainly depends on the apps. Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From owner-linux-cluster@nl.linux.org Tue Sep 18 20:39:17 2001 Received: (root@humbolt.nl.linux.org) by humbolt.nl.linux.org id ; Tue, 18 Sep 2001 20:39:01 +0200 Received: from plogic.com ([209.92.41.135]:25301 "EHLO otto.plogic.internal") by humbolt.nl.linux.org with ESMTP id ; Tue, 18 Sep 2001 20:38:53 +0200 Received: from rlatham by otto.plogic.internal with local (Exim 3.22 #1 (Debian)) id 15jPla-00073h-00 for ; Tue, 18 Sep 2001 14:38:42 -0400 Date: Tue, 18 Sep 2001 14:38:42 -0400 From: Rob Latham To: Lista Linux-Cluster Subject: Re: MPI for bproc Message-ID: <20010918143842.A26971@otto.plogic.internal> References: <20010918160558.A2117@werewolf.able.es> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <20010918160558.A2117@werewolf.able.es>; from jamagallon@able.es on Tue, Sep 18, 2001 at 04:05:58PM +0200 Sender: owner-linux-cluster@nl.linux.org Precedence: bulk Return-Path: X-Envelope-To: <"| /bin/marchive -a -m -f /home/majordomo/public_html/linux-cluster/folders/linux-cluster"> (uid 0) X-Orcpt: rfc822;linux-cluster-list On Tue, Sep 18, 2001 at 04:05:58PM +0200, J . A . Magallon wrote: > Hi, everyone. > > I am looking for an MPI version that runs over bproc. Any help ? > I have seen nothing specific in MPICH or LAM. the newly released mpich-1.2.2 or scyld's beompi (basically mpich-1.2.1 + all the bproc patches that scyld forwarded on to mpich ) > BTW, I tried to apply bproc 3.0.1 on a 2.4.9-ac tree. I got several > rejects (not many, I thought I would be a disaster). I will try to > correct ofssets and so on, but, is anybody doing it yet ? direct that question to bproc-users@lists.sourceforge.net ==rob -- [ Rob Latham Developer, Admin, Alchemist ] [ Paralogic Inc. - www.plogic.com ] [ ] [ EAE8 DE90 85BB 526F 3181 1FCF 51C4 B6CB 08CC 0897 ] Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From owner-linux-cluster@nl.linux.org Tue Sep 18 20:45:52 2001 Received: (root@humbolt.nl.linux.org) by humbolt.nl.linux.org id ; Tue, 18 Sep 2001 20:45:33 +0200 Received: from plogic.com ([209.92.41.135]:32981 "EHLO otto.plogic.internal") by humbolt.nl.linux.org with ESMTP id ; Tue, 18 Sep 2001 20:45:10 +0200 Received: from rlatham by otto.plogic.internal with local (Exim 3.22 #1 (Debian)) id 15jPrk-000751-00 for ; Tue, 18 Sep 2001 14:45:04 -0400 Date: Tue, 18 Sep 2001 14:45:04 -0400 From: Rob Latham To: linux-cluster@nl.linux.org Subject: Re: Hardware for a cluster Message-ID: <20010918144504.B26971@otto.plogic.internal> References: <20010918161452.C2117@werewolf.able.es> <01091818455901.03731@mioooldpc> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <01091818455901.03731@mioooldpc>; from mumismo@wanadoo.es on Tue, Sep 18, 2001 at 06:45:59PM +0200 Sender: owner-linux-cluster@nl.linux.org Precedence: bulk Return-Path: X-Envelope-To: <"| /bin/marchive -a -m -f /home/majordomo/public_html/linux-cluster/folders/linux-cluster"> (uid 0) X-Orcpt: rfc822;linux-cluster-list On Tue, Sep 18, 2001 at 06:45:59PM +0200, jordi wrote: > I don't know if PIV dual exist. they do. for one example ( i don't mean this as an endorsement or plug, just one option out of several (?) ) http://www.supermicro.com/PRODUCT/MotherBoards/860/SUPER_P4DC6_SPEC.htm ==rob -- [ Rob Latham Developer, Admin, Alchemist ] [ Paralogic Inc. - www.plogic.com ] [ ] [ EAE8 DE90 85BB 526F 3181 1FCF 51C4 B6CB 08CC 0897 ] Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From owner-linux-cluster@nl.linux.org Tue Sep 18 22:27:21 2001 Received: (root@humbolt.nl.linux.org) by humbolt.nl.linux.org id ; Tue, 18 Sep 2001 22:27:06 +0200 Received: from kotako.analogself.com ([207.181.249.20]:14086 "EHLO kotako.analogself.com") by humbolt.nl.linux.org with ESMTP id ; Tue, 18 Sep 2001 22:26:54 +0200 Received: by kotako.analogself.com (Postfix, from userid 1502) id 7FFADE801A; Tue, 18 Sep 2001 13:22:44 -0700 (PDT) Date: Tue, 18 Sep 2001 13:22:44 -0700 From: Jag To: Rob Latham Cc: Lista Linux-Cluster Subject: Re: MPI for bproc Message-ID: <20010918132244.A17753@kotako.analogself.com> Mail-Followup-To: Rob Latham , Lista Linux-Cluster References: <20010918160558.A2117@werewolf.able.es> <20010918143842.A26971@otto.plogic.internal> Mime-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-md5; protocol="application/pgp-signature"; boundary="envbJBWh7q8WU6mo" Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <20010918143842.A26971@otto.plogic.internal>; from rlatham@plogic.com on Tue, Sep 18, 2001 at 02:38:42PM -0400 Sender: owner-linux-cluster@nl.linux.org Precedence: bulk Return-Path: X-Envelope-To: <"| /bin/marchive -a -m -f /home/majordomo/public_html/linux-cluster/folders/linux-cluster"> (uid 0) X-Orcpt: rfc822;linux-cluster-list --envbJBWh7q8WU6mo Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Tue, 18 Sep 2001, Rob Latham wrote: > On Tue, Sep 18, 2001 at 04:05:58PM +0200, J . A . Magallon wrote: > > Hi, everyone. > >=20 > > I am looking for an MPI version that runs over bproc. Any help ? > > I have seen nothing specific in MPICH or LAM. >=20 > the newly released mpich-1.2.2 or scyld's beompi (basically > mpich-1.2.1 + all the bproc patches that scyld forwarded on to mpich ) beompi is something of the past.. the package is just mpich now --envbJBWh7q8WU6mo Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.6 (GNU/Linux) Comment: For info see http://www.gnupg.org iD8DBQE7p60U+pq97aGGtXARAmOpAKCfpz2gRKbLTXEBBXncYcRm5tMqkACgwMnO aly/c9hxpLy6YptHG/vWwiM= =3cpl -----END PGP SIGNATURE----- --envbJBWh7q8WU6mo-- Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From owner-linux-cluster@nl.linux.org Wed Sep 19 00:32:06 2001 Received: (root@humbolt.nl.linux.org) by humbolt.nl.linux.org id ; Wed, 19 Sep 2001 00:31:48 +0200 Received: from jalon.able.es ([212.97.163.2]:21704 "EHLO jalon.able.es") by humbolt.nl.linux.org with ESMTP id ; Wed, 19 Sep 2001 00:31:28 +0200 Received: from werewolf.able.es ([212.97.172.227]) by jalon.able.es (Netscape Messaging Server 4.15) with ESMTP id GJVQMY00.M0V; Wed, 19 Sep 2001 00:32:58 +0200 Date: Wed, 19 Sep 2001 00:31:23 +0200 From: "J . A . Magallon" To: Jag Cc: Rob Latham , Lista Linux-Cluster Subject: Re: MPI for bproc Message-ID: <20010919003123.A4847@werewolf.able.es> References: <20010918160558.A2117@werewolf.able.es> <20010918143842.A26971@otto.plogic.internal> <20010918132244.A17753@kotako.analogself.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit In-Reply-To: <20010918132244.A17753@kotako.analogself.com>; from agrajag@linuxpower.org on Tue, Sep 18, 2001 at 22:22:44 +0200 X-Mailer: Balsa 1.2.pre3 Content-Length: 922 Lines: 25 Sender: owner-linux-cluster@nl.linux.org Precedence: bulk Return-Path: X-Envelope-To: <"| /bin/marchive -a -m -f /home/majordomo/public_html/linux-cluster/folders/linux-cluster"> (uid 0) X-Orcpt: rfc822;linux-cluster-list On 20010918 Jag wrote: >On Tue, 18 Sep 2001, Rob Latham wrote: > >> On Tue, Sep 18, 2001 at 04:05:58PM +0200, J . A . Magallon wrote: >> > Hi, everyone. >> > >> > I am looking for an MPI version that runs over bproc. Any help ? >> > I have seen nothing specific in MPICH or LAM. >> >> the newly released mpich-1.2.2 or scyld's beompi (basically >> mpich-1.2.1 + all the bproc patches that scyld forwarded on to mpich ) > >beompi is something of the past.. the package is just mpich now > Yes, I downloaded mpich and it is verision 1.2.2.1, with bproc support. Anyways, Scylds tools look like based on kernel 2.2.19-bproc2.2, that is now unmantained. As I have read, bproc 3.0.1 is 2.4.x only. -- J.A. Magallon # Let the source be with you... mailto:jamagallon@able.es Mandrake Linux release 8.1 (Cooker) for i586 Linux werewolf 2.4.9-ac11 #1 SMP Mon Sep 17 00:00:56 CEST 2001 i686 Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From owner-linux-cluster@nl.linux.org Wed Sep 19 00:44:40 2001 Received: (root@humbolt.nl.linux.org) by humbolt.nl.linux.org id ; Wed, 19 Sep 2001 00:44:32 +0200 Received: from plogic.com ([209.92.41.135]:30936 "EHLO otto.plogic.internal") by humbolt.nl.linux.org with ESMTP id ; Wed, 19 Sep 2001 00:44:22 +0200 Received: from rlatham by otto.plogic.internal with local (Exim 3.22 #1 (Debian)) id 15jTbI-0007aa-00 for ; Tue, 18 Sep 2001 18:44:20 -0400 Date: Tue, 18 Sep 2001 18:44:20 -0400 From: Rob Latham To: Lista Linux-Cluster Subject: Re: MPI for bproc Message-ID: <20010918184420.A29108@otto.plogic.internal> References: <20010918160558.A2117@werewolf.able.es> <20010918143842.A26971@otto.plogic.internal> <20010918132244.A17753@kotako.analogself.com> <20010919003123.A4847@werewolf.able.es> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <20010919003123.A4847@werewolf.able.es>; from jamagallon@able.es on Wed, Sep 19, 2001 at 12:31:23AM +0200 Sender: owner-linux-cluster@nl.linux.org Precedence: bulk Return-Path: X-Envelope-To: <"| /bin/marchive -a -m -f /home/majordomo/public_html/linux-cluster/folders/linux-cluster"> (uid 0) X-Orcpt: rfc822;linux-cluster-list On Wed, Sep 19, 2001 at 12:31:23AM +0200, J . A . Magallon wrote: > > Yes, I downloaded mpich and it is verision 1.2.2.1, with bproc support. > Anyways, Scylds tools look like based on kernel 2.2.19-bproc2.2, that > is now unmantained. As I have read, bproc 3.0.1 is 2.4.x only. This thread is heading pretty off topic ( though i'm not sure exactly what the topic of this list is :> ), but try running scyld's mpich with bproc-3.0.1. as far as mpich is concerned, the differences should be small enough that you have a pretty good chance of getting away with it. ==rob -- [ Rob Latham Developer, Admin, Alchemist ] [ Paralogic Inc. - www.plogic.com ] [ ] [ EAE8 DE90 85BB 526F 3181 1FCF 51C4 B6CB 08CC 0897 ] Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From owner-linux-cluster@nl.linux.org Wed Sep 19 03:51:03 2001 Received: (root@humbolt.nl.linux.org) by humbolt.nl.linux.org id ; Wed, 19 Sep 2001 03:50:49 +0200 Received: from bos-gate3.raytheon.com ([199.46.198.232]:48874 "EHLO bos-gate3.raytheon.com") by humbolt.nl.linux.org with ESMTP id ; Wed, 19 Sep 2001 03:50:27 +0200 Received: from ds02w00.directory.ray.com (ds02w00.rsc.raytheon.com [147.25.146.118]) by bos-gate3.raytheon.com (8.11.0.Beta3/8.11.0.Beta3) with ESMTP id f8J1oPS15779 for ; Tue, 18 Sep 2001 21:50:26 -0400 (EDT) Received: from es-e01-mta1.rsc.raytheon.com (localhost [127.0.0.1]) by ds02w00.directory.ray.com (8.9.3/8.9.3) with ESMTP id SAA27950 for ; Tue, 18 Sep 2001 18:50:24 -0700 (PDT) Subject: Re: Hardware for a cluster To: linux-cluster@nl.linux.org X-Mailer: Lotus Notes Release 5.0.5 September 22, 2000 Message-ID: From: "Samuel R Warren" Date: Tue, 18 Sep 2001 18:48:31 -0700 X-MIMETrack: Serialize by Router on RWSMTA10/SRV/Raytheon/US(Release 5.0.8 |June 18, 2001) at 09/18/2001 06:48:34 PM MIME-Version: 1.0 Content-type: text/plain; charset=us-ascii Sender: owner-linux-cluster@nl.linux.org Precedence: bulk Return-Path: X-Envelope-To: <"| /bin/marchive -a -m -f /home/majordomo/public_html/linux-cluster/folders/linux-cluster"> (uid 0) X-Orcpt: rfc822;linux-cluster-list > On Tue, Sep 18, 2001 at 06:45:59PM +0200, jordi wrote: > > > I don't know if PIV dual exist. > > they do. for one example ( i don't mean this as an endorsement or > plug, just one option out of several (?) ) > http://www.supermicro.com/PRODUCT/MotherBoards/860/SUPER_P4DC6_SPEC.htm Here are a couple of name brand systems http://www.compaq.com/products/workstations/w6000/index.shtml http://www.dell.com/us/en/biz/products/model_precn_3_precn_530.htm Dell has put some impressive numbers at the SPEC website. http://www.spec.org/osg/cpu2000/results/res2001q2/cpu2000-20010507-00615.html http://www.spec.org/osg/cpu2000/results/res2001q2/cpu2000-20010507-00614.html http://www.spec.org/osg/cpu2000/results/res2001q2/cpu2000-20010507-00613.html http://www.spec.org/osg/cpu2000/results/res2001q2/cpu2000-20010507-00612.html Rob Latham @nl.linux.org on 09/18/2001 11:45:04 AM Sent by: owner-linux-cluster@nl.linux.org To: linux-cluster@nl.linux.org cc: Subject: Re: Hardware for a cluster On Tue, Sep 18, 2001 at 06:45:59PM +0200, jordi wrote: > I don't know if PIV dual exist. they do. for one example ( i don't mean this as an endorsement or plug, just one option out of several (?) ) http://www.supermicro.com/PRODUCT/MotherBoards/860/SUPER_P4DC6_SPEC.htm ==rob -- [ Rob Latham Developer, Admin, Alchemist ] [ Paralogic Inc. - www.plogic.com ] [ ] [ EAE8 DE90 85BB 526F 3181 1FCF 51C4 B6CB 08CC 0897 ] Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From owner-linux-cluster@nl.linux.org Wed Sep 26 16:25:06 2001 Received: (root@humbolt.nl.linux.org) by humbolt.nl.linux.org id ; Wed, 26 Sep 2001 16:24:47 +0200 Received: from gate.in-addr.de ([212.8.193.158]:12557 "EHLO mx.in-addr.de") by humbolt.nl.linux.org with ESMTP id ; Wed, 26 Sep 2001 16:24:28 +0200 Received: by mx.in-addr.de (mail.in-addr.de, from userid 10) id ACA514ADE8; Wed, 26 Sep 2001 16:23:57 +0200 (CEST) Received: by hermes.marowsky-bree.de (Postfix, from userid 500) id 531B0B829; Wed, 26 Sep 2001 16:24:42 +0200 (CEST) Date: Wed, 26 Sep 2001 16:24:42 +0200 From: Lars Marowsky-Bree To: linux-cluster@nl.linux.org, linux-ha@muc.de Subject: CFP: Linux Kongress 2001 Clustering Workshop Message-ID: <20010926162442.C4100@marowsky-bree.de> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit User-Agent: Mutt/1.3.16i X-Ctuhulu: HASTUR Sender: owner-linux-cluster@nl.linux.org Precedence: bulk Return-Path: X-Envelope-To: <"| /bin/marchive -a -m -f /home/majordomo/public_html/linux-cluster/folders/linux-cluster"> (uid 0) X-Orcpt: rfc822;linux-cluster-list [ Please pass this on to interested parties ] CALL FOR PARTICIPATION Linux Kongress 2001 Clustering Workshop November 26 - 28 2001 Enschede, NL We are pleased to announce a working group meeting for all clustering related topics on Linux, taking place as a part of and directly prior to the Linux Kongress 2001. The purpose of this meeting is to form a permanent working group on an Open Clustering Framework, define its charter and outline the roadmap for future work. An Open Clustering Framework is urgently needed to increase compatibility of projects, allow for applications to be portable across different implementations/installations and should ideally become part of the Linux Standards Base at some future date. The agenda for this meeting is: - Overview of proposed working group - Discussion of its charter - Presentation of current efforts - Discussion of initial roadmap - Work on roadmap items Interesting papers concerning current clustering efforts are invited, to allow for an overview of currently available solutions and open issues. Topics of interest from this point of view include, but are not limited to: - All aspects of High Availability Clustering - All aspects of High Performance Clustering - Specialised filesystems - Distributed Lock Managers - Kernel modifications for clustering - Status reports of existing projects - Research reports and Work In Progress reports While the focus is on open efforts, commercial vendors are explicitly solicited to present their technical offering and work with us on the Open Clustering Framework and other aspects of Linux clustering. The cooperation between commercial vendors and the community is vital on arriving at a viable solution from which all parties would greatly benefit. The intended audience of this working group meeting are clustering system architects and developers, cluster application developers, and interested users, whose input will be particularly welcome. On Monday and Tuesday, we expect to have a single track of presentations, 4-6 each day, each lasting 45-90 minutes. We will also have plenty of time for spontaneous BOF sessions. On Wednesday morning, we will conclude the meeting with a panel session on: "Clustering Efforts for Linux - Present and Future" After this, the main Linux Kongress 2001 will start and you are of course invited to stay or present a paper in the main tracks. If you are interested in presenting your work or participating in the working group meeting, please contact Lars Marowsky-Bree by Friday, 2001-11-02. Please send a short (<400 words) abstract of your topic, whether you would like to make it a presentation, working group meeting or tutorial. We are also pleased that we have a - albeit limited - budget for travel expenses. Please understand that travel expenses can only be paid for in few cases as our budget is rather limited, and that we would like to ask companies to pay for their employees. However, if you are an Open Source developer, student or working on an interesting project and have difficulty making the trip, please contact us also to see whether we can work something out. Sincerely, Lars Marowsky-Brée (Working Group Organizer) Alan Roberston (Linux HA Project) Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/