Untitled Document

From linux-cluster-bounce@nl.linux.org Sun Mar 3 02:08:53 2002 Received: from localhost.nl.linux.org ([IPv6:::ffff:127.0.0.1]:35820 "EHLO humbolt.") by humbolt.nl.linux.org with ESMTP id ; Sun, 3 Mar 2002 02:08:40 +0100 Received: with LISTAR (v1.0.0; list linux-cluster); Sun, 03 Mar 2002 02:08:09 +0100 (CET) Received: from [IPv6:::ffff:211.248.201.131] ([IPv6:::ffff:211.248.201.131]:59287 "HELO 211.248.201.131") by humbolt.nl.linux.org with SMTP id ; Sun, 3 Mar 2002 02:07:48 +0100 From: cauliz To: linux-cluster@nl.linux.org Subject: Just thinking about you. dgnua Mime-Version: 1.0 Content-Type: text/html; charset="iso-8859-1" Date: Sat, 2 Mar 2002 17:03:55 -0800 Message-Id: <20020303010755Z16706-24254+206@humbolt.nl.linux.org> X-listar-version: Listar v1.0.0 Sender: linux-cluster-bounce@nl.linux.org Errors-to: linux-cluster-bounce@nl.linux.org X-original-sender: mhliz@sky.net Precedence: bulk List-help: List-unsubscribe: List-software: Listar version 1.0.0 X-List-ID: List-subscribe: List-owner: List-post: List-archive: X-list: linux-cluster 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-archive@nl.linux.org Original-Recipient: rfc822;linux-cluster-archive@nl.linux.org Untitled Document

If you thought sex was all over for you, take heart and read on! There's every chance that very, very soon you'll actually be watching your penis stand up and salute with all the vigor, zest, and lusty appetite of a fresh, young recruit! The secret of success will be all yours - she need ever know - and this is how it works...

For more information please send an email to revolutionary_pills@yahoo.com with "MORE INFO" being the only two words on the subject line. To remove, place "REMOVE" being the only word on the subject line.

aqtstkqgwmgnljqyvkexikiqvetun - Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From linux-cluster-bounce@nl.linux.org Thu Mar 28 01:58:02 2002 Received: from localhost.nl.linux.org ([IPv6:::ffff:127.0.0.1]:24789 "EHLO humbolt.") by humbolt.nl.linux.org with ESMTP id ; Thu, 28 Mar 2002 01:57:47 +0100 Received: with LISTAR (v1.0.0; list linux-cluster); Thu, 28 Mar 2002 01:56:49 +0100 (CET) Received: from [IPv6:::ffff:206.251.243.130] ([IPv6:::ffff:206.251.243.130]:56331 "EHLO mailrelay") by humbolt.nl.linux.org with ESMTP id ; Thu, 28 Mar 2002 01:53:40 +0100 Received: from mail pickup service by mailrelay with Microsoft SMTPSVC; Wed, 27 Mar 2002 16:48:28 -0800 Received: from mail.signalpharm.com ([172.16.55.33]) by mailrelay with Microsoft SMTPSVC(5.0.2195.3779); Wed, 27 Mar 2002 16:48:28 -0800 Content-Transfer-Encoding: 7bit Received: by MAIL with Internet Mail Service (5.5.2653.19) id ; Wed, 27 Mar 2002 16:56:15 -0800 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4522.1200 Message-ID: <13A1ABE4769DD31192AC009027DE945DD7E4EB@MAIL> from: "Tracy Zhao" to: "linux-cluster@nl.linux.org" Subject: very basic cluster question Date: Wed, 27 Mar 2002 16:56:04 -0800 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain; charset="iso-8859-1" X-OriginalArrivalTime: 28 Mar 2002 00:48:28.0398 (UTC) FILETIME=[467550E0:01C1D5F2] X-Approved-By: riel@nl.linux.org X-listar-version: Listar v1.0.0 Sender: linux-cluster-bounce@nl.linux.org Errors-to: linux-cluster-bounce@nl.linux.org X-original-sender: TZhao@CELGENE.com Precedence: bulk List-help: List-unsubscribe: List-software: Listar version 1.0.0 X-List-ID: List-subscribe: List-owner: List-post: List-archive: X-list: linux-cluster 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-archive@nl.linux.org Original-Recipient: rfc822;linux-cluster-archive@nl.linux.org Hi, I have a very basic cluster question and I would be very amazed if anyone responded to my question becuase it's really beyond basic. Here is my question: I am trying to build a cluster and I've been trying to decide between building a Scyld Beowulf cluster and a Rocks cluster. I have been trying to find documentation everywhere for the differences between the two but I haven't been very sucessful. I've heard that the Scyld Beowolf cluster would give you a Single System Image which is cleaner than the set up of a Rocks cluster. I've even heard Rocks clusters being comparable to having a bunch of computers on a network and just writing code to delegate jobs to each machine. Could you tell me where I can find some basic information as to the differences between these clusters and how it would be beneficial to have a Rocks cluster vs. a bunch of machines on a network? And how about having a database running on a cluster? Sorry about these basic questions but I'm really lost as to where I can find this type of basic, before you start building a cluster type of information. Thanks, Tracy Zhao - Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From linux-cluster-bounce@nl.linux.org Thu Mar 28 02:33:23 2002 Received: from localhost.nl.linux.org ([IPv6:::ffff:127.0.0.1]:17632 "EHLO humbolt.") by humbolt.nl.linux.org with ESMTP id ; Thu, 28 Mar 2002 02:33:13 +0100 Received: with LISTAR (v1.0.0; list linux-cluster); Thu, 28 Mar 2002 02:32:48 +0100 (CET) Received: from nixon.xkey.com ([IPv6:::ffff:209.245.148.124]:34797 "HELO nixon.xkey.com") by humbolt.nl.linux.org with SMTP id ; Thu, 28 Mar 2002 02:32:40 +0100 Received: (qmail 5387 invoked from network); 28 Mar 2002 01:32:36 -0000 Received: from localhost (HELO localhost.conservativecomputer.com) (127.0.0.1) by localhost with SMTP; 28 Mar 2002 01:32:36 -0000 Received: (from lindahl@localhost) by localhost.conservativecomputer.com (8.11.6/8.11.0) id g2S1XQQ02852 for linux-cluster@nl.linux.org; Wed, 27 Mar 2002 20:33:26 -0500 X-Authentication-Warning: localhost.localdomain: lindahl set sender to lindahl@conservativecomputer.com using -f Date: Wed, 27 Mar 2002 20:33:25 -0500 From: Greg Lindahl To: "linux-cluster@nl.linux.org" Subject: Re: very basic cluster question Message-ID: <20020327203325.A2825@wumpus.aonnetworks.com> Mail-Followup-To: "linux-cluster@nl.linux.org" References: <13A1ABE4769DD31192AC009027DE945DD7E4EB@MAIL> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <13A1ABE4769DD31192AC009027DE945DD7E4EB@MAIL>; from TZhao@CELGENE.com on Wed, Mar 27, 2002 at 04:56:04PM -0800 X-listar-version: Listar v1.0.0 Sender: linux-cluster-bounce@nl.linux.org Errors-to: linux-cluster-bounce@nl.linux.org X-original-sender: lindahl@conservativecomputer.com Precedence: bulk List-help: List-unsubscribe: List-software: Listar version 1.0.0 X-List-ID: List-subscribe: List-owner: List-post: List-archive: X-list: linux-cluster 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-archive@nl.linux.org Original-Recipient: rfc822;linux-cluster-archive@nl.linux.org On Wed, Mar 27, 2002 at 04:56:04PM -0800, Tracy Zhao wrote: > I am trying to build a cluster and I've been trying to decide > between building a Scyld Beowulf cluster and a Rocks cluster. Tracy, You don't say what you want to do on your cluster. This is a generic cluster list, there is also a cluster list specific to Beowulf clusters. The main difference between Scyld and Rocks is administration and installation. Scyld is only a little more complicated than a single system to install, even for a fairly large cluster. Rocks uses a complete system disk on every node, and while it has a fairly automated install sequence, it isn't as simple as Scyld. The main question I'd have, though, is why you're "rolling your own" cluster instead of buying from a cluster vendor. Since you work for a business, I'd hope that one of the many Linux cluster vendors could meet your needs and save you time. > And how about having a database running on a cluster? It's easy to run a single database on a single node in a cluster, but parallel databases are far more complicated. In the genomics world, people usually run things like BLAST on their clusters, not their databases. greg - Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From linux-cluster-bounce@nl.linux.org Fri Mar 29 00:05:56 2002 Received: from localhost.nl.linux.org ([IPv6:::ffff:127.0.0.1]:55221 "EHLO humbolt.") by humbolt.nl.linux.org with ESMTP id ; Fri, 29 Mar 2002 00:05:52 +0100 Received: with LISTAR (v1.0.0; list linux-cluster); Fri, 29 Mar 2002 00:05:18 +0100 (CET) Received: from mail315.mail.bellsouth.net ([IPv6:::ffff:205.152.58.175]:58373 "EHLO imf15bis.bellsouth.net") by humbolt.nl.linux.org with ESMTP id convert rfc822-to-8bit; Fri, 29 Mar 2002 00:05:12 +0100 Received: from taz ([65.81.168.161]) by imf15bis.bellsouth.net (InterMail vM.5.01.04.05 201-253-122-122-105-20011231) with SMTP id <20020328230630.WQWS1237.imf15bis.bellsouth.net@taz>; Thu, 28 Mar 2002 18:06:30 -0500 Date: Thu, 28 Mar 2002 17:58:17 -0500 From: Greg Freemyer Subject: re: very basic cluster question To: Tracy Zhao , linux-cluster@nl.linux.org Mime-Version: 1.0 Organization: The NorcrossGroup X-Mailer: GoldMine [5.70.11111] Content-Type: Text/plain Content-Transfer-Encoding: 8BIT Message-Id: <20020328230630.WQWS1237.imf15bis.bellsouth.net@taz> X-listar-version: Listar v1.0.0 Sender: linux-cluster-bounce@nl.linux.org Errors-to: linux-cluster-bounce@nl.linux.org X-original-sender: freemyer@NorcrossGroup.com Precedence: bulk List-help: List-unsubscribe: List-software: Listar version 1.0.0 X-List-ID: List-subscribe: List-owner: List-post: List-archive: X-list: linux-cluster 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-archive@nl.linux.org Original-Recipient: rfc822;linux-cluster-archive@nl.linux.org >> And how about having a database running on a cluster? If you want a fail-over cluster, then many of the cluster technologies can handle this. If you want active-active clustering, then every Cluster aware DB I know is commercial. Oracle 9i RAC is the best known. I think they have a Linux version, but I don't know the cost. Greg Freemyer Internet Engineer Deployment and Integration Specialist The Norcross Group www.NorcrossGroup.com - Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From linux-cluster-bounce@nl.linux.org Sat Mar 30 06:08:49 2002 Received: from localhost.nl.linux.org ([IPv6:::ffff:127.0.0.1]:44689 "EHLO humbolt.") by humbolt.nl.linux.org with ESMTP id ; Sat, 30 Mar 2002 06:08:41 +0100 Received: with LISTAR (v1.0.0; list linux-cluster); Sat, 30 Mar 2002 06:07:50 +0100 (CET) Received: from web20009.mail.yahoo.com ([IPv6:::ffff:216.136.225.72]:33811 "HELO web20009.mail.yahoo.com") by humbolt.nl.linux.org with SMTP id ; Sat, 30 Mar 2002 06:07:41 +0100 Message-ID: <20020330050739.36370.qmail@web20009.mail.yahoo.com> Received: from [24.185.124.181] by web20009.mail.yahoo.com via HTTP; Fri, 29 Mar 2002 21:07:39 PST Date: Fri, 29 Mar 2002 21:07:39 -0800 (PST) From: z l Subject: cluster To: linux-cluster@nl.linux.org MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-listar-version: Listar v1.0.0 Sender: linux-cluster-bounce@nl.linux.org Errors-to: linux-cluster-bounce@nl.linux.org X-original-sender: z_ling11@yahoo.com Precedence: bulk List-help: List-unsubscribe: List-software: Listar version 1.0.0 X-List-ID: List-subscribe: List-owner: List-post: List-archive: X-list: linux-cluster 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-archive@nl.linux.org Original-Recipient: rfc822;linux-cluster-archive@nl.linux.org first i want to know if this list is active or not. if yes then please help answer some basic questions. HA module or load balancing solutions can also be called as cluster ? what exactly does a cluster denote. because i have implemented HA solutions like two servers, one being the hot standby for the other. can we call this a cluster ? can someone give various configurations of clusters and clustering that is being done today ? i know of fileservers, web servers and firewalls in HA or loadbalancing mode. would that suffice to be called as cluster ?? __________________________________________________ Do You Yahoo!? Yahoo! Greetings - send holiday greetings for Easter, Passover http://greetings.yahoo.com/ - Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From linux-cluster-bounce@nl.linux.org Sat Mar 30 17:58:17 2002 Received: from localhost.nl.linux.org ([IPv6:::ffff:127.0.0.1]:454 "EHLO humbolt.") by humbolt.nl.linux.org with ESMTP id ; Sat, 30 Mar 2002 17:58:11 +0100 Received: with LISTAR (v1.0.0; list linux-cluster); Sat, 30 Mar 2002 17:57:38 +0100 (CET) Received: from nixon.xkey.com ([IPv6:::ffff:209.245.148.124]:144 "HELO nixon.xkey.com") by humbolt.nl.linux.org with SMTP id ; Sat, 30 Mar 2002 17:57:26 +0100 Received: (qmail 23544 invoked from network); 30 Mar 2002 16:57:23 -0000 Received: from localhost (HELO localhost.conservativecomputer.com) (127.0.0.1) by localhost with SMTP; 30 Mar 2002 16:57:23 -0000 Received: (from lindahl@localhost) by localhost.conservativecomputer.com (8.11.6/8.11.0) id g2UGwLi02036 for linux-cluster@nl.linux.org; Sat, 30 Mar 2002 11:58:21 -0500 X-Authentication-Warning: localhost.localdomain: lindahl set sender to lindahl@keyresearch.com using -f Date: Sat, 30 Mar 2002 11:58:21 -0500 From: Greg Lindahl To: linux-cluster@nl.linux.org Subject: Re: cluster Message-ID: <20020330115821.B1805@wumpus.aonnetworks.com> Mail-Followup-To: linux-cluster@nl.linux.org References: <20020330050739.36370.qmail@web20009.mail.yahoo.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <20020330050739.36370.qmail@web20009.mail.yahoo.com>; from z_ling11@yahoo.com on Fri, Mar 29, 2002 at 09:07:39PM -0800 X-listar-version: Listar v1.0.0 Sender: linux-cluster-bounce@nl.linux.org Errors-to: linux-cluster-bounce@nl.linux.org X-original-sender: lindahl@keyresearch.com Precedence: bulk List-help: List-unsubscribe: List-software: Listar version 1.0.0 X-List-ID: List-subscribe: List-owner: List-post: List-archive: X-list: linux-cluster 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-archive@nl.linux.org Original-Recipient: rfc822;linux-cluster-archive@nl.linux.org On Fri, Mar 29, 2002 at 09:07:39PM -0800, z l wrote: > HA module or load balancing solutions can also be > called as cluster ? > what exactly does a cluster denote. Greg Pfister divides clusters into 3 kinds in his book: High Availability. That's the kind you've built before, often only 2 nodes. High Capacity. That's when you have many machines running the same serial program. A web farm is usually a combination of high capacity and high availablity -- the load balancer detects dead nodes and sends them no further work. High Performance. That's when you have many machines running a big parallel program, like an MPI code predicting the weather. "Beowulf" clusters often are in this category. Many clusters are combinations of the above. greg - Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/ From linux-cluster-bounce@nl.linux.org Sat Mar 30 20:42:39 2002 Received: from localhost.nl.linux.org ([IPv6:::ffff:127.0.0.1]:60371 "EHLO humbolt.") by humbolt.nl.linux.org with ESMTP id ; Sat, 30 Mar 2002 20:42:30 +0100 Received: with LISTAR (v1.0.0; list linux-cluster); Sat, 30 Mar 2002 20:42:02 +0100 (CET) Received: from gate.in-addr.de ([IPv6:::ffff:212.8.193.158]:24848 "EHLO mx.in-addr.de") by humbolt.nl.linux.org with ESMTP id ; Sat, 30 Mar 2002 20:41:54 +0100 Received: by mx.in-addr.de (mail.in-addr.de, from userid 10) id E0A2A64F0F; Sat, 30 Mar 2002 20:41:53 +0100 (CET) Received: by hermes.in-addr.de (Postfix, from userid 500) id 8038ED27; Sat, 30 Mar 2002 20:41:23 +0100 (CET) Date: Sat, 30 Mar 2002 20:41:23 +0100 From: Lars Marowsky-Bree To: z l Cc: linux-cluster@nl.linux.org Subject: Re: cluster Message-ID: <20020330204123.A28132@marowsky-bree.de> References: <20020330050739.36370.qmail@web20009.mail.yahoo.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <20020330050739.36370.qmail@web20009.mail.yahoo.com> User-Agent: Mutt/1.3.22.1i X-Ctuhulu: HASTUR X-listar-version: Listar v1.0.0 Sender: linux-cluster-bounce@nl.linux.org Errors-to: linux-cluster-bounce@nl.linux.org X-original-sender: lmb@suse.de Precedence: bulk List-help: List-unsubscribe: List-software: Listar version 1.0.0 X-List-ID: List-subscribe: List-owner: List-post: List-archive: X-list: linux-cluster 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-archive@nl.linux.org Original-Recipient: rfc822;linux-cluster-archive@nl.linux.org On 2002-03-29T21:07:39, z l said: > HA module or load balancing solutions can also be > called as cluster ? HA is usually achieved by "clustering" a few machines, yes. A load balancing solution like the Linux Virtual Server etc might in fact also be called a "cluster". > what exactly does a cluster denote. This is a very good question indeed; the answer isn't as straightforward as one might hope. The word is just "slightly" fuzzy. I would, short of recommending the excellent "In search of clusters" book by Gregory Pfister which was already mentioned, define it as A cluster couples a set of nodes (servers) which have separate processors, memory, local storage and networking each, in such a way that they serve as a "meta machine" (ie, cluster) for a given job. The "job" has different levels; from an enduser perspective, a webfarm is a meta-machine to serve HTTP and thus might be a cluster. But to a sysadmin, they do appear as very many separate machines to maintain, though some techniques may be used so that they appear as a single entity for some admin tasks etc - the job basically defines the "cluster boundary". > implemented HA solutions like two servers, one being > the hot standby for the other. can we call this a > cluster ? Yes, I would guess so. Sincerely, Lars Marowsky-Brée -- Immortality is an adequate definition of high availability for me. --- Gregory F. Pfister - Linux-cluster: generic cluster infrastructure for Linux Archive: http://mail.nl.linux.org/linux-cluster/