Editing Beowulf cluster

{{Short description|Type of computing cluster}}
{{More citations needed|date=November 2020}}
[[File:Beowulf-cluster-the-borg.jpg|thumb|The Borg, a 52-node Beowulf cluster used by the [[McGill University]] [[pulsar]] group to search for pulsations from binary pulsars]]
{{External media|image1=[https://www.computerhistory.org/revolution/supercomputers/10/71/909 The original Beowulf cluster built in 1994] by [[Thomas Sterling (computing)|Thomas Sterling]] and [[Donald Becker]] at [[NASA]]. The cluster comprises 16 [[white box (computer hardware)|white box]] desktops each running a [[i486]] DX4 processor clocked at 100&nbsp;MHz, each containing a 500&nbsp;MB hard disk drive, and each having 16&nbsp;MB of [[Random-access memory|RAM]], leading to a total of roughly 8&nbsp;GB of fixed disk storage and 256&nbsp;MB of RAM shared within the cluster and a performance benchmark of 500 M[[Floating point operations per second|FLOPS]].}}
A '''Beowulf cluster''' is a [[computer cluster]] of normally identical, commodity-grade computers networked into a small [[local area network]] with libraries and programs installed that allow processing to be shared among them. The result is a high-performance [[parallel computing]] cluster from inexpensive [[personal computer]] hardware.

==Original==
Beowulf originally referred to a specific computer built in 1994 by [[Thomas Sterling (computing)|Thomas Sterling]] and [[Donald Becker]] at [[NASA]].<ref>{{cite journal |journal=Proceedings, International Conference on Parallel Processing |date=1995 |volume=95|last1=Becker |first1=Donald J |last2=Sterling |first2=Thomas |last3=Savarese |first3=Daniel |last4=Dorband |first4=John E |last5=Ranawak |first5=Udaya A |last6=Packer |first6=Charles V |url=http://www.phy.duke.edu/~rgb/brahma/Resources/beowulf/papers/ICPP95/icpp95.html|title=BEOWULF: A parallel workstation for scientific computation}}</ref> They named it after the [[Old English]] epic poem, ''[[Beowulf<!--WP:OFTHESAMENAME-->]]''.<ref>
See [[Francis Barton Gummere]]'s 1909 translation, reprinted (for example) in {{cite book
| translator            = Francis B. Gummere
| year                  = 1909
| title                 = Beowulf
| url                   = https://books.google.com/books?id=ZCA4DbBCUlAC
| publisher             = Hayes Barton Press
| page                  = 20
| isbn                  = 9781593773700
| access-date           = 2014-01-16
| url-status            = 
}}{{dead link|date=September 2023}}
</ref>

==Systems==
No particular piece of software defines a cluster as a Beowulf. Typically only [[free and open source software]] is used, both to save cost and to allow customization. Most Beowulf clusters run a [[Unix-like]] operating system, such as [[Berkeley Software Distribution|BSD]], [[Linux]], or [[Solaris (operating system)|Solaris]]. Commonly used parallel processing libraries include [[Message Passing Interface]] (MPI) and [[Parallel Virtual Machine]] (PVM). Both of these permit the programmer to divide a task among a group of networked computers, and collect the results of processing. Examples of MPI software include [[Open MPI]]  or [[MPICH]]. There are additional MPI implementations available.

Beowulf systems operate worldwide, chiefly in support of [[scientific computing]]. Since 2017, every system on the [[Top500]] list of the world's fastest supercomputers has used Beowulf software methods and a [[Linux]] operating system. At this level, however, most are by no means just assemblages of commodity hardware; custom design work is often required for the nodes (often [[blade server]]s), the networking, and the cooling systems.

== Development ==
[[File:BSC-Beowulf-cluster.JPG|thumb|Detail of the first Beowulf cluster at [[Barcelona Supercomputing Center]]]]

A description of the Beowulf cluster, from the original "how-to", which was published by Jacek Radajewski and Douglas Eadline under the [[Linux Documentation Project]] in 1998:<ref>{{Cite web |title=Beowulf HOWTO |last1=Radajewski |first1=Radajewski |last2=Eadline |first2=Douglas |work=ibiblio.org |date=22 November 1998 |access-date=8 June 2021 |url= http://ibiblio.org/pub/Linux/docs/HOWTO/archive/Beowulf-HOWTO.html#ss2.2 |version=v1.1.1}}</ref>

{{Blockquote|
Beowulf is a multi-computer [[computer architecture|architecture]] which can be used for [[parallel computing|parallel computations]]. It is a system which usually consists of one server node, and one or more client nodes connected via [[Ethernet]] or some other network. It is a system built using commodity hardware components, like any PC capable of running a [[Unix-like]] operating system, with standard Ethernet adapters, and switches. It does not contain any custom hardware components and is trivially reproducible. Beowulf also uses commodity software like the FreeBSD, Linux or Solaris operating system, Parallel Virtual Machine ([[Parallel Virtual Machine|PVM]]) and Message Passing Interface ([[Message Passing Interface|MPI]]). The server node controls the whole cluster and serves files to the client nodes. It is also the cluster's console and [[gateway (computer networking)|gateway]] to the outside world. Large Beowulf machines might have more than one server node, and possibly other nodes dedicated to particular tasks, for example consoles or monitoring stations. In most cases, client nodes in a Beowulf system are dumb, the dumber the better. Nodes are configured and controlled by the server node, and do only what they are told to do. In a disk-less client configuration, a client node doesn't even know its [[IP address]] or name until the server tells it.

One of the main differences between Beowulf and a Cluster of Workstations (COW) is that Beowulf behaves more like a single machine rather than many workstations. In most cases client nodes do not have keyboards or monitors, and are accessed only via remote login or possibly serial terminal. Beowulf nodes can be thought of as a [[Central processing unit|CPU]] + memory package which can be plugged into the cluster, just like a CPU or memory module can be plugged into a motherboard.

Beowulf is not a special software package, new network topology, or the latest kernel hack. Beowulf is a technology of clustering computers to form a parallel, virtual supercomputer. Although there are many software packages such as kernel modifications, PVM and MPI libraries, and configuration tools which make the Beowulf architecture faster, easier to configure, and much more usable, one can build a Beowulf class machine using a standard Linux distribution without any additional software. If you have two networked computers which share at least the <code>/home</code> file system via [[Network File System|NFS]], and trust each other to execute remote shells ([[Remote Shell|rsh]]), then it could be argued that you have a simple, two node Beowulf machine.
}}

=== Operating systems ===
{{unreferenced section|date=November 2020}}
[[File:Beowulf.jpg|thumb|A home-built Beowulf cluster composed of [[White box (computer hardware)|white box]] PCs]]

{{As of | 2014}} a number of [[Linux distribution]]s, and at least one [[Berkeley Software Distribution|BSD]], are designed for building Beowulf clusters. These include:

* [[MOSIX]], geared toward computationally intensive, IO-low applications
* [[Rocks Cluster Distribution]], latest 2017
* [[DragonFly BSD]] OS, latest 2022
* [[Quantian]] OS latest 2006, a [[live CD|live DVD]] with scientific applications, remastered from [[Knoppix]] 
* [[Kentucky Linux Athlon Testbed]], physical installation at University of Kentucky

The following are no longer maintained:
* [[Kerrighed]] (EOL: 2013)
* [[OpenMosix]] (EOL: 2008), forked from MOSIX
* [[ClusterKnoppix]] OS, forked from [[Knoppix]] OS, forked from [[OpenMosix]]
* [[PelicanHPC]] OS latest 2016, based on [[Debian#Live images|Debian Live]]

A cluster can be set up by using Knoppix bootable CDs in combination with [[OpenMosix]]. The computers will automatically link together, without need for complex configurations, to form a Beowulf cluster using all CPUs and [[Random-access memory|RAM]] in the cluster. A Beowulf cluster is scalable to a nearly unlimited number of computers, limited only by the overhead of the network.

Provisioning of operating systems and other software for a Beowulf Cluster can be automated using software, such as [[Open Source Cluster Application Resources]]. OSCAR installs on top of a standard installation of a supported Linux distribution on a cluster's head node.

== See also ==
{{Portal|Free and open-source software}}

* [[Aiyara cluster]]
* [[HTCondor]]
* [[Grid computing]]
* [[Kentucky Linux Athlon Testbed]]
* [[Maui Cluster Scheduler]]
* [[Open Source Cluster Application Resources]]
* [[Stone Soupercomputer]]
* [[Oracle Grid Engine]]

== References ==

{{Reflist}}

==Bibliography==
* ''Beowulf Cluster Computing With Windows'' by Thomas Lawrence Sterling 2001 {{ISBN|0262692759}} MIT Press
* ''Beowulf Cluster Computing With Linux'' by Thomas Lawrence Sterling 2001 {{ISBN|0262692740}} MIT Press

== External links ==
* [http://www.beowulf.org/ The Beowulf Cluster Site]
* [https://www.final-labs.org/dev/protogonus A detailed description of building a Beowulf cluster]

{{Parallel Computing}}
{{Authority control}}

[[Category:Cluster computing]]
[[Category:Parallel computing]]
[[Category:Job scheduling]]
[[Category:NASA spin-off technologies]]
[[Category:Unix software]]