Mailing List ArchiveSupport open source code!
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]Re: tlug: Clustering Linux Boxes
- To: tlug@example.com
- Subject: Re: tlug: Clustering Linux Boxes
- From: "Stephen J. Turnbull" <turnbull@example.com>
- Date: Tue, 25 May 1999 15:20:56 +0900 (JST)
- Content-Transfer-Encoding: 7bit
- Content-Type: text/plain; charset=us-ascii
- In-Reply-To: <Pine.LNX.4.10.9905250151120.11652-100000@example.com>
- References: <19990524133153.14388.qmail@example.com><Pine.LNX.4.10.9905250151120.11652-100000@example.com>
- Reply-To: tlug@example.com
- Sender: owner-tlug@example.com
>>>>> "Chris" == Christopher Sekiya <wileyc@example.com> writes: Chris> On Mon, 24 May 1999, Austin Kurahone wrote: >> Now the question that I have is that will making a cluster(eg: >> Avalon with Extreme Linux) give me performace advantages for >> all my programs, or do they need to be specially coded to take >> advantage of the cluster. Chris> They need to be specially coded. Not if you distribute processes rather than threads. Also, there are (eg) FORTRAN compilers which will automatically take advantage of a certain amount of parallelism that is inherent in things like do loops. I doubt there are C compilers that do this (at all, let alone free code), but you could probably hack an optimization stage into GNU C/C++. What one person could do by September will probably suck, but it would be too cool a hack for people to let alone. It would not suck any more on October 1 ;-) And of course some programs will surely run more slowly in parallel ;-) >> If they need to be specially programed, what is the feasibility >> of coding/implementing a transarent layer(a module maybe) that >> will allow the Linux kernel to automatically distribute the >> separate threads/tasks to the various nodes in the cluster??? Chris> This is a userspace issue (nothing special needs to be done Chris> to a stock kernel to do clustering[1]). Check out MPICH Chris> and/or PVM. If you're using a distribution which allows searching package space (eg debian) search on `mpi' and `pvm'; there are several packages in Debian already (SPAlina or something like that is one) taking advantage of them. Since the interfaces are different the packages get a -mpi or -pvm suffix. -- University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Institute of Policy and Planning Sciences Tel/fax: +81 (298) 53-5091 __________________________________________________________________________ __________________________________________________________________________ What are those two straight lines for? "Free software rules." ------------------------------------------------------------------- Next Nomikai: May 26 (Wed), 19:00 Tengu TokyoEkiMae 03-3275-3691 *** Special guest: Eric S. Raymond Next Technical Meeting: June 19 (Sat), 12:30 place: Temple Univ. Topic: Linux SMP and the Intel SC450NX Quad Xeon Server platform ------------------------------------------------------------------- more info: http://www.tlug.gr.jp Sponsor: Global Online Japan
- References:
- tlug: Clustering Linux Boxes
- From: "Austin Kurahone" <a_downunder@example.com>
- Re: tlug: Clustering Linux Boxes
- From: Christopher Sekiya <wileyc@example.com>
Home | Main Index | Thread Index
- Prev by Date: RE: tlug: [reminder] 5/26 meeting
- Next by Date: Re: tlug: Clustering Linux Boxes
- Prev by thread: Re: tlug: Clustering Linux Boxes
- Next by thread: Re: tlug: Clustering Linux Boxes
- Index(es):
Home Page Mailing List Linux and Japan TLUG Members Links