SlackwareThis Forum is for the discussion of Slackware Linux.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
Hi all,
I am about to try and set up a linux cluster to run some programs like Folding at Home and other things of that sort and was wondering how would i go about setting up a linux cluster?
I did a little bit of googling and found mostly articles from 2003-2005 but nothing much newer. These are my two systems that i am planning to cluster with the possibility of more being added on:
AMD Duron 1.0ghz
192mb ram
Intel P3 450mhz
192mb ram
basically i have these two PCs sitting around and want to put them to use.
I'm pretty interested in Linux Clusters and I'm planning to set one up myself (I just need to find the time for a project like that), that's why your thread called my attention.
I don't know if you have already seen this, but maybe you could start with the book Engineering a Beowulf-style Compute Cluster by Robert Brown. You can find it online here:
Well, today I thought to ask you something: if you might be interested in documenting the building process of your cluster, like a log or something of that kind (I've seen little of that around, most of them talk about their cluster when it is working). Maybe when you're done you can share the "diary" with everybody. I promise I'll do that when I get to build my own cluster.
Any issues you might have I'll be glad to know about to try to help you solve them. It will be a great learning experience.
Definitley agreed, I have noticed that there are not many logs on the internet on how to do this kind of stuff and since i will be working with different processors and set-ups it should make for some interesting results. I'll report issues and things that i have on this thread until i can get my web server back up. unfortunately i have to assemble one of the PCs as it was taken apart for a deep cleaning.
Distribution: Mac OS X Leopard 10.6.2, Windows 2003 Server/Vista/7/XP/2000/NT/98, Ubuntux64, CentOS4.8/5.4
Posts: 2,986
Rep:
I have a job distribution cluster set up for work with about 16-CPU's, and I use CentOS. I use Sun's Grid Engine program. Easy to use and set up.
Also check out Condor.
And do people really like to read other's posts on setting up systems and clusters? If so, I can do this on my personal website! Very exciting, I think ...
Hi,
I don't know if you have already seen this, but maybe you could start with the book Engineering a Beowulf-style Compute Cluster by Robert Brown. You can find it online here:
This is the link that I would have recommended too, if Choucete hadn't beaten me to it.
A good portion of the book is about figuring out whether or not to build a cluster, how to build it, and how to write software that can take advantage of it.
My question: what applications are set up to use parallel processing without a lot of work?
Definitley agreed, I have noticed that there are not many logs on the internet on how to do this kind of stuff and since i will be working with different processors and set-ups it should make for some interesting results. I'll report issues and things that i have on this thread until i can get my web server back up. unfortunately i have to assemble one of the PCs as it was taken apart for a deep cleaning.
Glad to know this!
Quote:
Originally Posted by Micro420
And do people really like to read other's posts on setting up systems and clusters? If so, I can do this on my personal website! Very exciting, I think ...
At least I know I would found it interesting (and C/Sniper seems to agree with me). As I said before, it would be a great learning experience, and people reading from others experience might feel encouraged (or not) to build their own. In fact, I'd really like to know more about your cluster.
Quote:
Originally Posted by Lufbery
This is the link that I would have recommended too, if Choucete hadn't beaten me to it.
Thanks a lot for this, I feel flattered as the Linux newbie I am.
Quote:
Originally Posted by C-Sniper
--Update1--
acquired some more PCs
it is looking like i will have 5 Intel P3 450mhz computers as my nodes with the one Duron 1.0ghz as my main.
i am also interested, i have a 850mhz p3 with 768 ecc ram, and about 4-5 500-450 mhz pentium 2-3 with 128 megs of ram each. i go to school for network administration, but the crash courses they give are too smashed together and i dont really learn much from cramming. please do tell more! =)
isnt school evil? i did like 3 mcse classes, then a unix class and now theres no way i could go back to an MCSE class to learn microsoft shit. the bad part is that i HAVE to if i want to graduate. i would rather take more unix classes, but they only offer one.
Setting up a Beowulf cluster is very simple, but finding practical software that makes use of MPI or PVM is another story entirely. Unless you have a very specific (probably scientific) use for it, there is little reason to put together a Beowulf.
That is why I always liked OpenMosix. It didn't require any special applications, it treated different nodes in the cluster as if they were different CPUs in the same machine; the master node just handed out processes to the slaves as it saw fit based on their performance and level of activity. It let you have a relatively normal desktop environment on the front end, but a clustered work distribution system running in the background. Unfortunately it is no longer being developed (at least by the original authors).
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.