Forgot your password?
typodupeerror
News

The Globus Project and Computational Grids? 8

Posted by Cliff
from the throwing-bogomips-at-every-problem dept.
SigmoidCurve asks: "The latest issue of The Economist (sorry no URL, using the quaint paper version) has an article about software used to create 'computational grids' like those used with SETI@home. They mention the Globus project which is released under an open source license. The Economist article speculates on the future of the Grid and what kinds of projects would benefit from becoming massively parallelized. What does the Slashdot community think of this and similar projects? Does anyone have any comments about Globus or other such software? What projects do you know about that could take advantage of the untold billions of idle processor cycles lying in wait?" For those of you looking for more information on the Globus Project, you can check out their FAQ.
This discussion has been archived. No new comments can be posted.

On Globus and the Computational Grid?

Comments Filter:
  • It's here. [economist.com]

    Furthermore, "Grid Computing" != SETI@Home. Grid Computing is quite a bit more than just donating your spare CPU cycles to something like SETI@Home, distributed.net and such. Grid computing is going to happen. It may not include everyone and their grandmother's desktop machines as cycle providers.

  • There's a very good book called "The Grid: blueprint for a new computing infrastructure" that describes the grid concept, its future and many of its possibles application. It's basically a collection of papers written by the main people behind this. I don't have a link at hand, but I'm sure Amazon has it. Is a very expensive book, so I borrowed it from my university library.

    I remember a seminar that I attended at the Pittsburgh Supercomputing Center regarding Globus and the Grid, and thinking "uau, THIS is the future of computing". It was presented by people from NCSA ( National Center for Supercomputing Applications ) who were extremely knowledgeable and open to all kind of questions. Among other things ( 3 days of seminar ) they showed how they connected multiple Crays and SGI Origin computer across the world that were running Globus ( one from Germany, one from PSC and I believe one from UCSD or NCSA ) for a real-time calculation for a phyical experiment.
  • This story was submitted 3 or 4 hours ago, during this robust time of mid-afternoon EDT "surf don't work" time, and yet there is still no lame FRIST PROST posts.

    The world has gone mad, unplugged, and I am for some reason the only person able to access Slashdot.

    I'm afraid.

    Hold me.

    <disclaimer class="this is not a first post"/>

  • Actually, the whole series of articles seems to be available. The issue with this supplement came out a couple of weeks ago, and I'm still just starting to look though it. Interesting stuff.

    The series. Some articles are short -- a couple paragraphs -- but most are pretty in depth, like "Computing Power On Tap" as linked to above:

    Sorry if I was being a wiseass with the not-first-post. I'm interested in the subject, and was disappointed that no one had commented yet. I browse via the search page now just so that all the articles that never make it to the front page don't slip through; often they're more interesting & the discussion is at a more interesting level. Or not, this is Slashdot after all. Anyway...

  • Actually, I've noticed that every once in a while, these sort of "Ask Slashdot" questions don't get a frist prost-kind of response. And if your post is not "first post" then does that make mine the first post? This really *IS* the Twilight Zone! My first post is first post!
  • Globus has a free link to the article at The Enconomist here [economist.com]. I guess the $2.95 is to pay for the privilege of using the search box at the Enconomist ;-)
  • The electronic version can be found here [economist.com]. However, it requires a subscription or $2.95 to read the article. I have the latter, but no desire to part with it for this :-) Babbage: I think there are so few responses because this article isn't visible on the front page. It takes a search or clicking on "older stuff" to find it. At least that's how I did. I think using spare CPU cycles for such projects is great, as long as:
    • the results from the public's participation are made public, not kept for private gain as was reported with some drug and genome distributed research projects. If they want to make money from it, then they should pay for the CPU time, and state what they're doing up front
    • computers aren't left on simply to run such programs
    When the power crisis hit here in California, SETI@Home [berkeley.edu] specifically requested that SETI users turn off their computers during stage 2 and 3 power alerts, and has since added a page on "running green" [berkeley.edu], which also suggests that you turn off your monitor when not in use, and requests that you not keep your computer on just to run SETI@Home. And it has some handy links to sources about conserving energy, too.
  • An article on Sun's "Best Practices" website [sun.com] describes the risks of running distributed computing environment clients on a production network.

When you don't know what to do, walk fast and look worried.

Working...