Follow Slashdot blog updates by subscribing to our blog RSS feed

 



Forgot your password?
typodupeerror
Mandriva Businesses

MandrakeClustering Shows Off At ISC2003 157

Posted by timothy
from the snap-together-modules dept.
joestar writes "Just released today at ISC2003, Germany, is "MandrakeClustering", a high-performance computing Linux distribution/solution, which sounds interesting, at least in the PR: Pentium support with optimizations made with the Intel compiler, 64-bit Opteron support (with in this case, up to 16 GB of RAM for each cluster's node!), parallelized URPMI (Mandrake's apt-get) and other dedicated tools. This product is based on a one-year research project "CLIC" involving MandrakeSoft and partners. A good snapshot of the product running a 3D real-time demo is available here. The interesting point now: MandrakeClustering's goal is to provide a system which is easy to deploy, easy to administer and use. Well... Mum would certainly love to play Quake with this toy."
This discussion has been archived. No new comments can be posted.

MandrakeClustering Shows Off At ISC2003

Comments Filter:
  • by ObviousGuy (578567) <ObviousGuy@hotmail.com> on Tuesday June 24, 2003 @07:30PM (#6290114) Homepage Journal
    But what are people doing with these things in the wild? Is everyone running rendering farms?

    Do file/mail servers really need this kind of horsepower (assuming you aren't Google or Yahoo, of course)?
    • Actually...

      They're running demos of high-resource rendering demos people who buy this setup can run at Comdex 2004.

    • if i had the money, I'd be running a render farm right now instead of setting my render to low priority just so I can check slashdot. The software looks VERY nice, but unfortunately, unlike the OS, the hardware can't be had for nothing.
    • The cluster management software may be useful for things like offices, class rooms, etc, where one needs to maintain a bunch of identical systems.
    • by Anonymous Coward
      I would imagine that educational institutions might like a quick and dirty rendering farm. Complex projects, like CAD drawings or animations, could be shuffled off to the local render farm, instead of being run all night on one computer. Heaven help you if there is a power outage, hardware failure, or the occasional human interference. I personally think that there is a market that could be created. Just not necessarily for Quake III or Doom III.
    • by bm_luethke (253362) <luethkebNO@SPAMcomcast.net> on Tuesday June 24, 2003 @08:40PM (#6290586)
      Where I work (www.ornl.gov) we run weather models, nuclear stuff (explosions, fallout tracking, power plants, nuclear medical materials simulations), genome projects, particle flow analisys, a bunch of stuff (though unfortunatly for Mr. Dyslexic here no speel checker for slashdot :) ). Oil companies process seismic data to try and map what's underground, medical companies help develope medicines and gene therapy.

      Some of the projects (human genome most currently) may run for weeks over 10-20 terabytes of data. The algorithms used are pretty mundane - you just need to do them A LOT. Some of the particle flow stuff may model individual molecules in a furnace for example (not rendering the frames, but modelling the interactions and how to create a more effecient jet engine).
    • by Anonymous Coward
      CFD is one of the major uses of clusters
    • by Anonymous Coward
      But slightly related. :-)

      This clustering thing gave me the idea -- pardon me if not original:

      Mandrake could sell CPU time! Like those SETI, cancer drug molecule manipulation, code breaking/ prime number finding etc.

      We, Mandrake users, could (on a voluntary basis) donate hours of unused computer time to give them what to sell.

      Is there a market for this?

      Is this possible/viable?

      At least, they could use such processing for their distro making related tasks.

      Additionally:

      They could install automatically
    • by InodoroPereyra (514794) on Wednesday June 25, 2003 @12:30AM (#6291738)
      But what are people doing with these things in the wild?

      Huh ? How got this modded as insightful ? Anyways, if you read the announcement, this is for high performance computing. It is a turn-key solution to easily set up, maintain and monitor a parallel cluster. Parallel clusters are essentially arrangements of computers that allow to process data at each computer at the same time ("in parallel"), and then collect the results and combine them, thus making possible to achieve supercomputing power out of inexpensive hardware.

      The announcement very clearly states which is the market niche they target (i.e. research labs and the like):

      "MandrakeClustering is not only a impressive example of innovative Linux technology, but it's also a complete "all-in-one" solution that includes support & assistance for "real-world" clustering deployments and critical research lab activities. We are extremely confident that this product will answer the needs of numerous technology research centers where time of deployment and ease of administration are highly valued", said FranÃois Bancilhon, MandrakeSoft CEO.

      By the way. It is so nice to see Mandrake aiming at corporate/government customers, this is going to be very healthy for the company. I would expect and encourage more of this ...

    • Video babee - good quality cg video takes some serious processing. My blender models (after just a month of playing with it) are starting to take hours to render a few seconds (it's those image maps for mirror balls that kills you - it has to render 6 points of view for each frame, running 30 seconds per frame) so yes I NEED A CLUSTER! Also cinelerra has support for extra render nodes built in.

      My goal is to produce my own cg nightly news program just like the big boys - reality made up (or biased toward)
  • Imagine a Beowulf cluster of these ! I'd kick your ass at quake with one !
    • yes (Score:4, Interesting)

      by SHEENmaster (581283) <travis&utk,edu> on Tuesday June 24, 2003 @07:51PM (#6290264) Homepage Journal
      because we all know that render times kick ass when processors have a 12mb/s wifi link between them. What were multi-processor manufacturers thinking building the processors into one unit when beowulfing is available!

      (Beowulf clusters have long-term power but slow reaction times. Firewire networking (maybe scsi?) would help with this, but the node->node bandwidth is still an issue.)

      Why would Mandrake use Intel's compiler for the Opteron?
      • by mindstrm (20013)
        Firewire can do 1Gbps?
      • by draziw (7737) *
        There are better ways to connect systems. eg: Myricom [myri.com] and Quadrics [quadrics.com] (More money, but faster)
      • by mandolin (7248)
        because we all know that render times kick ass when processors have a 12mb/s wifi link between them.

        Even "normal" people can afford gigabit ethernet these days.

  • by INMCM (209310) on Tuesday June 24, 2003 @07:31PM (#6290120) Homepage
    Is it as fast as the new Macs?
    • by Anonymous Coward
      Wait until you read it in the NY Times
    • When run at standard settings yes, but if you turn off the compiler optimizations, floating point optimizations, memory optimizations, user optimizations, reality optimizations, and go-fast-now optimizations, the Macs pull ahead by a whole point(when compared to one node)!
  • openmosix easier (Score:2, Informative)

    by cdc179 (561916)
    It's easier to manage an openmosix cluster than this mandrake cluster. Plus you can run diskless nodes using ClusterKnoppix!.
    • Re:openmosix easier (Score:3, Informative)

      by KeyserDK (301544)
      READ! It clearly states that nodes can boot via PXE.
      • Re:openmosix easier (Score:2, Informative)

        by cdc179 (561916)
        Some organizations management might not allow them to set up PXE. In this case this gives them the convienence of booting workstation nodes into the cluster at night with just a CD. You can read all you want, but you missed the whole point of using the CD for non-Linux nodes.
        • Some organizations management might not allow them to set up PXE.

          So you think someone is going to setup a high performance cluster on the same network segment as the rest of the corporate network?

          Spend a fortune on high performance compute nodes, and not buy a seperate switch?
    • I've been using Clic for a while on some old diskless PIIIs- too old for PXE- but it wasn't hard to just clone a system and send it out over netboot. If I make major changes to the setup I have to make a new boot image, but that's easy.

      It's an old system, but it still gets used (mostly Monte Carlo simulations, a few molecular dynamics, and some other home brewed science type codes).

      The nodes (and now even the head!) don't have cd drives, because I'm too lazy to go to re-app and get some...
      • Whaddya mean "too old for PXE?" That's a function of the network card. If, by chance, you're using an oldy-moldy Intel PRO100B adapter (circa 1995 vintage), check the PXE ROM socket to see if there's a chip in it. Newer ones (circa 1998) had the chip soldered onto the card, which turned out to be cheaper when flash prices came down.
    • You cannot compare the two, as Openmosix and MandrakeClustering have really different goals. MandrakeClustering is for High Performance Computing. This includes stuff like a batch system for job submission, MPI and a fast interconnect.

      Openmosix is for load balancing, it can migrate jobs between the nodes in a cluster, but it's not suited for the parallel number crunshing.

  • by RLiegh (247921) on Tuesday June 24, 2003 @07:34PM (#6290142) Homepage Journal
    If so, then I think it's a pretty good investment. :)

    So much for the "bailing out a failing business, let the market sort it out" mindset. If Mandrake hadn't appealed to the public, they would not have had the $ to come up with this. :)
  • It's not like there's anything free [bofh.be] out there that does this.
    • My bad (Score:4, Informative)

      by mikeophile (647318) on Tuesday June 24, 2003 @07:51PM (#6290256)
      It can be downloaded here [mandrakesoft.com]

      The $2,960 is to get support with it.

      • No, you can't download it. That link is for the predecessor Mandrake CLIC, which isn't the same as MandrakeClustering.
        But I'm sure you can get the sources of all the parts developed by Mandrake, as usually all their stuff if GPLed and most of MandrakeClustering is based on the usual Mandrake 9.0.
  • "Imagine a Mandrake cluster of ..."

    That just doesn't have the same ring to it :-(
  • Mandrake (Score:3, Interesting)

    by StarTux (230379) on Tuesday June 24, 2003 @07:44PM (#6290205) Journal
    Wow, first /. posting of something Mandrake related in some time. Is it just me, or have a lot of announcements not even mentioned Mandrake recently?

    Even though I don't use Mandrake I do wish them the best, especially with a number of Linux distro's competing for the desktop. Hopefully they will make some money from this venture.

    StarTux
    • I agree. Althought I currently use Gentoo now that I am more experienced, Mandrake was a pretty enjoyable and painless introduction to linux for me
  • by Geminus (602334) * on Tuesday June 24, 2003 @07:51PM (#6290258)
    An OS that surrenders under pressure?
    • I know it's trendy to make French jokes (Freedom jokes?) right now, but while the US is still pissing over Iraq (Oil exports have started again. w00t! Still no real infrastructure for the locals, but hey, we've got the OIL) and even Afghanistan, the French are in the DR Congo fighting for the people there.

      Props to the French for getting on with the business where it really counts.

      And No, I'm not European.

    • It appears to me the French are the only ones who did not surrender to Bush's pressure to go fight in Iraq, but instead adamantly opposed the whole idea.
      • by phoxix (161744) on Tuesday June 24, 2003 @09:38PM (#6290930)
        This has nothing to do with French people. I am plenty sure that there are many people from around the world that have worked on linux, *AND* that you probably disagree with outright.

        But none of that matters. What matters is that while our people disagree, and our governments fight, we still can come together on something at least. And in our geeky world, that something is free software.

        Sunny Dubey

        PS: The right to disagree is rather powerful, sadly people don not see that when they do disagree with each other.

        • The right to disagree: Isn't it a nice right to have? And just think of the billions of people in the world who do not even have it. What we take for granted, there are billions who could only dream to be able to say "I disagree" without fear of getting thrown into prison or shot on the spot.

          And, as a mongrel, I can attest that I'm proof positive that men and women of all races, creeds and religions can get together in at least one aspect.
  • Should be in next week. I'll have my Mom drive it, too. Wait for the slashdot article, "Mom Meets Linux (Super Computer Ed)"
  • by molnarcs (675885) <molnarcsNO@SPAMgmail.com> on Tuesday June 24, 2003 @08:12PM (#6290395) Homepage Journal
    I've been using Mandrake for a while, and I would hate to see this company go - they have many great tools, and they put every single line of code they wrote under the gpl (unlike YAST for instance, even though I love SuSe too). They probably need as much PR as they can get, and this was a good idea.

    Its time for some features, like their excellent urpm* tools to get more attention (I wonder why it received such scare coverage, for it is the only package management tool that is on par with apt-get among rpm-based distributions - maybe with the exception of apt-rpm). Another great tool, excellent in large deployments is draksync.

    Check out these sites:
    urpmi mini-howto [tiscali.co.uk] and easy-urpmi [zarb.org]
  • by codepunk (167897) on Tuesday June 24, 2003 @08:13PM (#6290399)
    It ain't a cluster until it has a clustered file system. Until it has that it is nothing more than another high availability solution.
    • by GigsVT (208848) *
      I guess you might want to tell that to all the people that do high performance clustering with nothing more than ssh.
    • So why does this get a "5 Insightful" exactly ? It is basically incorrect
    • by Götz (18854)
      I guess you should read a few papers about what clusters really are. I shouldn't hurt to RTFA either. MandrakeClustering isn't a high availability solution, it's for number crunching. Stuff like Openmosix is for high availability.

      There are basically these two kinds of clusters, one for the traditional high performance computing stuff to replace expensive supercomputers, the other one is for high availability, e.g. for preventing the Slashdot effect.

  • Spend $20k on Plasma Displays
    Spend $40k on Cluster Hardware
    Leave the cords of the subwoofers and amplifier exposed ...
    Priceless....

    Monster Linux Garage Mantra: If you are going to build something cool, it should look cool too.

    Sample AMD Case-mod [ocmodshop.com]
  • That's 3 separate instances of windowmaker, right? With Xinerama, wouldn't the icons only be on 1 desktop? Dunno, never used it before.
  • Clusters, Mandrake, Opteron, 16GB. Add in an Xbox and Natalie Portman and you've got yourself a Slashdotter's wet dream. Well, I guess Natalie accomplishes that on her own, but I'm pretty sure if I had the rest of that stuff it would impress Natalie enough for her to date me. At least I hope so, or I've wasted a LOT of money over the years.
  • Maybe they should've just gotten an ATI 9800 Pro to do the rendering instead of all those loud and expensive boxen ::)
  • by rune2 (547599)
    Let's hope that their webserver is a MandrakeCluster too what with the good ol'-fashioned Slashdotting it's getting now.
  • Imagine a Mandrake cluster of those....
    I mean... the opterons ;)
  • The 3D Demo (Score:2, Informative)

    by Hufo (684441)
    I am the main programmer of the 3d demo shown on the screenshots. This demo combines a distributed interactive fluid simulation with graphical intensive rendering using pixel shaders. It uses Net Juggler [sf.net] to manage the distribution on the cluster. You canse some picture and a video of the fluid simulation in our gallery [sf.net] . The shader for the rabbit's fur is explained on this page [planet-d.net] of my website (please be gentle with the server...).
  • Enough about Quake (Score:3, Interesting)

    by SleezyG (466461) on Wednesday June 25, 2003 @07:38AM (#6292877)
    Here it comes... I'm going to vent. How is that ./ readers (clearly amongst the most well-educated of the 'net surfing masses) always fall for the bigger == better claim? First of all, Quake, or any computer game that I can think of, does not have multi-node support. You have to rewrite the code to support the PVM (Parallel Virtual Machine) message passing library, for example.

    And it's not just with clusters we seem to have this problem. Let's look at Apple computers and the G5. A 64-bit machine will simply allow you to add/subtract/multiply/etc really big integers faster. How often do think you use numbers that large? Encrypting or decrypting emails is the only thing that comes to mind for even an "above average" user. Unless you run software that supports multiple processors such as web and database servers or want to watch a DVD while you play Quake, a multiprocessor machine is not for you. The relative performance increases are negligible. It's simply a marketing ploy that allows Apple to make a high profit on the low volume of units they ship. They throw an additional $300 uP in their computers and charge an extra $1000!

    But in Apple's defense, their displays rock. Any luck getting one of those running on a PC? I'm done ranting. Have a nice day.
  • hey (Score:2, Funny)

    by pulse2600 (625694)
    Imagine a beowulf cluster of MandrakeClusters!
  • Now we know what type of performance we can get when we get all of these penguins to swim in the same direction all day.

Mr. Cole's Axiom: The sum of the intelligence on the planet is a constant; the population is growing.

Working...