Please create an account to participate in the Slashdot moderation system

 


Forgot your password?
Close
typodupeerror
×
Hardware

Online Game Cluster 107

Posted by timothy from the drool-pools dept.
axehind writes "Carlo Daffara posted to the openMosix mailing list about his game cluster. It's a 6 node cluster using Athlon XP cpus and running linux & openmosix (with some qdisc trickery) for the OS. It is used to host approximately 1000 users playing online games, like Jedi Knight and Quake III. Here's a link (italian) to the pictures."
This discussion has been archived. No new comments can be posted.

Online Game Cluster More

Online Game Cluster

Comments Filter:

  • FP (Score:4, Funny)

    by Anonymous Coward on Saturday November 09, 2002 @01:05PM (#4632746)
    Don't imagine a beowulf cluster of these.

    • Re:FP (Score:2)

      by jo42 ( 227475 )

      Query: What do you call a cluster of slashdot Linux geeks?

      Response: The boys that called "Beowulf!".

  • RDMA for GigEthernet (Score:3, Insightful)

    by selectspec ( 74651 ) on Saturday November 09, 2002 @01:12PM (#4632786)
    This is slightly offtopic, but RDMA for GigE is nearing finalization. I would imagine this will play as a huge win for these clustering technologies.

  • /.ed (Score:5, Funny)

    by psergiu ( 67614 ) on Saturday November 09, 2002 @01:13PM (#4632788)
    or:

    Il luogo è shashdotado. :)

  • waiting (Score:5, Funny)

    by jdkane ( 588293 ) on Saturday November 09, 2002 @01:13PM (#4632789)
    It is used to host approximately 1000 users playing online games
    More like "It was used to..."
    During the standard slash/dot.ing period, the cluster probably serves about two Quake players at max.

  • Imagine... (Score:5, Interesting)

    by SlimFastForYou ( 578183 ) <konsoleman@yaAUDENhoo.com minus poet> on Saturday November 09, 2002 @01:15PM (#4632803) Journal
    That same cluster spec running 1 instance of a game for someone to play on :). Someday I hope for clustering software to be to the point where when someone at a LAN party goes to sleep, their processing power can be used to help other machines.

  • Comment removed (Score:3, Funny)

    by account_deleted ( 4530225 ) on Saturday November 09, 2002 @01:19PM (#4632822)
    Comment removed based on user account deletion

  • Link to Google Cache (Score:5, Informative)

    by wmspringer ( 569211 ) on Saturday November 09, 2002 @01:20PM (#4632825) Homepage Journal
    See the site here [216.239.53.100].

    (I

  • For once they might not be lying....... (Score:1, Funny)

    by Anonymous Coward
    QuakePlayer: omfg, BS!!~ laaaaagggggg!!!!11

  • Why use Mosix ? (Score:4, Interesting)

    by boaworm ( 180781 ) <boaworm@gmail.com> on Saturday November 09, 2002 @01:20PM (#4632831) Homepage Journal
    OpenMosix is used for load balancing, allowing processess to migrate during runtime.. I wonder what happens if you are trying to "gib/frag" someone and the server process migrates to another node .. ;-)

    It seems to me it would make more sense to use only the mentioned queue-system to position the games evenly.
    • There were recent posts with problems regarding the game freezing during migrations, not with this cluster, but with a similar one someone is trying to set up.

      From the post:
      We used openmosix, and some linux qdisc trick to guarantee smooth play even in quite heavy network activity.

      The bottom line was to just have a lot of bandwith. It looks like he is doing some traffic shaping or something... I don't read his language.

  • supercomputer gaming (Score:5, Informative)

    by mmca ( 180858 ) on Saturday November 09, 2002 @01:22PM (#4632844) Homepage
    http://www.supercomputergaming.com/

    is doing something similar. If you cut through all the marketing hype it looks pretty cool.

    -M

  • No wait (Score:3, Funny)

    by mao che minh ( 611166 ) on Saturday November 09, 2002 @01:24PM (#4632855) Journal
    Now imagine a CLIC [mandrakesoft.com] cluster of these....

    • Re:No wait (Score:3, Insightful)

      by zaqattack911 ( 532040 )
      Wait.. does a CLIC cluster work the way openMosix does? openMosix behind the scenes migrates processes (or server connections) to the least busy machine at that time. It doesn't do threads, and I don't even know how it works if it's not a network drive.

      How does CLIC do it?
      Don't assume it's the same thing... christ you could even call my webserver a cluster just cuz it's webpages spread amoung two different machines... but they (the servers) really have no concept of eachother.. and it doesn't share threads or diskspace or anything really.

      Clust is more of a buz word. It is what is technically going on that I'm most concerned with.

      • Re:No wait (Score:5, Funny)

        by The_Sock ( 17010 ) on Saturday November 09, 2002 @02:03PM (#4633026) Homepage
        Clusternig is not a buzz word. It's a very real solution to many of the problems faced by enterprise applications and infastructure. Combined with .Net can be used to aid in providing web services, giving you One Degree of Seperation for your customers. Combined with XML and RDBMS, clusters can provide you with the speed you need for your enterprise software, and the 5-9s reliability that you require, without vendor lock-in.

        In short, clustering is the solutions for tomorrows problems today.

        • Clusternig is not a buzz word.

          where am dat warty melon

          8)
        • Clustering most definitely is a buzz word, Like 'OO' and 'XML' before it. Sure there is real technology behind it, but the term has been pretty much overused to death.

          The first thing that you have to ask when someone says cluster is if they mean a high peformance or high availability cluster (alternatively if this is a computational, load-balancing, or failover case).

          Oh wait, re-reading your MS marketspeak flack, I fear I've been trolled. Let me put it this way: Clustering is only the "solution for tomorrow's problems today" if you're talking about vapourware like .Net - for the rest of us, clustering has been around for years, long before it became a buzzword.
          • First of all OO or Object Oriented is not a buz word... it's a programming design/structure.

            and I certainly don't read about OO in advertizements for the moron public (read: Managers).

            XML is also a standard.. but yes it's a buz word now... which annoys the crap out of me.

            HOWEVER clustering is complete meaningless.. and has yet to even have an actual definition.. WTF IS clustering!? Don't tell me it's just a bunch of servers that sit in the same room.

        • Then what is it? It's not in any of my dictionaries...

  • As seen on the openMosix board... (Score:4, Funny)

    by deadhammer ( 576762 ) on Saturday November 09, 2002 @01:26PM (#4632863)
    Somebody should post this to Slashdot.org: How to assemble your own game server with openMosix

    The fools! Muhahahahahahaha!!!

  • Careful what you wish for... (Score:5, Funny)

    by Queuetue ( 156269 ) <.moc.liamg. .ta. .euteueuq.> on Saturday November 09, 2002 @01:29PM (#4632879) Homepage
    Moshe said:

    Somebody should post this to Slashdot.org: How to assemble your own game server with openMosix

    BTW, has anyone else seen the AppAssure soft-core porn ads here on slashdot? Better than the .net ads, by far! :)
  • but does this really count as ONE BIG SERVER for each game? if not, what is the advantage to having a single-entrance point of failure for the whole lot, when you could just use multiple independent servers?

  • Why not make ONE game... (Score:4, Interesting)

    by SexyKellyOsbourne ( 606860 ) on Saturday November 09, 2002 @02:02PM (#4633018) Journal
    That uses all CPU resources of a cluster, and leap 10 years into the future?

    The AI would use genetic algorithms, the sound would have every echo and diffusion effect possible, the graphics would use real-time raytracing, and the level count would be as extreme poly as possible. Simply spare nothing when it comes to CPU power, and just let it fly.

    Just use nothing but outright raw CPU power to render the whole thing.
    • Just use nothing but outright raw CPU power to render the whole thing.

      Yeah, 'cause, ya know, most games these days are either powered by combustion engines or mules.
    • Except you'd need like 400 nodes for it to run at some 30 fps.

    • Re:Why not make ONE game... (Score:4, Interesting)

      by Screaming Lunatic ( 526975 ) on Saturday November 09, 2002 @02:24PM (#4633080) Homepage
      That uses all CPU resources of a cluster, and leap 10 years into the future?

      The same reason that game companies don't make games for machines with 8 CPUs or 4 CPUs, or even 2 CPUs. People just don't have machines out there to play the game.

      You can do it server-side, because some geek just has to put together a bunch of computers. On the client-side, can you imagine that 12 year old that keeps fraggin you when you're playing RtCW online putting together a cluster of computers? (That was rhetorical)

      The AI would use genetic algorithms

      "True to life" AI, does not imply good "Game" AI.

      the sound would have every echo and diffusion effect possible,the graphics would use real-time raytracing, and the level count would be as extreme poly as possible

      The network latency would kill you. Motherboards are moving to AGP 8x so that they can get more bandwidth to the graphics card.

      • Genetic algorithms (Score:2, Informative)

        by jetlag11235 ( 605532 )
        "True to life" AI, does not imply good "Game" AI.

        Genetic algorithms do not imply "true to life" algorithms. They are merely a general method used to search for good solutions in difficult optimization problems.

        -- jetlag --
    • >That uses all CPU resources of a cluster, and leap 10 years into the future?

      Hmm, that is a good idea. According to my back-of-the-envelope math, 10 years would be seven cycles of Moore's law. Which means we are roughly dealing with a 128 times increase in CPU power. A cluster computer will waste much of its CPU time, so let's say we'd need 256 modern computers to do what you want.

      Not something the average person could do, it is the sort of thing that a bored university student with access to a lab of computers could do...

    • That uses all CPU resources of a cluster, and leap 10 years into the future?

      Minimum Hardware Requirements:
      8 speed CD-ROM
      Direct3D 8 or later graphics accelerator card
      50 2.4GHz PCs (100 2.4Ghz PCs recommended)

      Actually though, it's a good idea for the server to d othis rather than the client. You don't want to be sending pre-rendered graphics over the network, but you ought to be able to take advantage of the large CPU power available for good AI and physics modelling.
    • That's certainly an interesting idea, if I get you right you suggest that someone should program that very resource intensive game and make it really good so that it could be ready to be played by power users, say, in 5 years and by ordinary people in 10. OK...

      Only problem is: bandwith. Games require lots of realtime data flow between components. It wouldn't even be possible to play a modern game that way (given that most calculations take place on the "client"). So you'd have all the processing power but no data links to fit all the information that needs to be exchanged between CPUs for this in order to work!

      But I think you could build a custom PC architecture where, say, about 8 machines are sharing a really fast bus. And each machine could have number-crunching PCI cards in it to give them an additional boost. That's more like it, but I digress...

  • Interesting (Score:5, Funny)

    by ekrout ( 139379 ) on Saturday November 09, 2002 @02:07PM (#4633037) Journal
    Here's a link (italian) to the pictures.

    I've always heard that a picture is worth 1000 words, but no one ever mentioned I'd have to translate them all into English.

  • Slashdot effect. OT (Score:2, Informative)

    by FreeLinux ( 555387 )
    The Slashdot effect is beyond "getting old" now. I'm sick of story after story that is inaccessible.

    It's just so lame now. Oh, by the way, posting is still broken.

    • Local Cacheing (Score:3, Interesting)

      by Catskul ( 323619 )

      How about: before posting we see if we can get permission to cache locally (on slashdot). Obviously this isnt always possible, but in the case of small websites it might be practical, and even polite to do.

      • Re:Local Cacheing (Score:3, Insightful)

        by DeadBugs ( 546475 )
        Even though Slashdot handles all of its users fine now. Serving up large files like images and video instead of just text would cause it to suffer its own effect.
    • It works like this:

      1. Story with links is posted.

      2. One thousand Steve Barkto [essential.org] set their bots to obliterate the sites linked.

      3. Steve Barktos then submit bogus and self moderated crap, including "slashdot effect complaints".

      In this case, the link was quicly mirrored by someone with VA System like networking bandwith and ability to kill bots. It must be tough reading Slashdot from a Microsoft owned IP address .

      I'd love to see Slashdot prove my assertions, just as much as I'd love to see all of these threads modded to -1 off topic. It gets in the way.

  • phew (Score:3, Funny)

    by shimbee ( 444430 ) on Saturday November 09, 2002 @02:24PM (#4633081)
    good thing it's in italy and not in greece [theregus.com]. gaming times 1000 counts of conspiracy would be bad ;).

  • Migrating processes (Score:2, Interesting)

    by tribes ( 613022 )
    We toyed around with openmosix, "borrowing" some hardware at work. It was interesting to see what types of processes would be migrated to other cpus and which ones were simply not movable. We ended up running distributed.net threads until our hardware hijacking was discovered (hard to miss ~30 2RU servers). I would guess their tweaks facilitate easier process migration?
    • What was the point of running d.net in a mosix cluster? You're not going to see any performance gain like that.
      • Clustering is not necessarily used to get performance increases. We're looking at using openMosix to make better use of a room full of lab computers. Certainly during the night their CPU's are completely unused. Getting processes to migrate to them and make use of the CPU is the goal NOT to gain some huge parallel computing benefit. The original poster might not want to get higher performance but rather to "steal" cycles from idle desktops, that otherwise would go to waste.
      • Nope...we sure didn't, d.net wasn't the only thing we toyed with. We played with povray and such, but we mainly just wanted to watch the process migration, d.net offered the bonus of helping out a worthy project oncd we got bored playing with the cluster and were just going to let it run. After all, once we cabled and powered that number of boxes, really should use them for something semi-productive, no?
  • Why dosen't /. show cached pages, we get the point... offering cached pages would sure be nice.

    • Because, well, then, they'd /. themselves?

      • Id every article was cached slashdot would probably slashdot itself, but if you are only casheing one or two frond page posts links (ie only cache the links to small sites that cant handel the load) then it shouldnt be a problem. Remember September 11th when alot of other sites were in deadlock because of the mass of traffic. Slashdot was handeling the load just fine. I think Slashdot has a pretty meaty back end.
  • Do they actaully treat it as one node, and let mosix take care of everything....
    or do they use it as a normal cluster, running different servers on different nodes, and just have mosix there for the hell of it.

  • are there tutorials out there on this kind of stuff? this seems very interesting to me and I would like to play around at home with a few machines I have.
  • I wonder how hot and loud six Athlon XPs get. Seeing as how MY box is about -30dB of noise, with ONE Athlon XP, I can't imagine what this would be like. Then again, it might be good if it's an aluminium box -- they could just put a coffee decanter on top of it for those long nights of fragging.
  • This thing must eat up a whole bunch of power that w/ six PSUs there.
  • None of our men are "experts." We have most unfortunately found it necessary
    to get rid of a man as soon as he thinks himself an expert -- because no one
    ever considers himself expert if he really knows his job. A man who knows a
    job sees so much more to be done than he has done, that he is always pressing
    forward and never gives up an instant of thought to how good and how efficient
    he is. Thinking always ahead, thinking always of trying to do more, brings a
    state of mind in which nothing is impossible. The moment one gets into the
    "expert" state of mind a great number of things become impossible.
    -- From Henry Ford Sr., "My Life and Work"

    - this post brought to you by the Automated Last Post Generator...

Slashdot Top Deals

Last yeer I kudn't spel Engineer. Now I are won.

Close