| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
Comments:
<martian67> i like cli
<theUg> Dagmar Lokey probably all of us were n00bs once.
<martian67> Dagmar, why?
<n0dl> has anyone eher built a beowulf cluster?
<martian67> its no longer experimental
<tkup> Dagmar, martian67 n00rul , host2, which has a wireless card, connects to the internet just fine. It's host1 that's not getting through although it can ping the AP's internal gw address and all other hosts, except the itnernet.
<martian67> n0dl, it can be insanely easy
<Lokey> i wouldnt trust writing locally to ntfs : now over the network i have no problem
<martian67> there is even a livecd
<martian67> that can take any computer booted
<Dagmar> kokoko1: Honestly, you should really take into account which is more important... A few seconds of CPU time, or you wasting the rest of your evening obsessing over how to fix something that can be fixed in a different way in a matter of minutes
<n0dl> martian67: can be?
<theUg> By the way, no one was talking to Dagmar, yet he again turned to trolling.
<martian67> and auto add it to a local cluster
<n00rul> tkup: let me know the route table of host2
<n00rul> tkup: sorry host1
<Dagmar> Lokey: Until I see the caveats about what you can write with NTFS go away from the Help page in the kernel, I'm just not trusting it
<theUg> martian67, and again, if it can me done, is it a security issue? For any OS on a mchine with CD?
<theUg> *be done
<n00rul> tkup: It should be like this. Your host1 should send the packets to host2.
<Lokey> Dagmar: i agree w. ya on that one : then again, why would someone want an ntfs drive in a linux box ;)
<martian67> well obviously it is
<n00rul> tkup: I mean internet packets
<n0dl> martian67: hmmm i c... does it have to be redhat>?
<kokoko1> Dagmar, but in this case i have to loose the cache build aren't I ?
<martian67> if you are a compentent admin you know what to do about it
<tkup> n00rul, it has 192.168.2.0 via 192.168.3.1 dev eth0; 192.168.3.0/24 dev eth0; default via 192.168.3.1 dev eth0
<martian67> n0dl, not something you really want to ask me
<martian67> try Dagmar
<kokoko1> So why i go for the solution which cost cache lose to me?
<n0dl> Dagmar: have you built a cluster?
<tkup> 192.168.3.1 is host2's interface
<martian67> kokoko1,
<martian67> he has explained it multiple times
<martian67> if you dont want to use it
<martian67> go away
<martian67> get help from somone else
<martian67> QUIT BITCHING
<tkup> n00rul, I can ping all the way to the AP 192.168.2.1. but not beyond that
<kokoko1> martian67, Thanks mate
<n00rul> tkup: wait
<Dagmar> n0dl: Yes. Do I have one now? No. Don't need it.
<Lokey> tkup: why a x.x.2.1 addy?
<Lokey> shouldnt it be 1?
<n00rul> tkup: why this subnetting
<Dagmar> kokoko1: You already basically destroyed the cache you have now
<n00rul> tkup: you n/w setup looks strange
<theUg> martian67, I just doodle, speculate that is. What would be common hardware/software data protection measures?
<Dagmar> tkup: Then your AP is likely the problem
<n0dl> Dagmar: i c... is there a specific distro to run a cluster?
<martian67> theUg, disabling it in bios
<martian67> p***word the bios
<kokoko1> Dagmar, actaully I have kept the old swap.state while in case of fallback option :)
<martian67> easy enough
<n0dl> Dagmar: where can i find more info (i tried to google and got some stuff)
<n0dl> besides the official site
<martian67> n0dl, there is a bootable live cd
<martian67> that auto-sets it up
<theUg> martian67 just as easy to reset CMOS?
<n0dl> martian67: i presume you mean clusterknoppix
<Lokey> n0dl: poke around Portland Oregon State Uni's site : you might find something on their 1/4million beowulf cluster
<Dagmar> n0dl: There's one distro specifically for clustering but I can't rmemeber hte name of it right now. Most people I know of use Mosix most of the time
<martian67> theUg, and how are they going to do that?
<martian67> n0dl, yes
<theUg> martian67 open up the case, and move a jumper, innit?
<Dagmar> Lokey: They have 250,000 nodes there?
<n00rul> tkup: What's your host1 ip
<martian67> theUg, if you have public computers
<martian67> you are either gonna supervise them
<Dagmar> Lokey: We're just pushing close to 800 nodes here
<martian67> to make sure they cant be opened
<martian67> or lock them up
<Lokey> nah Dagmar : its just a quarter million $$ worth of hardware Intel gave em
<martian67> if you dont... you are a fool
<Dagmar> Lokey: Ah, okay. I think we're probably over that amount by far now. Heh
<kokoko1> Dagmar, okay know i'm going to restart the squid with the backup copy of swap.state file, coz reading from the mailing list explains its not bad to get ..
<kokoko1> 2003/02/25 10:16:10| storeAufsOpenDone: (2) No such file or directory
<kokoko1> > 2003/02/25 10:16:10| /cache/0F/6A/000F6A8F
<Dagmar> There is a point at which parallel computing clusters seem to reach a critical m***
<n0dl> aaa i c openmosix
<n0dl> interesting
<Lokey> scary thing was that Intel DONATED it
<kokoko1> "Maybe one could state that aio is not 'cache safe', but the thing
<kokoko1> is that squid will fetch the object again from the remote webserver. "
<Lokey> right after they laid off like 4000 people
<martian67> kokoko1, then go try it
<theUg> martian67 Well it is pretty obvious. But it would be curious to see if there is some ingenous software solution
<martian67> Dagmar already explained why its a bad idea
<Dagmar> kokoko1: You're going to be ignoring the rest of those "not found" messages then because your cache index isn't going to have info about files added after the backup was made, nor will it know that some items have expired and were deleted.
<martian67> but if you are going to sit in here and bitch and bitch and bitch
<martian67> about why Dagmar is wrong
<Dagmar> kokoko1: Just delete and reinit the damn cache and be done with it
<martian67> you arnt helping anyone
<PeanutHorst> martian67, help!
<n0dl> how do i runa .run file?
<martian67> lest of all youself
<martian67> PeanutHorst, hmm?
<PeanutHorst> hilfe! as the Knoppix maintainers would sayh
<Dagmar> n0dl: chmod o+x it and then run it or us something like sh blah.run
<Dagmar> s/us/use/
<theUg> martian67, say like custom mount p***word
<PeanutHorst> http://ftp.knoppix.nl/os/Linux/distr/knoppix/ <----- which is the CD folder? I can't burn DVDs
<bleenshaker> martian67: dagmar cant even sublimate the concept of 1+1=2 :)
<n00rul> tkup: Is the issue solved?
<Lokey> 0+1=2
<martian67> 1+1=10
<Dagmar> PeanutHorst: Don't make me come over there and click the mouse for you
<martian67> heh :)
<lmm> events/0 has just started eating all my cpu. Anyone have any idea why this might be?
<martian67> PeanutHorst,
<martian67> its just the .iso
<martian67> files in that dir
<martian67> remeber EN = ENGLISH
<Dagmar> PeanutHorst: You're IN the directory with the right ISOs dude
<Dagmar> lmm: COMMUNISM!
<PeanutHorst> don't make me use this........
<martian67> omg communists eating our megahurtz!!!
<martian67> PeanutHorst hmm?
<lmm> Dagmar: so I need to have a revolution and, erm, free my workers?
<Lokey> 500 PeanutHorst ?
<martian67> dont know how to burn an iso?
<Dagmar> lmm: Perhaps. I'd just give it a reboot. The events kernel thread is a bit new so I'm basically in the dark about what the hell it does
<lmm> Dagmar: I do reboot, it still does it. Any way I can leave it out?
<martian67> Dagmar, well you are further ahead then everyone else in this channel :)
<Dagmar> lmm: Disable pre-emptive multitasking maybe?
<theUg> martian67 It is just funny how with all that technological mumbo jumbo it all comes down to a sentry. That very human factor we always trying to remove from equation, is the only thing that ultimately protect us.
<kokoko1> Dagmar, nice taking with ya
<kokoko1> talking*
<Dagmar> lmm: Lemme see if I can Google and find out what it's tied to
<lmm> Dagmar: thanks
<PeanutHorst> fsck fsck fsck .... where is the fscking latest bleedin'-edge beta ISO?
<Dagmar> kokoko1: Sorry I could't be of more help but honestly, squid is ill-prepared to deal with unexpected damage to it's cache indexes, and generally it's just way faster to wipe 'em all and start over unless you're trying to hunt down an actual bug that causes the cache to become truly out of sync
<Dagmar> The *occasional* message about objects being missing is fine and just a small race issue
<martian67> hmm
<martian67> damn modem
<PeanutHorst> martian, i repeat
<PeanutHorst> fsck fsck fsck .... where is the fscking latest bleedin'-edge beta ISO?
<theUg> PeanutHorst, what happened to the good old ****?
<PeanutHorst> theUg, lilo banned it :(
<martian67> PeanutHorst, look hard :)
<Dagmar> lmm: I'm seeing signs that it's possibly the part that deals with USB/hotplug events
<theUg> Ooh, sorry :)\
<martian67> its the highest version number
<theUg> Nice to know. :)
<Dagmar> lmm: Try unjacking your USB stuff (except maybe your keyboard and mouse that you might well be needing still. heh)
<kokoko1> Dagmar, no mate you did well today and i learn lot of new things from you :)
<PeanutHorst> 4.0.2
<PeanutHorst> that isn't beta
<martian67> PeanutHorst, no idea
<Lokey> PeanutHorst: why do you wanna be on the bleeding edge : typically it cuts you along w. the bleeding
Return to
#linux
or
Go to some related
logs:
python urllib3
#linux
multipart/form data not allowed
#bash
human gtk1.x theme
#web
Xm/PrimitiveP.h: No such file or directory
#lisp
#perl
mydsl telnetd install
|
|
