Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 


Forgot your password?
Close
typodupeerror
×
Slashdot.org News

We Are Experiencing Technical Difficulties 146

Posted by CmdrTaco from the please-stand-by dept.
So without a doubt, the best time for the power supply in your server to die is when you're out of town. Oh, and the line to your boxes at home should die too. And the only machine with a working modem in your hotel room should be an NT box (without ssh installed) and then the connection you dial through should be 19 hops away from anything (routing from San Jose to NY, DC, Boston, and back to Frisco and conveniently losing almost all of the packets) Anyway, we're back up and kinda hobbling now, (thanks to Jesse & Dan and UP Networks for being jonny's-on-the-spot) but I'm trying to fix some stuff as fast as I can. In the meantime, things are gonna be a bit zany, so don't flame me to loud. And don't worry, we've been working for the last month to build a new system with redundancy and stuff so that this won't happen again (knock on wood).
This discussion has been archived. No new comments can be posted.

We Are Experiencing Technical Difficulties More

We Are Experiencing Technical Difficulties

Comments Filter:

  • Um ... no. You get what you pay for. (Score:1)

    by Anonymous Coward
    A good dual 400 watt or dual 600 watt 2x unit will run you $900 to $1200. That is for seriously fat watts. I usually get stuff from Magtech out in El Paso (where CalPC gets their power supplies made). More than seven years, and two total failures, out of 600+ out there working hard with (mostly) BSDi and FreeBSD (I am sure that Linux would stress them as well). They are supposed to be that expensive, they are supposed to weigh 15+ pounds each, and they are supposed to have hot-swappable redundant fans. The new ones have serial jacks to let you keep an eye on them. Properly made by a solid manufacturer, power supplies almost never fail. You get what you pay for. If youy got a deal, it was used or you didn't. Example: 386/16, running continuously for almost ten years, almost non-stop. First Interactive, then SCO, now FreeBSD. No failures. Supported a large array of mirrored SCSI drives that had to power up sequentially. It is now my firewall/proxy. The power supply is huge. It rode out power glitches that nuked everyone else's system in the office before we rewired and added some SOLA units. Example: 486/100 with 5 SCSI disks and 128MB RAM, serving about 600 people. One really large power supply -- took up the whole top of the case. It is now on my own LAN, serving my office. No problems. 400 watts rated. Example: 300 watt power supply running the main web server for a friend's ISP. Build on a budget, got used. Survived a lighting strike. Still running. Went from a P90 to a k6-3/400 and 768MB of RAM and 10k drives, and no problems. Standard AT size, weighs 10 pounds. I could go on, but I hope you get my point. I have always spent a lot on power supplies (I have always seen them as the heartbeat of the system) and I have never, save 2 times (one Dr. Pepper, one a major electrical "problem" due to a stoned electrician who auto-Darwinated in the process of taking out a whole shopping center's computer systems and POS systems) have I had trouble. Don't see it? Open up one of those bargain power supplies in a $39.95 case and look at the wiring. Now open up (if you can find one) a good power supply. Notice the difference? Here's a nickel kid. Get yourself a real computer.
  • I'm sure the fine folks who house slashdot's server know how to test and diagnose a broken power supply. If Rob says it's broke, I'd bet that it's broke.
  • Number 3.14: 'Walk' your server down the hall, from outlet to outlet to outlet...

    (also good if you have redundant ethernet)
  • Let's see if Andover will spring for some real hardware...

    I can highly recommend the Compaq Proliant 7000...hot-swap redundant PSU's hotswap RAID in hardware, hotswap PCI slots, integrated hardware monitoring and alerting . . . and a groovy LCD panel on the front...

    You never realise the value of hot-swap mirrored drives until you're demoing a front office trading system to the powers-that-be in the company taking yours over (the 'lets try and keep our jobs' demo) ...when a drive fails in the SQL server. One quick (and hushed) call to the DataCentre, and a new driver goes in and gets built on the fly...users never noticed a thing...

    I think I aged 5 years in 10 minutes...

  • ...why are you suggesting an Intel based box?

    Only 'cos I have real-world experience of this kit saving my arse one or twice, and saving our company several millions of dollars...

    If you want to send some other eval kit that works as well, I'll sing its praises too...

    Regarding hot-swap, it's the hardwares job to hide it from the OS. Example - I can walk up to our production Proliant 7K, pull one PSU and 7 drives from it, and not one of the users will notice the difference (although Data Centre security might give me a hard time). Hot-swapping a PCI network card is a litte more work, as you have to use a provided utility to down the power to the slot before you yank the card. And all this under NT too...

    The only thing the OS needs to know about in the hot-swap world is the CPU's. I don't know of anything that will support CPU hot-swap...yet.

  • That doesn't mean you can put mission critical systems on any commodity Pentium box. As someone mentioned, cheap pwer supplies and IDE drives are no way to run a non-stop sight.

    Sun hardware (with the exception of the 5/10, heh) is built very solid. They offer redundant power supplies, RAID built-in among other nice features.

    It's not Linux so much but really the choice of hardware. You get what you pay for.

    It's also dependant on your design. If you need 24/7 service then you design two or more of everything into the mix.

    I'm setting up a nice site with two E250's with redundant power supplies, RAIDs and Resonate. I don't want to be called at 3am... :)

  • Why don't we create a section called slashdown or something, which posts nothing but downtime news. It would definitely fill in on the weekends when only 1 or 2 stories go on the main page. Then again the amount of downtime news would crash the server. Luckily there's msn.com, which never goes down.
  • Sounds like /. doesn't have much fail over. Perhaps it's about time to test Andover.net's commitment to /. and get some more hardware, etc.
  • ...since he did not specify which century, he could be right. Of course you'd have to assume that the century that is the topic of discussion started on 1900/01/01 or somewhere around that date.

    Of course if the particular century in question began with 1901/01/01, he'd be wrong or if he's talking about the "20th" century he'd be wrong, or if he's talking about the century that began today, well he'd be wrong again. This would of course be the first Friday 13th of *that* century, which means that it can also be the first of *this* century.

    Extrapolation from century to millennium is left as an exercise for the reader.
  • Maybe the problems are related to the record ~1300 commments for the Kansas - education - Bible-thumper - evolution - creation - everybody's pet hot button - jaw-dropping incredible - coming out of the woodwork story?
    --
  • You should definately consider that Rob. I work for the company that provides hardware and connection [dedicatedserver.com] to Andover.net (and now freshmeat) and I can say we'd love to have you aboard.
  • Shame on ya Rob. With all that Linux hardware about and you had an NT box lying about.

    Shameful... ;-)
  • "Whatever he did" was to have the server go down so nobody was viewing any pages or taking up any bandwidth; I'm sure that it'll get nice and slow again once everyone realizes that it's back up.
    ---
    "'Is not a quine' is not a quine" is a quine.
  • "So, what do we call the years between 1 AD and 1 BC?"
    Don't know about you, but I personally call 'em non-existent.

  • "No problem"
    Gee, I'll bet that's what e-Bay said too.

  • As I learned (much to my embarassment) use of the term "Frisco" pretty much marks you as a tourist in the bay region. It tends to grate on the locals' ears. "'Cisco" tends to be better received. However, since the context was packet routing, every packet passed through [a] 'Cisco [router] several times, at least (haw haw).

    "That's the great thing about kill-bots, you can always make more."

  • With a little clever scripting, you CAN do this! I did once (of course the box I was configuring was only across the room -- I suppose that's rather tragically lazy). Basically, whip up a shell script that runs your new ifconfig command, then sleeps for a minute or so, then ifdown's and ifconfig's back to the old (known working) command. If the new command works, your connection comes back up, and you just log in and kill the script and go on your merry way. If it fails, it'll be back up in a minute. Pretty nifty.
    ----------------------
    "This moon-cheese will make me very rich! Very rich indeed!
  • pretty neat ;)
  • Of the few SSH clients for windows, my favorite is putty. Single exe file, nice terminal. No port forwarding, but what can you do? The first site that I found with it when I type "putty ssh" into a search engine is:
    http://www.sorted.org/~chris/ssh/
  • <Sigh.>

    It's really a pity that something that just had to have been intended as humor was marked down as flamebait.

    Blargh.
  • Can you PLEASE turn off this "Slashdot overload" function??!!
  • I'm not if anyone noticed yet.
  • Yep. Slashdot never really was an shining example of stability and availability...
  • No , my ISP is using Solaris ... but I think the problem is that ./ is _not_ using something like Solaris. Don't you think ?

  • 13 (Score:1)

    by mcc ( 14761 )
    heh.. maybe the slashdot servers refused to go up on a friday the 13th, so rob had to set the clock forward a month so it would be a monday?
  • mail.earthlink.net has also been down for most of the day.
  • Without much installation or downloading, there is a freeware application out there called PuTTY.EXE [greenend.org.uk]. Enjoy.
  • btw... this is a SSH/Telnet client.
  • I've been running 2 and 3 power segment redundant power supplies on 4 server farms for the last few years and had one fail for the first time only recently.

    A nice non-tech fellow from accounting was interested in checking out our premise equipment---just out of curiosity. Toward the end of the tour, he tripped on a slim fiber bundle behind a relay rack after answering his cell phone. In his other hand was a tall 20+ ounce styrofoam cup filled with tea. Of course, it's contents hit the rear of a Dell PowerEdge 6300 (running Oracle and all of great importance) alarming two units in the triple-redundant power supply. The server ran on the third. The other two were destroyed but later replaced hot. Slide in the new and press a button.

    No kidding!
  • Of course, one could also interpret this to mean that if you can maintain a Linux system from NT, you can maintain it from anything.
  • >The last total solar eclipse of the centry
    >(notice I didn't say 'millenium', is there one
    >next year anytime?)

    You should be aware that the next century begins
    On Jan. 1, 2001, for exactly the same reason
    that the next millenium begins on the same day.

  • here i was thinking that slashdot got slashdotted...silly me
  • You might have a hard time finding someone who got in on Red Hat's IPO invitation. Rumor has it they're all pretty rich now..all three of them. ;)
    Bowie J. Poag

  • Hehehe.. dont worry, we still love you, Rob. :)

    Bowie
    PROPAGANDA [system12.com]
    Bowie J. Poag
  • when my ISP or a site I visit regularly like /. goes down for a day or 2, it rarely bothers me. "It'll be back up in a day or 2," I say to myself. No problem.
  • _my_SYSLOG {
    /mnt/dskPic 100% full;
    /mnt/dskEverything 100% full;

    badly written code trashes database files;
    users hysterical;
    need beer;
    exec rm -rf;
    return "need_more_space" }
  • Personally I use mindterm, a java applet that works as a ssh terminal. It's great. Search for mindterm on altavista to find it's homepage. Or have a look at it on http://ion.ath.cx/ssh (Sorry, you can't even try to login at the moment because I'm in the middle of a big restructuring, and there's no sshd running...).
  • Blame it on thunderstorms.


  • I know it might be too late to cry over spilled milk, but way back when, vendors started offering a very nifty thing called a "redundant power supply" or "n+1 power"... This would have prevented all these problem.



    Sun Tzu must have been running Linux...
    - Hold out baits to entice the enemy. Feign disorder, and crush him. (Sun Tzu, The art of war)

  • Happy Friday the 13th, everybody...
  • though you notice the ad banner at the top still works... nice to know that though slashdot may fall, capitalism still lives..
  • 1999 is part of "this century" as well as "this millenium". So is 2000. Jan 1, 2001 is the first day of the new millenium.

    He said that today was the last Friday the 13th of this century. Since Friday, October 13, 2000 is a Friday the 13th in this century but is not today, that makes him wrong.
    ---
    Put Hemos through English 101!
    "An armed society is a polite society" -- Robert Heinlein
  • October 2000 has a Friday the 13th.

    However, today IS the last Friday the 13th in "the 1900's".
    ---
    Put Hemos through English 101!
    "An armed society is a polite society" -- Robert Heinlein
  • Weighing in at 198K binary, this [greenend.org.uk] is pretty easy to transport around. No port forwarding, but it will get you a shell.

    Oh, and it's illegal to use in the US without an RSA license. Damn patent laws.

  • The aversion to the use of the term "Frisco" is generally limited to those of us who were born here. Since a vast chunk of the City is now more than ever comprised of People From Somewhere Else -- of a beautifully dizzying diversity -- said aversion borders on historic.

    It annoys me, but I never correct anyone.
  • The @Home network? The only reason the @Home network has LOTS of hops is that they're running optical stuff (packet over DWDM), and with that, as opposed to ATM-type links, EVERY link is visible in a traceroute.

    BTW, traceroute to Slashdot from @Home:
    Tracing route to slashdot.org [206.170.14.75]
    over a maximum of 30 hops:

    1 27 ms 30 ms 3 ms ottawa-fe.net.rogers.wave.ca [24.112.33.1]
    2 16 ms 35 ms 24 ms 10.0.184.53
    3 41 ms 22 ms 30 ms 24.2.9.9
    4 15 ms 26 ms 27 ms c1-pos9-1.bflony1.home.net [24.7.72.245]
    5 17 ms 29 ms 33 ms c1-pos1-0.clevoh1.home.net [24.7.65.5]
    6 28 ms 30 ms 30 ms c1-pos5-3.cmdnnj1.home.net [24.7.67.158]
    7 33 ms 29 ms 30 ms c1-pos1-0.washdc1.home.net [24.7.65.85]
    8 35 ms 32 ms 56 ms bb1-pos2-1-0.mae-e.nap.home.net [24.7.72.54]
    9 72 ms 72 ms 43 ms mae-east.ibm.net [192.41.177.110]
    10 59 ms 66 ms 52 ms beth1sr2-11-0-0.md.us.ibm.net [198.133.27.10]
    11 68 ms 71 ms 60 ms bethjbr1-ge-1-0-0-0.md.us.ibm.net [165.87.29.122
    ]
    12 106 ms 95 ms 93 ms sfra1br1-at-2-0-0-2.ca.us.ibm.net [165.87.230.98
    ]
    13 121 ms 89 ms 91 ms sfra1sr3-so-0-0-0-0.ca.us.ibm.net [165.87.13.30]

    14 101 ms 112 ms 93 ms 165.87.161.73
    15 97 ms 122 ms * ded1-fa11-1-0.snfc21.pbi.net [209.232.130.4]
    16 113 ms 95 ms 92 ms 209.232.138.214
    17 111 ms 92 ms 92 ms slashdot.org [206.170.14.75]

    Personally, I'd say the reason for LOTS of hops is because Slashdot uses PacBell's backbone, which seems to ONLY talk to the world via IBM, etc...
  • Interesting!
    Does it beat teraterm + ttssh?
    (Not as though I use that these days. I'm almost totally linuxed now, just the occasional box out there that it helps to have slightly secure access from, eg at home...)
    ~Tim
    --
  • I dunno, but the person that copyrighted/trademarked "01-01-00" was an idiot, IMHO. If we take the real millienium (jan 1, 2001), it turns into the nicer string "01-01-01". We should trademark/copyright/patent/etc that >G

    Odd fact: 010100b = 20d, and 010101b = 21d. Which reflects the ages of certain people I know. kewl.
  • I guess this would be the wrong time to ask for a new poll. *sigh*

    J:)
  • Dude - the century and the millenium will be ending at precisely the same time.

    BTW, this is the last total eclipse of either.

    Tim

  • PuTTY is nice-just started using it, and it is pretty neat. Univ. of Chicago is using a java applet for web based ssh-I'm not sure exactly how it works, but it is a neat idea, and I believe it is available on the web. It's the mindbright SSH Java applet, http://www.mindbright.se/english/ has more details. It's worked pretty well for us, although you do have to have the latest flavor of Java for it to work.

  • Date- (Score:1)

    by zanzar ( 33471 )
    "Posted by CmdrTaco on Monday September 13, @12:23PM EDT"

    Anyone else notice this? :) September? Cool... Temporal distortions on slashdot...


    -Larry

  • Hey.. (Score:1)

    by zanzar ( 33471 )
    Now if someone had just hooked me up with some redhat shares, I'd be in the position to donate some hardware to you....
    hehehe

    -Larry
  • This is the problem with flame wars - people say things just to make the other angry. How can you say that this guy was part of a group that critizises but never stands behind his words? For all we know, he could be running a quadruple-redundant server with 10 different leased lines and 4 different power supplies. And for that matter, what part of what he said made him an incompetant IT?

    Okay, so you don't like what he said. Fine. Agree to disagree and be done with it. Flames make me sick.
  • You can set that in your user preferences.
  • Absolutely - I agree also. However, I would strongly suggest that drinking fluids _NOT_ be allowed near the back of servers sufficiently important as a PE 6300. (Of course, I'm sure you guys do this now...)

    --bdj

  • Most standard P/S's that come with cases are _quite_ crappy. For servers and high-availability machines, I use PC Power & Cooling [pcpowercooling.com]'s products - they've been around forever, and have a very good reputation AFAIK.

    --bdj

  • I was trying to make some updates to my ethernet card, and remotely did an ifconfig on the card I was connected to. Something didn't work and I lost the connection.

    Oh well, the machine is only twenty miles away, and I have to now take a trip. It was a good try though:)

    Lesson: Don't remotely configure the ethernet card that you are connected to!

    It worked once before, honest!
  • Of course, mine wasn't [computer] hardware problems, but try having the power go out as soon as you get up, be off for 2 hours, come on for 10 mins, go off again for approx. and hour, and then come back on? Besides having a laptop monitor that's spazzing on the video (at some angles, the screen won't display.. but move it a bit, and the picture comes back). And a few other minor details.

    But hey, getting back to topic here, tis ok. Stuff happens.. hee hee hee. and always at the most convenient time(s).
  • Well, actually, it's really 2035, if you want to get technical, but then you'd have to admit that this is a religious calendar and thus deeply flawed.

    So, what do we call the years between 1 AD and 1 BC?

  • Yeah, if it weren't for the SEC, it probably would have been 3, but I think it's up near the 8000 mark thanks to them.

    Not that I've got any shares yet ...

  • Hmmm. I wonder ...

    How about the question noone wants to ask? Like: Who do you blame for the Red Hat IPO disaster?
    A. Red Hat
    B. E*Trade/E*Offering
    C. SEC
    D. All of the above
    E. Y2K
    F. Will in Seattle
    G. The penguins in the cage

    I vote for B. But I'm not sure about those penguins ...

  • H. The Lunar Eclipse (or was that the Solar Eclipse? Not sure which it was)

  • What do you mean, you're not? You should still have a hangover ... it's only Friday when you posted.

  • Really, it's 24/7.

    C'mon, they guarantee 95% uptime.

    Isn't that enough?

    ;-)

  • It's funny how it's ALWAYS the power supplies that go bad first on Intel boxen. Assuming the PC components survive the first 3 months of their lifespan, the next thing to need replacing (within 2 years) has always been the power supply. Why is this?
    Is it just the heat that limits their lifespan? Or are power supplies used in most Intels just crappy??
    I once (1 time only) bought a US made PowerCool 300 watt P/S for $150. It's lasted quite a long time but I'm not sure that justifies the hefty price.
    I'm curious. What are other folks' experiences?
    Has anyone found a good reliable P/S for AT/ATX's??
    Is there some secret to it??

    (JTOL)
  • Comment removed based on user account deletion

  • Don't forget, OPIE has its problems too. Most notably, it is still vulnerable to connection hijacking, so frequent use of OPIE in a particularly hostile environment is generally a bad idea. Intelligently handling su in an environment with lots of root users is nontrivial at best. (Remember, any passwords sent over an OPIE connection can still be sniffed.)

    I'm not saying OPIE is bad, or that it shouldn't be used. I'm just saying it needs to be used cautiously.

    --
    Fourth law of programming: Anything that can go wrong wi

  • ...it wasn't my NT proxy server or my NT exchange server or my SWB-provided ISDN or my Bubba's Basement ISP. (Yes, I'm in hell. Another reason for the /. handle.)

    I *was* going to "Ask Slashdot" if nerds were superstitious. Hmmm...

    Hang in there guys, we've all been there, with the Great Unwashed standing at the server room door, saying; "Do you have an estimated time?" "Did you know that accounting's down, too?" "What's the problem?" "Should we call someone?" "Can I get my email?" "Should I call my customers?" "Do the phone lines work?"

    Mph!

  • Sorry about the nitpick...really.

    Although your use of double entendre is technically correct, it is, in most cases, used for more risque puns.

    Again, sorry.
  • You spent the night in a hotel with an NT box?
    You've got a lot of explaining to do, young man.
  • Oh, Bill, it's just you. LOL
    What a crazy little borg you are!

    The Divine Creatrix in a Mortal Shell that stays Crunchy in Milk
  • If Andover is going to spring for some "real" hardware then why are you suggesting an Intel based box?
    I would have to say go with Alpha, or UltraSPARC. Although a Quad-Xeon machine ain't bad, a dual-processor Ultra would probably be a whole lot better for this job. Some Ultras are hot swappable, but the problem with hot swap is whether or not the OS can handle it.
  • I couldn't get on all morning, but it wasn't a big deal as I was rather busy. However at about 10AM I set up a job to run, and settled in to read a little /. . Much to my surprise I got yesterday's edition. Everything was from about 4PM yesterday. Very odd, and when I refreshed to get the newest version, I got nothing. Odd, but I survived. I agree with past posters who were in favor of getting Andover.net to purchase some new equipment.
  • Your comment makes no sense. Who are you to bash anyone especially since you don't explain how the previous poster is a "luser"?

    It figures an anonymous bastard like yourself would post such trash.

  • I've hand many chances to play with machines with redundant power supplies. In two words, THEY SUCK! Failure rate on redundant power supplies is ENORMOUS. In most cases (double entendre intended) two power supplies must occupy the space normally inhabited by a single power supply. They fail an order of magnitude more often than large single power supplies. IMHO, your overall best bet is redundant servers. Linux doesn't have many built-in high-availability features but the High-Availability HOWTO [unc.edu] is a great place to start. But redundant power supplies -- PHEW! They stink!
  • If you don't already have it, you need it! :')

    http://www.lm-sensors.nu [lm-sensors.nu]

    You might have received an email warning that the voltages were beginning to go wacky. I found a dying power-supply and replaced it before services when down with this software. It can tell you when a CPU fan is slowing or stopped, or if things are getting too hot, too. Nifty!

  • You should be aware that the next century begins On Jan. 1, 2001, for exactly the same reason that the next millenium begins on the same day.

    I don't care if the millenium "officially" starts Jan 2001 or not. When it comes not partying this New Years Eve, I'm going to be thinking of other things besides calendar math.

    TheGeek
    http://www.geekrights.org [geekrights.org]

  • Is this why I haven't gotten any headlines mailed to me either? I'm not smart enough to get to the site myself. I need to be prompted!
  • Flamebait. Have a nice day. Not.
  • Nahh... If NT had not existed, that m/c would still be there but running a different OS. Therefore, NT had only an indirect effect.

    james
  • Funny, always has been for me...is your ISP using NT?
  • Who knows how much those ads cover? Is Slashdot turning a profit? Is it supposed to? I know kids in high school who have link exchange banners up on their sites and pull (very marginal) money. Does this mean they're not hobbyists? Ack...

  • Today is the LAST Friday the 13th of this century. Think about it. I think supernatural forces are against Rob!

    About this redundancy problem, anyone got any information? Like a book? We need to do that here.



  • Sounds like your packets are being routed through the @home network!
  • they said they are working on it, and to switch a complete production site is not that easy, and if I am not totally mistaken it is still partly a hobby site.

    Besides, when they start to charge membership fees you can start complaining, until then it is annoying but not much you can do.
  • Wrong? He didn't say "the 1900's", he said "this century". It's different.

    Also, he could have said "the millenium".

  • ack (Score:2)

    by drwiii ( 434 )
    And the only machine with a working modem in your hotel room should be an NT box (without ssh installed)...

    It pains me to see that Slashdot is back up and running thanks in part to NT.

    :-)

  • The aversion to the use of the term "Frisco" is a yuppie/upper class thing (Remember Herb Caen's book?). "Frisco" is A-OK among the blue collar/"Regular Guy" crowd.

    --
    Interested in XFMail? New XFMail home page [slappy.org]

  • Re:Date- (Score:2)

    by jd ( 1658 )
    That explains the run-down clocks, the meteorites and the Solar Eclipse. :) Ah, it was the Kansas story that did it! Mixing the old and the new causes breaches in time.

    Sapphire & Steel have been assigned!

  • Chill. I'm guessing /. is still "in trasition" to Andover, so multi-redundant backup and high-availabilty is still in the works. Si?

  • MCI WorldCom is experiencing some major network difficulties [techweb.com].
  • This lady walks into a bar and says to the bartender, "I'd like a Double Entendre."

    So he gives it to her.


    -Chris
  • Doh! Yep of course, you're right. Either not enough or too much caffeine is to blame. ;-)

  • Heh, beat me to it. :p

    Microsoft in danger of being ripped apart by the Government on anti-trust.

    A Linux distrubutor goes public and its stock increases like 600% the first week.

    The last total solar eclipse of the centry (notice I didn't say 'millenium', is there one next year anytime?)

    Forecast in hell next week: colder with possible flurries...

  • Well, if you measure the "millenium" to end on 12/31/2000, then how are you counting "centuries"? If you're going by the (erroneous) calendar with no year zero, then both century and millenium end on the same day. As for me, I prefer to go by the big high-order digits. 12/31/1999 is close enough for me... but then I'm something of a slacker. ;)
  • First off, finding any site(s) without banner ads of some sort is nearly impossible these days. Internet connectivity isn't free. Web hosting isn't free either -- and don't even get me started on those "free" sites that do nothing but litter one's desktop with ads...

    /. [slashdot.org] is (or was) a hobby site. As such, do you honestly expect people to fork out their (hard earned?) cash for redundant, "it will never go down", high availablity hardware? If you do, then you should seek professional help.

    Let us do some math in reference to banner ads and "hobby sites"...

    Let's make a few assumptions (alright, ALOT of assumptions.) Say your hobby site is fairly popular getting 100k hits per day with an average transfer of 8k per hit. <math> 100000 * 8 / (60 * 60 * 24) == 9.26K/s</math> That's assuming the traffic is perfectly evenly distributed -- which we all know isn't true.

    To continuously transmit 9.26K/s would require a 128k connection. A standard analog modem tops out in the 3K/s range. A single ISDN B channel tops out at 7.5K/s. A dual channel ISDN (2B) or 128k frame connection tops out at 15.2K/s. Let's assume you go with ISDN (it's generally the lower cost.)
    SO, you'd need an ISDN router (the first person to say "ISDN Modem" gets shot) at a cost of 700$US, an ISDN phone line at a cost of 100$US/month and 200$US installation, and an ISP at a cost of 150$US/month plus 200$US installation (what a rip-off.) That's 1100$US to start plus 250$US per month for your "hobby"... and you haven't even bought a machine to be your web server.

    Seeing as no one is being charged for access to /. [slashdot.org], no one has a leg to stand on if/when it fails. Face it, machines break. I think the /.'ers [cmdrtaco.net] put more than enough of their time and money into making this geek haven. If you cannot get your /. fix [slashdot.org], then that should be taken as an omen to get (back) to work.

    As always, Your Milage MUST Vary!

  • For those times without SSH...OTP. (Score:3)

    by Greg@RageNet ( 39860 ) on Friday August 13, 1999 @07:01AM (#1748394) Homepage
    Traveling occasionally and being forced to use other's machines (usually windoze boxes) I've found a good way to get in without SSH and still keep the crackers off my back. I've installed OPIE, a one time password login mechanism; in addition I installed PilotOTP on my Palm 3 which travels with me everywhere. TCP wrappers are set up to give local network users the standard login prompt while 'twist'ing the rest of the internet to an OTP login..

    hosts.allow has a rule for in.telnetd allowing local network standard access.
    in hosts.deny:
    in.telnetd: ALL : twist /usr/sbin/in.telnetd \
    -L /bin/login.opie

    Hope this helps!

    -- Greg

Slashdot Top Deals

Research is to see what everybody else has seen, and think what nobody else has thought.

Close