Slashdot Mirror


Surviving Slashdotting with a Small Server

S.BartFarst writes "Our little departmental server has been slashdotted twice in the last year and survived! Implementation of a two-headed redundant hardware scheme using linux virtual server and backup and failover capabilities enhanced by the linux high-availability tools has produced a nifty low-cost solution. Gotta love those little white boxes! (also having a university-supplied BIG PIPE doesn't hurt). More interesting is the documentation of the apparent exponentially decaying attention span of slashdotters. Anybody else observed similar phenomena?"

307 comments

  1. Third Time's a Charm? by sweeney37 · · Score: 5, Funny

    Our little departmental server has been slashdotted twice in the last year and survived!

    Wait... is this a challenge?

    Mike

    1. Re:Third Time's a Charm? by Joe+Jordan · · Score: 0, Troll

      ... begin repeatedly pressing the refresh button like a maniac..... NOW!!!

    2. Re:Third Time's a Charm? by Ninja+Master+Gara · · Score: 0, Redundant

      It's the S-Prize. Soon John Carmack will be in on it and it'll get huge computer-media attention.

      --

      ---
      When I grow up, I want to be a kid again.
    3. Re:Third Time's a Charm? by Anonymous Coward · · Score: 0

      Here's a big file to download.

    4. Re:Third Time's a Charm? by bad_fx · · Score: 3, Funny
    5. Re:Third Time's a Charm? by BWJones · · Score: 5, Insightful

      Our little departmental server has been slashdotted twice in the last year and survived!

      Oh, come on. Even my little old G3 iMac is capable of handling quite a load from Slashdot and this site is serving up graphics intensive stuff. What you need to prevent a good Slashdotting is bandwidth that universities provide. T3 backbone connections are a wonderful thing. :-)

      Go ahead click all you want.

      --
      Visit Jonesblog and say hello.
    6. Re:Third Time's a Charm? by Anonymous Coward · · Score: 3, Insightful
      a friend of mine had his site featured on the front page of slashdot without any problems. Page hits spiked for a couple days, but his setup (P4 x86 linux box, apache, busines DSL) handled it fine. And, after being hit he discovered his bandwidth was being capped too low by his ISP :).

      Most slashdottings come from limited bandwidth or php/perl/asp/mysql/ etc being unable to handle it. Most dynamic content could be static content. (Most slashdot pages are static caches).

    7. Re:Third Time's a Charm? by Anonymous Coward · · Score: 0

      Only a tiny percentage of slashdot readers ever actually view these comments. Sure, you can get a fair bit of traffic from being linked in the comments, but it's nowhere near being on the same scale as getting "slashdotted" -- that is, linked from the front page.

    8. Re:Third Time's a Charm? by TedCheshireAcad · · Score: 1

      Not so bad for 2 simple Dell workstation-class PCs. Glad to see they can take that kind of hammering and still live; cheers to the High Availability development crew.

    9. Re:Third Time's a Charm? by FCKGW · · Score: 5, Interesting

      Even though only a small percentage of Slashdot readers look at the comments, Slashdot's readership is so huge that the number of people reading the comments is still significant. It's not enough to kill a server, but I posted links to three images, around 80KB each, on my home server a few days ago fairly deep down in the discussion and got 3904 hits from it. It didn't kill my server (Pentium 133MHz, 64MB RAM, Debian 3.0, Apache 1.3.26, 3000/256 cable) and didn't result in any nasty letters from my ISP.

      OT: It was interesting reading the logs. There are quite a few Linux users on here (but even more Windows users), and I saw lots of people using Mozilla, Opera, Safari, etc. Compare that to sites aimed at the average user where 95% of visitors are using IE or AOL and don't know that there's anything better out there.

      --
      It's an operating system, not a religion.
    10. Re:Third Time's a Charm? by dnoyeb · · Score: 1

      hehe, so did I.

      I think this qualifies as something to re-post monday morning ~1pm. ;)

      Challenge!

    11. Re:Third Time's a Charm? by strange_attract0r · · Score: 1
      I think this qualifies as something to re-post monday morning ~1pm. ;)

      ummm ... monday morning when?

      --
      This sentence no verb
    12. Re:Third Time's a Charm? by Anonymous Coward · · Score: 2, Interesting

      What's the big deal? .... KPLUG.ORG (http://www.kplug.org) survived a slashdotting just fine... and we run on a P75 w/ 48 megs of RAM.

      Static pages and a big pipe. What's the mystery?

    13. Re:Third Time's a Charm? by Syre · · Score: 4, Informative

      Static content is right.

      In 1998 I was working at a startup and we served Olympics news to Excite.com, which was one of the very largest sites on the web back then. Excite linked to pages on our server which were at some URL like olympics.excite.com that pointed to us.

      What they didn't know was that we were serving all this off of ONE Sun Ultra 1 workstation on my desk (which was all we could afford at the time). I had it set up with Squid so everything was coming out of memory.

      It worked fine, even at peak times everything popped up about as fast as any other page on Excite.

      Moral: if you need speed and the page doesn't really have to be dynamic make it static or cache it.

    14. Re:Third Time's a Charm? by Stonent1 · · Score: 1

      Funny, I did that too, make sure to hit ctrl when you do it or it will redisplay from cache. (At least in IE)

    15. Re:Third Time's a Charm? by Stonent1 · · Score: 1, Funny

      What they didn't know was that we were serving all this off of ONE Sun Ultra 1 workstation on my desk

      Meanwhile, after work, the Filipino cleaning lady that doesn't speak english (Do not unplug Olypic news server!) unplugs it so she can plug in her vacuum.

    16. Re:Third Time's a Charm? by Anonymous Coward · · Score: 0

      Karma whoring by appealing to Mac fanboys is the very worst kind.

    17. Re:Third Time's a Charm? by NevermindPhreak · · Score: 1
      thats blatently racist to say that the cleaning lady who cant speak english was filipino.

      its filipino-american, thank you very much.

    18. Re:Third Time's a Charm? by CowboyMeal · · Score: 1

      Even though only a small percentage of Slashdot readers look at the comments

      Are you saying people on slashdot actually RTFA?

      --
      Your credit card information wants to be free.
    19. Re:Third Time's a Charm? by Anonymous Coward · · Score: 0

      Why would an ISP send you a letter for using the bandwidth you are paying for?

    20. Re:Third Time's a Charm? by fireboy1919 · · Score: 1

      I too have a website, and I've got the logs in a nice, easy to read form. Most of the readership of my site is from slashdot.

      Notice how many people are using Windows/IE?

      Of course, the only reason people usually visit my site is because I said something interesting and they wanted to see who I was. Maybe IE/Windows users are more prone to webwandering.

      --
      Mod me down and I will become more powerful than you can possibly imagine!
    21. Re:Third Time's a Charm? by TheMidget · · Score: 2, Funny
      It's not enough to kill a server, but I posted links to three images, around 80KB each, on my home server a few days ago fairly deep down in the discussion and got 3904 hits from it.

      Wow. Impressive. However, I would have hoped that by now the Slashbots would have learned not to click on image links occuring in a random Slashdot comment...

    22. Re:Third Time's a Charm? by pmz · · Score: 1

      Oh, come on. Even my little old G3 iMac is capable of handling quite a load from Slashdot...

      Aceshardware.com, also a Slashdot survivor, is run from a comparatively dinky computer: A single Sun Blade 100 (500MHz UltraSPARC IIe CPU; a decent low-wattage CPU, but a SPEC contender it ain't).

    23. Re:Third Time's a Charm? by xteddy · · Score: 1

      Yeah, I probably was on of them:
      fake_user_agent Mozilla/4.0 (compatible; MSIE 5.0; Windows 98; DigExt)

    24. Re:Third Time's a Charm? by anotherGuy214 · · Score: 1

      Ahhhh...that would only matter if SMU had a T3. A few years back when I was in school at the godforsaken(literaly, they are no longer recognized by the methodist church) place I believe there were a whopping two or three T1's feeding the campus. Which sucks a** when you live on campus. The IT dept. there is less then fun to deal with to get anything done...aside from the bandwidth efficiencies. But hey...as long as someone is still running a meth lab in the meadows arts building the IT budget should expand over time.

    25. Re:Third Time's a Charm? by chrisbtoo · · Score: 1

      After one of my ex-employers' rants about how we just needed to turn off our monitors at night, rather than purchase air conditioning, our friendly security guard took to going round turning off all the screens he could see.

      Many's the morning you'd come to work and all the workstations would be frozen because he'd turned off the serial console to our NFS server.

      --
      Registering accounts later than some other chrisb since 1997
    26. Re:Third Time's a Charm? by BigBlockMopar · · Score: 1

      ...but I posted links to three images, around 80KB each, on my home server a few days ago fairly deep down in the discussion and got 3904 hits from it. It didn't kill my server (Pentium 133MHz, 64MB RAM, Debian 3.0, Apache 1.3.26, 3000/256 cable) and didn't result in any nasty letters from my ISP.

      Yeah, I've been hammered a few times with exactly the same effect. (Pentium 90, 48 megs RAM, simple dynamic content, Apache, Linux.) I believe the primary limitation in my case is the speed of my DSL (1200kbps/128kbps less PPPoE overheads); when I've been Slashdotted, I've checked my outbound traffic with iptraf and found that I'd maxed out my upload speed and page requests were occasionally timing out rather than melting down the server. (Though seeing literally nothing but CGI scripts and httpd daemons on top's display is exhilarating, almost as good as skydiving but not quite as good as a big-block V8 or sex.)

      --
      Fire and Meat. Yummy.
    27. Re:Third Time's a Charm? by BigBlockMopar · · Score: 1

      Even my little old G3 iMac is capable of handling quite a load from Slashdot and this site is serving up graphics intensive stuff.

      I suspect, given the subject matter of your site, the chosen backgrounds and text colors are designed to test a theory that extreme garishness may cause blindness?

      Mac people are supposed to have style!

      --
      Fire and Meat. Yummy.
    28. Re:Third Time's a Charm? by BWJones · · Score: 1

      I suspect, given the subject matter of your site, the chosen backgrounds and text colors are designed to test a theory that extreme garishness may cause blindness?
      Mac people are supposed to have style!


      Yeah, so the deal is this site is essentially as it was back in 1993. I am recreating the whole site as we speak.

      --
      Visit Jonesblog and say hello.
    29. Re:Third Time's a Charm? by bhtooefr · · Score: 1

      Nice, but how do I make my FrontPage web use it? I've got it on my IIS server, but I need to convert my pages to PHP. Any quick way to do THAT?

    30. Re:Third Time's a Charm? by Stonent1 · · Score: 1

      Who said she was a filipino-american? She could be a resident alien. I worked with a guy that was British, he wasn't British american. He was a permanent resident.

    31. Re:Third Time's a Charm? by BigBlockMopar · · Score: 1

      Yeah, so the deal is this site is essentially as it was back in 1993. I am recreating the whole site as we speak.

      Keep a copy somewhere for an online museum!

      Sorry... usually a site like that is designed by teenage girls using Geocities accounts to post tributes to their favorite teen idol. Flaming them for their animated GIFs, horrible background, and MIDI assaults when I accidentally happen onto their pages is, of course, pointless. I guess I had a lot of built-up tension. Sorry.

      Ever heard the Growing Pains theme in MIDI? Urk. B.J. Thomas and Jennifer Warnes must not be impressed.

      --
      Fire and Meat. Yummy.
  2. I'm scared... by z-kungfu · · Score: 1

    to post anything, article wise or mirror, I think it'll bring my little ol' server to it's knees...
    plus having only a 384/384 pipe doesn't help.

  3. Apparently... by yanbusa · · Score: 5, Funny

    They are asking for another test.

    --
    What's in a sig?
  4. Weeee.... by Ridge · · Score: 2, Funny

    Third time's a char... DOH!

    Heh, well they're actually still kicking, oh well, it would've been so apropos.

  5. Let's get serious by termos · · Score: 1

    Okey people, let's prove them wrong!
    Open the links in the story twice from now on ;-)

    --
    Note to self: get smarter troll to guard door.
  6. Thou shall not by Anonymous Coward · · Score: 5, Funny

    Thou shall not survive thrice. You're insolence will not be tolerated. You'll servers will suffer a slashdotting not hence seen....

    1. Re:Thou shall not by Anonymous Coward · · Score: 0

      Your inattention to proper grammar will not be tolerated. Report to the termination chambers immediately.

    2. Re:Thou shall not by Sethus · · Score: 3, Funny

      Are you sure this is the third time? I count five...

      [Arthur, King of the Britons] One, Two, Five!
      [Knight] Three, Sir!
      [Arthur, King of the Britons] Three!
      **HALLELUIAH**
      *BooM*

      --
      Posting with out proof reading since 2001.
    3. Re:Thou shall not by Anonymous Coward · · Score: 0

      You are insolence? You will servers? WTF?

  7. Clever, clever by Frodo2002 · · Score: 4, Insightful

    Notice this comment was posted on a slow Sunday afternoon (EST). Very clever, because they know that /.'ers can't resist a challenge like that. Feel sorry for them on Monday morning though...

    1. Re:Clever, clever by Anonymous Coward · · Score: 0

      More interesting is the documentation of the apparent exponentially decaying attention span of slashdotters

      What were you talking about? I liek Linux.

    2. Re:Clever, clever by ashitaka · · Score: 4, Funny

      Feel sorry for them on Monday morning though...

      Monday morning!!?? You're kidding, right?

      They have already noticed the "exponentially decaying attention span" of Slashdotters.

      By Monday morning this story and the site will be relegated to un-clicked graveyard of "Older Stuff"

      --
      If you don't want to repeat the past, stop living in it.
    3. Re:Clever, clever by Anonymous Coward · · Score: 4, Funny

      Commander Taco will repost it on Tuesday....

    4. Re:Clever, clever by dnoyeb · · Score: 1

      Indeed. Around here this is dinner time :D

    5. Re:Clever, clever by LostCluster · · Score: 1

      Also, notice the challenge is issued on Sunday of one of the deadest weeks in the academic calendar. In most places, there's nobody in the dorms. Summer session is over, but the incoming freshmen don't move in until at least next week. This is when the computer departments are allowed to disrupt services and do the moves they've been planning since January (when the holiday break is the other dead time on campus) because there's hardly anybody left on campus to complain.

    6. Re:Clever, clever by non · · Score: 1

      its 10:39AM monday morning, at least where i am(GMT-2). they do have a couple of large m$ word documents, 1590k and 1460k, but all the mpegs have broken links!

      --
      ...vividly encapsulates that post-Watergate/pre-punk/coked-up moment when you could trust no one, least of all yourself.
  8. Wouldn't you need an expensive switch? by Billly+Gates · · Score: 4, Interesting
    I find it hard to believe that a cheap dual channel switch for 2 pc's could handle the slashdot effect.

    I was under the impression that a 20k fiber or 100mbs one that can dynamically shift traffic would be needed.

    1. Re:Wouldn't you need an expensive switch? by OverlordQ · · Score: 2, Insightful

      From the Article:
      Each machine has its own independent 100 Mbit connection to the Gigabit SMU internet service,

      I really didn't see 'cheap dual-channel switch' in there anywhere . . I DID see independent 100 Mbit connection though.

      --
      Your hair look like poop, Bob! - Wanker.
    2. Re:Wouldn't you need an expensive switch? by PurpleFloyd · · Score: 3, Insightful
      Assuming the page was fairly small, without lots of pictures or movies, I'd imagine that a small 10/100 switch would handle the load fairly well.

      On the stats link provided for a January slashdotting, the most bandwidth used in a day was a little under 7.4 GB. Assume that it was posted on /. at noon, so there were 12 hours to spread the big hit over: about 617 MB/hr. That's a little over 10 MB/Min, or 170 KB/sec. Even if we assume that the initial hit was double that, I can easily stream a 1000 KB/sec Divx movie over my 100 Mbps switched home LAN. The limiting factors here are the servers, routers and bandwidth to the Internet, not the local network connecting the servers.

      --

      That's it. I'm no longer part of Team Sanity.
    3. Re:Wouldn't you need an expensive switch? by Limburgher · · Score: 4, Interesting
      It's not the switch/throughput that's the issue. It's the amount of page generation being done by the servers themselves. The high throughput and individual links merely allows the maximum possible number of simulatneous hits from a bandwidth standpoint. The trick here is to do that, AND have your webserver survive, hence the dual box Linux Virtual Server setup. That's the piece that smells the /.ing and tries to load-balance the page generation. That ability should allow this setup to survive most /.ings.

      Very inexpensively, I might add. Beats having to buy the hardware for an, I dunno, . . .

      Beowulf cluster?

      --

      You are not the customer.

    4. Re:Wouldn't you need an expensive switch? by Anonymous Coward · · Score: 0

      Well, it seems the parent to this post can't read a bar graph...the 7.4 gb transfer isn't total it was just the greatest amount of information transfered in an hour, which comes out to 2.1 mb/sec. Not exactly the puny 170 kb/sec...

    5. Re:Wouldn't you need an expensive switch? by Ezdaloth · · Score: 1

      Still ... you'r box must be total crap, if it can't serve 2 MB/s of static content! My old p2/350 can serve 5 or 6 MB/s (ok, when i copy a single file, like a movie or iso, via ftp. but still.) Certainly as long as the content fits in the RAM, it is no deal at all. After that, your cheap ide disk might have a hard time. :P

    6. Re:Wouldn't you need an expensive switch? by jonadab · · Score: 1

      > I find it hard to believe that a cheap dual channel switch
      > for 2 pc's could handle the slashdot effect.

      You're on crack. A simple 100BaseT fast ethernet switch (which
      will run you $50 or so) running in full duplex mode can handle as
      much traffic as 66 T1 lines -- that's more traffic than one T3
      line can carry. The slashdot effect is a lot of bandwidth, but
      it's not _that_ much bandwidth. (LAN technology carries much more
      bandwidth for the buck than WAN technology; the reason it's cheaper
      is mostly because it doesn't have to handle the long distances.)

      The real danger of the slashdot effect is to people on narrow pipes
      (Cable/DSL -- or of course dialup) or with bandwidth caps, or to
      sites that carry ginormous bandwidth-intensive downloads with heavy
      geek appeal (ISO images anyone? What about a 500MB computer-
      generated animation of Darl McBride getting spanked?), or to sites
      with server-side active content that does something that consumes
      a lot of resources (such as database queries). Static HTML (and a
      few GIFs and sanely-sized PNGs) served over a university connection
      (probably T3 or better) isn't in any really serious danger.

      --
      Cut that out, or I will ship you to Norilsk in a box.
    7. Re:Wouldn't you need an expensive switch? by jonadab · · Score: 2, Insightful

      > After that, your cheap ide disk might have a hard time

      You do realise, don't you, that a cheap IDE disk can handle
      more throughput than a T3 line or a 10/100 switch? Refer to
      the speed chart below.

      Although, if you don't have adequate cooling, making it do so
      continually for a few hours could have a detrimental impact on
      its lifespan. But a university department server is probably
      in an air-conditioned building, which makes adequate cooling
      fairly easy, because ambient temperature stays low no matter
      how much heat you blow out the back of the box. So you slap
      a cooler on the HD and add one case fan and you're set, yes?

      Speed Chart:
      CPU, RAM, HD, Ethernet/T3, T1, Dialup, Keyboard
      Each step is _at least_ an order of magnitude slower than the
      previous step on the chart. Thus, if your bottleneck is the
      T3 line, even the cheapest IDE hard drive can easily keep up.
      Removable drives are well slower than the HD, but I don't know
      how they compare to the networking technologies. Residential
      broadband is generally not better than T1 and often is closer
      to dialup. (Dialup at "56k" gets real speeds up to 45kbps; T1
      is about 1500kbps, so 128 or even 256kbps is closer to dialup
      than it is to T1. 512kbps is (geometrically) closer to the T1,
      however, and so can be classified as true broadband.)

      --
      Cut that out, or I will ship you to Norilsk in a box.
    8. Re:Wouldn't you need an expensive switch? by gregmac · · Score: 1
      Beats having to buy the hardware for an, I dunno, . . .

      Nice.. The first time I've actually laughed at a comment with that word in it since .. well, ever. :p

      --
      Speak before you think
    9. Re:Wouldn't you need an expensive switch? by devnullify · · Score: 1

      CPU -> RAM: ~1.5GB/s (166MHz DDR) 7200RPM HD: ~20-50MB/s (sequential) 100mbit Ethernet: ~12.5MB/s T3: ~5.5MB/s T1: ~180KB/s 56k: ~5KB/s Keyboard: ~2.5KB/s Obviously all those figures are under optimal conditions, so yeah you're right...though I don't see how the hard drive would get much hotter from lots of access...it's the drive motor (along with air friction) that generates the most heat, not the heat actuator. As for broadband...I dunno where you're from, but around here residential broadband is generally 1.5mbit+ down, and 512kbit+ up. You should move :P

    10. Re:Wouldn't you need an expensive switch? by Anonymous Coward · · Score: 0

      There is one more problem: not how FAST you can send data, but how SLOW. Sure, you can handle three clients at a time with infinite bandwidth. No problem. Can you handle three hundred clients at a time on 28.8 modem (yeah, they still exist)? If you don't have enough memory things will get really slow. And then they will press refresh, all of them. And then you have 300 slowly dying processes/threads and 300 trying to start.
      My computer can serve stuff at >10 MB/sec, but that dosen't matter if I have to use swap space.
      So maybe the file fits in RAM, and the processes don't.

    11. Re:Wouldn't you need an expensive switch? by julesh · · Score: 1

      Well, it seems the parent to this post can't read a bar graph...the 7.4 gb transfer isn't total it was just the greatest amount of information transfered in an hour

      I thought this as well, at first glance, but if you look carefully, each bar represents a day, not an hour... I personally wouldn't have expected a slashdotting to continue for 3 days, but there you have it...

    12. Re:Wouldn't you need an expensive switch? by jo42 · · Score: 1


      Darl McBride getting spanked is here.

    13. Re:Wouldn't you need an expensive switch? by jonadab · · Score: 1

      > As for broadband...I dunno where you're from
      From the internet. Oh, you mean physically? Galion, Ohio.

      > around here residential broadband is generally 1.5mbit+ down,
      > and 512kbit+ up
      The 1.5 down is as good as T1. 512 up isn't quite, but it's
      still way better than what I could get here.

      > You should move
      Well, I'm going to colocate a server very soon; that'll help.
      For client stuff, in a pinch I can go in to work during my off
      time and use the T1. At home, I've adapted pretty well to my
      shared 33.6 dialup. wget rocks. Also, Mozilla's tabbed browsing
      lets me queue pages and continue reading the first page while
      they are retrieved. (Very nice for slashdot; scroll down the
      main page middle-clicking the interesting links, and by the time
      I get to the bottom the first story has loaded.) The new
      prefetching is nice too. The thing that really hurts is all the
      spam I get. Filtering helps with the time I have to spend
      _reading_ my mail, but it still takes thirty minutes to _download_
      it all. I do something else while I wait for it, but it's annoying.

      --
      Cut that out, or I will ship you to Norilsk in a box.
  9. well golly gee... by JeanBaptiste · · Score: 5, Insightful
    (also having a university-supplied BIG PIPE doesn't hurt).


    well there you go... having a massive amount of bandwidth will allow you to survive a slashdotting. In most cases of slashdotting, I dont think the server was the bottleneck... its no problem for a server to dish out static pages... its the bandwidth, especially for serving pictures or videos....
    1. Re:well golly gee... by Anonymous Coward · · Score: 0

      Well out of their experience one server isn't enough to handle a slashdotting.

      So yeah I guess if you link to static pages, especially with lots of data (pictures/videos) bandwidth and not the server is usually the problem.

      But in their case one server is not enough, so they use two.

      I guess if they used just one of those boxes, they wouldn't be able to handle the slashdottig, hence the dual system.

      However, a dual AMD MP 2600+ with 2 gigs of ram might be able to handle a slashdotting too if all it had to do was serve static pages and bandwidth wasn't limiting.

      In short read the article to see how much server power a slashdotting requires, and why they needed a dual server setup.

      Kudos goes to the guys though for their setup, though it isn't really *that* fantastic, it's more then I could set up ;)

      Now let's see if they could handle a slashdotting serving up freshly generated pages...
      (Like many news sites have)

    2. Re:well golly gee... by MinutiaeMan · · Score: 1

      Well, yeah. Like how my site got taken offline for a full 16 days after someone in a reply linked directly to a 600 KB image on my server. My entire monthly bandwidth quota of 6 GB was consumed in just 12 hours, and I couldn't afford to pay any overage fees, so my site languished for the rest of the month.

      I've taken precautions since then... mod_rewrite is a godsend! 'Course it won't deflect the tidal wave completely -- but like any good sea wall, it provides decent protection for all but the most massive floods.

    3. Re:well golly gee... by Lumpy · · Score: 4, Informative

      nahhh... you can survuve a slashdotting much easier...

      simply use thttpd.

      it has throttling built in so that the slashdotting wont take down the server. It simply stops serving images at anything but a really slow rate and eliminates multiple requests from same hosts... (A bitch for NAT'ed companies)

      There are httpd servers out there that are much better than apache for handling insane loads.

      --
      Do not look at laser with remaining good eye.
    4. Re:well golly gee... by mo · · Score: 4, Insightful

      You're right. Bandwidth also serves a big purpose in finishing requests quickly. For example, let's see what happens when I have a 1.5mbps line with 512 concurrent requests. First of all, if you're using apache 1.3.X or 2.0 prefork, you've filled 2 gigs of ram by spawning 512 clients. Furthermore, you're bandwidth allocation per-client is 384 bytes/sec. This means you're spoon feeding your pages to your clients which makes it really hard for your server to get that 512 number down to something manageable.

      The problem here, is that the bandwidth bottleneck will make your server either (a) run out of processes/threads, (b) run out of ports/sockets, or (c) run out of memory from spawning all of the processes/threads to handle all of the stalled connections.

      Once this happens, people no longer can connect to you, and you're toast. The crazy thing here is that this can happen even at 10mbit/sec if you're machine is configured well enough, and if the content you're serving is large enough (IE: image/media serving).

      So cheers for these guys at keeping their bandwidth/server ratio high, I actually really like their architecture. But note that the greatest architecture in the world won't save you from a slashdotting if your server(s) are on a business dsl line.

    5. Re:well golly gee... by Jeremiah+Blatz · · Score: 4, Interesting
      JeanBaptiste writes:
      well there you go... having a massive amount of bandwidth will allow you to survive a slashdotting. In most cases of slashdotting, I dont think the server was the bottleneck... its no problem for a server to dish out static pages... its the bandwidth, especially for serving pictures or videos....

      With static web pages, server power is rarely a problem, it's all about the pipes. However, if the pages are dynamically generated, and don't have a lot of caching, then you've got yourself a big problem.

      So take, for example, loading a forum page in UltimateBB. AntiOnline handily tells you how many db requests it takes to create a page, and how long it took. This one over here says 61 requests and .3 seconds. Now, the poster claims to be peaking at ~37000 page views/hour, which is 10 hits per second. Now in that .3 seconds, where 61 database connections were established, there were another 3 requests coming in, making it an average of 240 database connections every .3 seconds. That's not an unreasonable number of connections, but what if your DB server can't keep up? What if, due to the load, the queries take 10x longer than usual? At that point, over .3 seconds, you get 240 connections, but only service 24 of them. Over the next .3 seconds, you get another 240 requests, but service only another 24, leaving you with 436 pending. After 30 seconds, you've serviced 2400 requests, but have another 21,600 pending. before too long, you're out of possible TCP ports.

      There are ways to keep your servers from crapping out under heavy load. One is to buy a studly, fire-breathing DB server that can process requests faster than your web servers can send them. Another (cheaper) solution would be to pool and marshall your DB requests, being sure to remove requests from the queue when the remote user times out (either by clicking the stop button or running up against a built-in limit of their browser). This way, your site may get slow, but nothing will crash. A final method is to use enough caching on the web server that you pages are, essentially, static. This is, for instance, what Vignette does, which is why all the major news sites use it. This method combines the flexibility of database-backed CMS systems with the database load of static web pages.

      So, essentially, there are many ways to let your database-backed web site survive a slashdoting, but embedding a bunch of PHP SQL queries against a locally-running installation of MySQL is not one of them. Unless you have a big honkin' cluster.

    6. Re:well golly gee... by Ezdaloth · · Score: 1

      indeed. 100mbit (as the university suplies to me, gna, gna) is quite usefull for that purpose :) An old p2/350 can easily sustain that full bandwith with static content like images and html.

    7. Re:well golly gee... by caouchouc · · Score: 1

      Of course, those are old GX-line Dell Optiplex machines. They're Intel-based uniprocessor workstations and lack things that help servers serve up pages quickly & correctly, like a 64bit/66mhz PCI bus, SCSI, RAID and ECC memory.

      SCSI & RAID could possibly be installed, but on a University department's IT budget I don't see why they'd go to the expense when it may or may not help in some instances, and they could just grab a second box from storage/surplus.

      Those two old clunkers combined don't even come close to a dual AMD MP 2600+ w/2GB DDR ram. The dual AMD would handle a static-page slashdotting of the described magnitude easily, bandwidth permitting.

    8. Re:well golly gee... by seanadams.com · · Score: 5, Interesting

      It simply stops serving images at anything but a really slow rate

      What's the point? Either way you're slashdotted.

      Besides, I think in the case of server overload (as opposed to network overload), throttling will only exacerbate the problem by increasing the number of slow clients you have to deal with. This is the #1 bottleneck in web servers, the more clients you have, the longer it takes to deal with each one of them. Losts of processes to switch between, long arrays in an out of select(), etc.

      Also, when a user doesn't get a page in his browser, what does he do? He clicks the link again and again.... even more connections to handle.

      Really the only way to cure an overloaded server is to drop incoming SYNs. Any other measure is just pouring gasoline on the fire.

    9. Re:well golly gee... by evilviper · · Score: 2, Interesting
      But note that the greatest architecture in the world won't save you from a slashdotting if your server(s) are on a business dsl line.

      That's not true at all.

      Configure your maximum number of users to something your server/bandwidth can handle, and then send everyone else an error page that says they need to wait. That error page can automatically try to re-load the main page every 1-5 minutes, until it is successful.

      Even on your dial-up connection you could survive a slashdotting.
      --
      Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
    10. Re:well golly gee... by gbjbaanb · · Score: 1

      I find it interesting that they claim to survive a slashdotting when both times the links were posted on a sunday. Just looking at the number of comments posts get on sundays show that maybe 1/10 of the normal /. crowd gather - so really they should try a monday-morning post to see if they can stand up.

    11. Re:well golly gee... by JeanBaptiste · · Score: 1

      yeah pretty much....... any p2 properly configured should be able to handle all that and then some. the cpu is STILL not the bottleneck, its the bandwidth.

    12. Re:well golly gee... by Ezdaloth · · Score: 1

      indeed. until you would find out i run phpBB(doh! fully dynamic php bulletin board) and awstats (dynamically generated logfile stats from a perl script, takes A LOT of cpu !!) :P My p2 certainly doesn't pull that off. It hardly does that with 15 or 20 concurrent adsl users :)

    13. Re:well golly gee... by jemfinch · · Score: 1

      The problem here, is that the bandwidth bottleneck will make your server either (a) run out of processes/threads, (b) run out of ports/sockets, or (c) run out of memory from spawning all of the processes/threads to handle all of the stalled connections.


      Problems (a) and (c) are really problems with Apache's chosen server model (thread/process per connection) rather than with the slashdotting itself. Choosing a more appropriate server (such as thttpd, which also does throttling) pretty much solves those problems for you.

      Jeremy
    14. Re:well golly gee... by KPU · · Score: 2, Insightful

      Um sending people an error page is NOT surviving a slashdotting.

    15. Re:well golly gee... by evilviper · · Score: 4, Insightful

      We aren't talking about an actual error, that's just what it is called.

      What we are talking about is sending a different, low-bandwidth page, which will reload the main page after a certain ammount of time.

      Sure, you aren't immediately serving up the page to everyone who is requesting it, but people will only see a delay of a minute or so, rather than the server not serving up anyone, and crashing and burning for a day before anyone resets the thing. Or maybe just saturating the line, so nobody gets anything faster than 0.00001kbps, which is far far worse than having to leave the error page open in the background a couple minutes before it has a chance to load the main page.

      --
      Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
    16. Re:well golly gee... by Syre · · Score: 4, Informative

      Vignette is a rather trivial hack which was developed at C|Net and licensed to Vignette. It basically just writes out static pages, nothing fancy.

      As soon as Vignette licensed it, they were able to claim C|Net as a "Big Customer", which of course was a fib. This let them sell loads of software to organizations that didn't know any better.

      Moral of the story? Static pages are a great idea. If you can't do static easily, put a cache in front of your dynamic pages and decide how dynamic you really need them (Is 15 minutes delayed OK for you or is each page totally dynamic?). You don't need to spend $250K for Vignette to get this to work. In fact, you can do it with free software quite easily.

      Web caching and optimization is a big topic but in general the more clever you can be the less money you have to spend on big iron.

    17. Re:well golly gee... by Anonymous Coward · · Score: 0

      if you just drop the syns the hosts will keep retrying. you have to deny the connections... send out a fin.

    18. Re:well golly gee... by aaarrrgggh · · Score: 1

      Quick summary: if you have "unlimited" bandwidth and static pages, it is easy to survive a slashdotting.

      I can imagine that a little Via machine with a 10/T connection can survive a slashdotting... if all you do is serve text pages. Each image you serve cuts down that ability linearly.

      When you link to a database... the math gets much harder! If your link is saturated, and you have to maintain more connections at a time, you are out of luck.

    19. Re:well golly gee... by Czmyt · · Score: 1

      How do you run out of TCP ports? If my Web server has 1,000,000 requests pending, they may be coming from random ports on various IP addresses, but they're all coming to port 80 on my machines' IP addresses.

    20. Re:well golly gee... by seanadams.com · · Score: 1

      if you just drop the syns the hosts will keep retrying. you have to deny the connections... send out a fin.

      I think you mean a RST, but in any case yes, that's another way which avoids generating any state server side. But it does generate outgoing traffic.

      With web connections, the user's just going to keep trying anyway so it doesn't make much difference.

    21. Re:well golly gee... by Anonymous Coward · · Score: 1, Interesting

      I did a little test a while ago when setting up a site (CGI applications in C with MySQL API) on a dual Athlon MP 2000+.
      Apache could serve about 1100 static pages/sec (~40 kB file), 1800 404s/sec (~300 bytes or so) and 41 CGI app requests/sec (~4 kB).
      No one is expecting 41 hits/sec anytime soon so it's no problem, but I can imagine what would happen if I pasted a link here.

    22. Re:well golly gee... by paganizer · · Score: 1

      One of the machines I have cluttering up my mini-farm is a IBM server 320; the manuals for it state that it's capable of handling 10 concurrent hits on one processor, 30 with both.
      While I'd say the most important factor is a big pipe, having adequate CPU power to handle the requests is sort of important, also.

      And if you don't mind me asking, why did you make me your foe? am I doing something right?

      FREENET=FREESPEECH

      --
      Why, yes, I AM a Pagan Libertarian.
    23. Re:well golly gee... by 0x12d3 · · Score: 2, Funny

      huh huh. Hey Butthead he said "big pipe"

      --uuhh. That was cool.

    24. Re:well golly gee... by pmz · · Score: 1

      This is the #1 bottleneck in web servers, the more clients you have, the longer it takes to deal with each one of them. Losts of processes to switch between, long arrays in an out of select(), etc.

      With fair to decent rack-mount servers so cheap, any more (a $1000 server should handle a few hundred thousand hits/day for simple content, right?), adding more servers is a better solution than throttling, etc., too.

      Get a little 2-foot-high rack, sit it on a desk, and pack in a few CPUs, a load balancer, and a switch into it, and Slashdot should be nothing more than an annoying mosquito of the Internet.

    25. Re:well golly gee... by Jeremiah+Blatz · · Score: 1
      Czmyt writes:
      How do you run out of TCP ports? If my Web server has 1,000,000 requests pending, they may be coming from random ports on various IP addresses, but they're all coming to port 80 on my machines' IP addresses.
      Blah, nevermind, I was confused. I almost remembered how TCP works, but not quite. Running out of TCP ports is a problem on the database client, not the server. Each time you make a connection, the two machines negotiate a port to use on the machine that initiated the connection, in this case the web server, not the db server. There are only 16 bits worth of ports, afaik, so a max of ~32,000. Running out of ports might still be a problem in this example, but only if you have a really cool web server that can manage to support enough users to spawn 32k+ connections to the db server.
    26. Re:well golly gee... by Jeremiah+Blatz · · Score: 1

      I thought the cool bit about Vignette was the CMS workflow, and the serving part was just a decent cache implementation. Anyway, the other thing to note about Vignette was that they were doing the caching thing a long time ago, when you coulnd't just grab some free software to do it. I expect that CNN doesn't really have a lot of interest in trashing all the infrastructure that they've built around Vignette.

    27. Re:well golly gee... by Syre · · Score: 1

      Actually AFAIK, C|NET never did implement Vignette. They stuck with their old stuff and kept developing it.

      Vignette was a fork that C|Net ignored cuz it didn't do what they needed it to do.

  10. Is it decaying attention span? by Eevee · · Score: 5, Insightful

    Or is it where the article is at any given time? Top of front page gives lots of hits. As it drifts down, the hits slow as fewer read; to the sidebar, fewer but still substantial hits; then off to the specialty pages such as Science or Games, then only a few will read.

    Of course, the only test would be to repost the article and see if there's the same number of hits... Nah, slashdot would never go for duplicate stories.

    1. Re:Is it decaying attention span? by SpaceLifeForm · · Score: 1

      LOL. So, are you saying to try again at, say, just for hypothetical purposes, 2003-08-11 16:30 GMT?

      --
      You are being MICROattacked, from various angles, in a SOFT manner.
    2. Re:Is it decaying attention span? by anthonyrcalgary · · Score: 1

      Well they could, but instead of load testing the thing people would just bitch about the duplicate.

      --
      When someone might yell at me, it has to be OpenBSD.
    3. Re:Is it decaying attention span? by Anonymous Coward · · Score: 0

      Of course, the only test would be to repost the article and see if there's the same number of hits...

      yeah that will never happen, oh wait

  11. That hits graph by KingDaveRa · · Score: 3, Insightful

    Interesting how it peaks, drops off slowly then rises again a little before dropping off again. Maybe some 'behind the curve' slashdot readers?

    1. Re:That hits graph by Anonymous Coward · · Score: 0

      Nope, not "behind the curve" slashdot readers, those in other parts of the world (eg: Australia and Europe).

    2. Re:That hits graph by Pharmboy · · Score: 1

      or some of us who once we see the site is /.ed, wait a while before trying again. or give up since it cant be that damn important. or someone (often) posts a mirror of the /.ed article. or someone quotes it.

      there is probably more of an explanation than just a short attention span. Wait, that is what you were talking about, right? ;-)

      --
      Tequila: It's not just for breakfast anymore!
    3. Re:That hits graph by Anonymous Coward · · Score: 0

      Australians are "behind the curve."

    4. Re:That hits graph by Anonymous Coward · · Score: 0

      More typically Australians are to be found behind the sheep, giving it a good reaming from behind. BAAA BAAAA BAAAAA

    5. Re:That hits graph by xyvimur · · Score: 3, Informative

      Not quite - look at the plot it is monthly plot - there are 31 bars - one for each day. If we had daily plot - the time zone effect would be visible...
      However I don't know how to explain this phenomenon (Slashdot is good for making research and writing some thesis).

    6. Re:That hits graph by Anonymous Coward · · Score: 0

      Hell, in New Zealand they sell velcro gloves at the 5 and dime.

    7. Re:That hits graph by Anonymous Coward · · Score: 0

      Perhaps the increase is a weekend when more people have spare time and the ability to access slashdot from home machines (while not supposed to be doing so at work during the week). Just a theory, not actually checked if those dates correspond.

    8. Re:That hits graph by gspr · · Score: 1

      A lot of "minor" news site get their stories from slashdot, and they might be a bit late. Maybe the visitors coming from these sites make up the final rise?

    9. Re:That hits graph by xyvimur · · Score: 1

      January 2003:
      18th - Saturday
      19th - Sunday (hmm, logical)
      26th - Sunday (beginning of second `peak')
      31st - Friday (end of plot)

      Aww, the saturdays and sundays are marked on the plot (I used calendar)

    10. Re:That hits graph by Anonymous Coward · · Score: 0

      Interesting how it peaks, drops off slowly then rises again a little before dropping off again. Maybe some 'behind the curve' slashdot readers?

      Contrary to popular belief, some slashdot readers actually do other stuff rather than refresh slashdot every 5 minutes waiting for a new story.

    11. Re:That hits graph by freeweed · · Score: 2, Funny

      Or dupes.

      --
      Endless arguments over trivial contradictions in books written by ignorant savages to explain thunder in the dark.
    12. Re:That hits graph by tqft · · Score: 1

      24h later could be "people" (if /.er's qualify for that title) hitting it from the email summary

      --
      The Singularity is closer than you think
      Quant
    13. Re:That hits graph by dam · · Score: 1

      The slight rise is probably the weekly readers. Not all of us read slashdot _all_ the time ;-) As the rise in this case starts monday, it implies that a lot of people read slashdot at work.

      --
      Cheers, Duncan
  12. well... by painehope · · Score: 2, Funny

    I've got some exponentially decaying pieces of chicken on my table...

    and some exponentially growing forms of life in some beer cans...

    does that count?

    --
    PC moderators can suck my White pierced, tattooed dick. If you think pride == hate, s/dick/Aryan meat mallet/g.
  13. Need.... more.... band.... width..... by caluml · · Score: 1

    I only wish I had enough bandwidth to be able to test the hardware out properly.
    2Mb? It's nothing for a nice server.

  14. A.D.D. crowd by Anonymous Coward · · Score: 5, Funny

    More interesting is the documentation of the apparent exponentially decaying attention span of slashdotters.

    Well, I was gonna reply, but I forgot what the post was about.

    1. Re:A.D.D. crowd by macshune · · Score: 3, Funny

      I think the decaying attention spans of slashdot users has to do with the rise of modern media and the fast-paced moving images in the contemporary media paradigm. Further, I think that A.D.D. is the result of HEY!!! LET'S GO SWING!!!!!!

    2. Re:A.D.D. crowd by finallyHasANickname · · Score: 0, Offtopic
      I was just thinking about A to D converters, all of which wind up offering Data, which seems reasonable for something whose acronym would be A.D.D.--what do you mean by swing? Golf? Baseball? Java? Music? Sex?

      Hey wait a minute! What was it we were talking about?

    3. Re:A.D.D. crowd by Kibo · · Score: 1

      I here by virtually mod you +1 Funny. You'll have to keep score yourself, sorry.

      But slightly more seriously, I think ADD is probably more a former evolutionary norm that humanity might be moving away from. It facilitated one kind of group coordination, where small groups of people had to perform many types of tasks and be open to targets of opportunity. But the larger kind of coordination, and hyper-specialization that we now engage in to make the economy go round finds it a little antiquated. So since we've gone through the trouble of inventing neurochemistry....

      But on another level, you're right, Clone High was completely sweet and the only thing worth watching on MTV. Ghandi's struggle with ADD, and nasal fixation, not only made me laugh but gave me hope. Which reminds us of Equilibrium, where if the evil puppet masters decide we all should be medicated, for our own good, we should kill the obviously emotional nazi bastards. And really, why couldn't they have just been a little more consistant within the confines of the story? Gah!

      --
      --Jimmy has fancy plans; and pants to match.
  15. Oh, oh, the dupes, the dupes. by slackingme · · Score: 2, Funny
    "Our little departmental server has been slashdotted twice in the last year and survived!"

    I hope the dupe jokes that get posted are better than what I had here before I decided to just post this in hopes of drawing better jokes later in the thread. They weren't good! But when are they?

  16. zerg by Lord+Omlette · · Score: 2, Funny

    This is almost as stupid as wearing a mask of Donald Rumsfeld and visiting 3ID troops to tell them "Are we planning on letting you go home anytime soon? Gosh, no!"

    These SMU guys are obviously begging for it, I say we give them hell!

    --
    [o]_O
  17. A slashdotting? by Chmarr · · Score: 3, Funny

    Just under 40,000 hits in the busiest day... this is a slashdotting? Come back when you get into the millions. :)

    1. Re:A slashdotting? by Pharmboy · · Score: 1

      then again, we COULD just wget the whole damn site, delete and re-wget it again. simple "while true" script it would seem. that would tend to increase the hits at an exponential level.

      while true
      do
      wget --delete-after -m -p http://www.geology.smu.edu
      done

      or something like that.

      --
      Tequila: It's not just for breakfast anymore!
    2. Re:A slashdotting? by Anonymous Coward · · Score: 0

      Just wanted to say: You are a very funny person...

      NOT!!!
      Come back when you've learned how humor works.

    3. Re:A slashdotting? by user32.ExitWindowsEx · · Score: 1

      Just store it in /dev/null

      --
      "Evil will always triumph because good is dumb." -- Dark Helmet
    4. Re:A slashdotting? by Pharmboy · · Score: 1

      yea, i could, but my /dev/null partition is almost full ;-) yea, either way is fine, both take as much to type, guess im just used to the built in way of doing it.

      If open a few more shells or vt's, and get like 4 or 12 or 50 wget's going with -m and -p, its a nice way to make the server thrash as well, assuming the site has considerably more content than ram (which it probably does, with all the media files i saw). I went from pulling 80k to 50k currently, so they have bandwidth and cpu to spare. need to hammer them a bit more i guess. I just started a few more instances, we will see if that helps them.

      --
      Tequila: It's not just for breakfast anymore!
    5. Re:A slashdotting? by FreeLists_Guy · · Score: 1

      Yeah, I agree. Something didn't seem right about his numbers, and mine are regularly *on average* over 150k per day.

    6. Re:A slashdotting? by Chmarr · · Score: 1

      About 4 million hits, or 400 thousand pageviews, per day, here :)

  18. evil by Anonymous Coward · · Score: 1, Interesting

    hey look at this mpeg about rocks or somthing!!! MUAHAHAHAHAA

  19. Nope... by ryanvm · · Score: 5, Funny

    Anybody else observed similar phenomena?

    Nope. In our jobs they make us do work.

  20. Yeah but it is a small home page... by twoslice · · Score: 2, Informative

    There is a total of 80K of information on the entire front page! Dig a little deeper guys and perhaps we can find a few gigantic image downloads. If you find any do share =)

    --

    From excellent karma to terible karma with a single +5 funny post...
    1. Re:Yeah but it is a small home page... by LostCluster · · Score: 1

      Memo to web designers who think having a 2 MB Flash animation on the first page is cool: Nobody's gonna be able to see your wonderful piece of art.

      Having a small home page that can be served quickly is a key factor in not losing visitors when the heavy usage comes.

    2. Re:Yeah but it is a small home page... by DaLiNKz · · Score: 1

      160kb/s (my max).. smooth downloading...

      --
      I've left to find myself. If you happen to see me, please, keep me there until I return.
  21. Nice but your big pipes the real secret by plierhead · · Score: 2, Informative
    Nice work, but of course your amazing resilience to the multi-headed hydra that is the slashdot audience is all to do with your BIG PIPE and nothing much to do with your dual headed linux configuration.

    While your setup may make you real safe from machine outages, the effects of a slashdotting are to flood your resources rather than break them. So your configuration gives you at best the performance of two machines instead of one - which you could also have achieved by just ramping up the CPU or memory.

    --

    [x] auto-moderate all posts by this user as insightful

  22. [GASP!] You mean LOAD BALANCING HELPS! Stunning! by siberian · · Score: 2, Insightful

    Its incredible, this person has actually proven that LOAD BALANCING MULTIPLE SERVERS INCREASES YOUR LOAD CAPACITY! This is incredible news! Wow, I am sure glad it made it as an article, stunning.

    Every medium to large website out there will be pleased to know that what they have been doing for the last 8 years is actually VALID, thanks guys!

    I think the only reason this made it to the front page is the slashdot self-reference.

  23. Eleventh Post by fm6 · · Score: 2, Insightful
    The first ten all make the same lame joke. Get a life people!
    More interesting is the documentation of the apparent exponentially decaying attention span of slashdotters.
    I'm not sure the decay fits any simple math model. Here's a more practical explanation: It takes about 24 hours for a Slashdot story to scroll off the main page. So for the first day and part of the second you're getting hits from every slashdotter. After that you're only get hits from people who compulsively look through old stories and/or browse Slashdot through RSS feeds and other offline tools. And after that, of course it's old news.

    I'll bet if you chart the data hour-by-hour, you'll see a sudden dropoff at the very moment the story scrolls off.

    1. Re:Eleventh Post by Scooter · · Score: 1

      Indeed - that graph does not measure attention span at all (how long each person looked at the page might come close but then thats kinda hard to measure - once they have the page there endeth the interest of your web server in the transaction).

      Maybe the ./ community should rent itself out as a load tester :) Then again unless you have vast tracts of bandwidth, the traffic is effectively throttled by the pipe anyway.

    2. Re:Eleventh Post by ralphdaugherty · · Score: 1

      I'm not sure the decay fits any simple math model. Here's a more practical explanation: It takes about 24 hours for a Slashdot story to scroll off the main page. So for the first day and part of the second you're getting hits from every slashdotter. After that you're only get hits from people who compulsively look through old stories and/or browse Slashdot through RSS feeds and other offline tools. And after that, of course it's old news.

      I'll bet if you chart the data hour-by-hour, you'll see a sudden dropoff at the very moment the story scrolls off.


      It's just a little worse than that even. All it takes is one or two of the articles pushing it down the page to be controversial (M$ or SCO). Then posts for the first article will stop even before it gets pushed off the page.

      Why aren't these replies also simultaneously posted to a forum board like phpBB so a dialogue can develop and be followed longer than an article lasts on the front page?

      rd

  24. Exponential? by s-orbital · · Score: 1

    Woudn't that be a logarithmic decay rate?

    --
    Patent: from Latin patere, to be open
    1. Re:Exponential? by 26199 · · Score: 2, Insightful

      Er... no. Logarithms increase.

      Exponential decay is of the form:

      y = e^-a x

      Where a is positive. It's exponential in that a fixed increase in x multiplies y by a fixed number; it's just that this number is less than one.

    2. Re:Exponential? by s-orbital · · Score: 1

      Well, I said that because expoential growth starts slow, and then skyrockets (Like the earth population over the last 1000 years)
      And logarithmic growth is more like the acceleration rate of a car taking off from a stop light.
      Anyway, I was applying that to a decay, since it decays fast at first, but then the decay rate slows, but this was in error I suppose.

      --
      Patent: from Latin patere, to be open
  25. Bandwidth? by Henry+V+.009 · · Score: 1

    I always figured the majority of slashdotting occured because of bandwidth constraints. I suppose a few servers crapped out, but that has got to be few and far between, hasn't it?

    Don't let me crap all over your cutesy linux saves the day story, though.

    1. Re:Bandwidth? by Anonymous Coward · · Score: 0

      I often thought that maximum number of connection limits in web servers (I think apache typically defaults to 50). Also on some systems there will be a maximum number of processes limit, if each request forks a new process this may be hit rather quickly.

    2. Re:Bandwidth? by ldspartan · · Score: 1

      Its definitely bandwidth. ticalc.org, which has always had a puny server, has survived several /.'ings thanks to its large, large pipe.

      That and intelligent site design. Most of the pages on ticalc.org (most =~ all) are pregenerated.

      --
      lds

  26. Attention sp...hold on be right back. by Anonymous Coward · · Score: 0

    Huh?

  27. decay pattern by tomdarch · · Score: 2, Interesting
    I'm surprised that there was so much traffic on the second and third days (given that stories drop off the front page)

    It's also interesting that there was a second little bump about a week later. Anyone have any ideas why?

    1. Re:decay pattern by automatix · · Score: 1

      dupe, of course :)

    2. Re:decay pattern by robolemon · · Score: 1

      People steal links from Slashdot and put them on their blogs or Fark or somewhere. I've found several stories this way, thought "wow I can't believe this person found this out!", only to go to Slashdot later that day and see it way down on the list.

      --

      I design user interfaces for a free network management application,

    3. Re:decay pattern by ameoba · · Score: 1

      Yeah... I tend to see /. stories pop up on some of the mailing lists I'm on after being on the fron page...

      --
      my sig's at the bottom of the page.
    4. Re:decay pattern by pipingguy · · Score: 1

      I think the article mentions something aboot a TV press conference / media attention about their recording of the seismics of the event, plus wasn't the explosion on a Saturday? Those without the net at home would undoubtedly surf for the event on Monday.

    5. Re:decay pattern by Drakonian · · Score: 1

      If you read CNET or Ars Technica or eWeek or a few places like that, you'll see Slashdot "steals" the links from there. You can stay ahead of the curve by reading other sites first. There is some original Slashdot content like interviews, but not a lot. I guess the really valuable service that Slashdot provides is the links to those little cool homebrew projects yoiu never would have found on your own. But of course by the time you see it on /., the server is long dead.

      --
      Random is the New Order.
  28. Are you asking for a challenge??!? by DuctTape · · Score: 0, Redundant
    DOUBLE-DEUCE!!!!

    had to be said... sorry.

    DT

    --
    Is this thing on? Hello?
  29. You didn't get Slashdotted by 0x0d0a · · Score: 4, Insightful

    You didn't get Slashdotted if the server was still operating normally. You just had some people from Slashdot visit.

  30. umm.. whaaa? by Squareball · · Score: 1

    More interesting is the documentation of the apparent exponentially decaying attention span of slashdotters.

    I think this is why Slashdot makes you wait 20 seconds before you can submit a comment. They figure this will weed out 80% of the responses as the people get bored of waiting for the 20 seconds to be up and they just leave.

  31. Well its sorta dead by Gherald · · Score: 1

    $ ping -t www.geology.smu.edu

    Pinging geology.heroy.smu.edu [129.119.223.84] with 32 bytes of data:

    Request timed out.
    Request timed out.
    Reply from 129.119.223.84: bytes=32 time=1317ms TTL=240
    Request timed out.
    Reply from 129.119.223.84: bytes=32 time=1254ms TTL=240
    Request timed out.
    Reply from 129.119.223.84: bytes=32 time=1538ms TTL=240
    Request timed out.

    1. Re:Well its sorta dead by Latent+IT · · Score: 4, Informative

      Are you serious? Because... this is what I get:

      Pinging geology.heroy.smu.edu [129.119.223.84] with 32 bytes of data:

      Reply from 129.119.223.84: bytes=32 time=58ms TTL=232
      Reply from 129.119.223.84: bytes=32 time=61ms TTL=232
      Reply from 129.119.223.84: bytes=32 time=60ms TTL=232
      Reply from 129.119.223.84: bytes=32 time=56ms TTL=232
      Reply from 129.119.223.84: bytes=32 time=58ms TTL=232
      Reply from 129.119.223.84: bytes=32 time=74ms TTL=232
      Reply from 129.119.223.84: bytes=32 time=67ms TTL=232 ...blah,blah,blah...

      Ping statistics for 129.119.223.84:
      Packets: Sent = 16, Received = 16, Lost = 0 (0% loss),
      Approximate round trip times in milli-seconds:
      Minimum = 56ms, Maximum = 74ms, Average = 60ms

      I mean, that's pretty damn smooth for a 30 minute old story. That's probably peak /. effect time, too. I'm really proud of them, and their little beige boxes. =)

    2. Re:Well its sorta dead by Gherald · · Score: 1

      I was serious. And I am on a very reliable 600 up 400 down ADSL connection.

      Strange...

    3. Re:Well its sorta dead by revmoo · · Score: 2, Funny

      And I am on a very reliable 600 up 400 down ADSL connection.

      Apparently not :-)

      --
      I would expect such blatant racism on Fark, but on Slashdot? Mods please ban this asshole.
  32. How much bandwidth do you need to survive? by captainboogerhead · · Score: 3, Interesting

    What I want to know is, how fat a pipe do you need to survive a slashdotting, given that your server structure is viable? Will a 10mbps pipe keep the barbarians from trampling the gate?

    1. Re:How much bandwidth do you need to survive? by Isaac-Lew · · Score: 1

      It's really a combination of bandwidth, hardware & apache configuration. A T1 or higher link, fast enough processor (or SMP system) with 512 MB RAM or more & a decent apache configuration (sorry I'm not on Linux right now, or I'd be more specific) should be enough to survive a slashdotting.

    2. Re:How much bandwidth do you need to survive? by pjrc · · Score: 1
      A T1 or higher link [and other stuff] should be enough to survive a slashdotting

      In this case (Sunday afternoon), it looks like the peak was about 10 page view per second, of a page containing approximately 80 kbytes of data. Not including http and tcp/ip protocol overhead, that's about 800000 bytes/second.... quite a bit faster than a T1 line at approx 150000 bytes/second.

      As many others have pointed out, static pages and plenty of bandwidth is a no-brainer.

    3. Re:How much bandwidth do you need to survive? by SuiteSisterMary · · Score: 1

      I've yet to see, but maybe I'm not looking for, a web-programming langague that does auto-magic query caching.

      What I mean is this. Say you've got a query that retrives comments from your database. This query gets run, with little to no adjustment (say, in the form of /., you basically pick your threshhold and your story) many many times.

      So be able to define that as a recurring query in the web-programming language. Then, give some directives on how often that query should actually be run; say, once every five minutes, or every 300 executions. Then, when your code calls that query, it'll get a normal result set back, but that result-set will be either a) returned from memory, or b) an actual database hit, depending on conditions.

      I know a good database will do lots of this on it's own end, but the ability to not even have data going from webserver to database and back would be good.

      --
      Vintage computer games and RPG books available. Email me if you're interested.
  33. The Slashdot worm. by caluml · · Score: 4, Funny

    Lets help them out.

    while :; do wget http://www.geology.smu.edu/~dpa-www/venus/mpeg/atl a1.mpg -O /dev/null -o /dev/null ; done

    Don't forget to fix the space in the URL.

    1. Re:The Slashdot worm. by Pharmboy · · Score: 4, Funny

      i already posted mine before i saw yours. my solution was

      while true
      do
      wget --delete-after -m -p http://www.geology.smu.edu/
      done

      which has the advantage of not only hammering their bandwidth, but since it requests everything on the entire site, including images and many multiple mb files, then the server cant cache it all, and will have to read from the disk alot. also, since it deletes after downloading, it doesn't take up your precious drive space, lol.

      but nice to know another sick fucker had the same basic idea.

      --
      Tequila: It's not just for breakfast anymore!
    2. Re:The Slashdot worm. by berenddeboer · · Score: 1

      What do those presumptuous geologists think they cannot be slashdotted? This will be the /. of the century!

      --
      If I had a sig, I would put it here.
    3. Re:The Slashdot worm. by berenddeboer · · Score: 1

      It seems people noticed your script :-) They're definitively slacking now.

      Two remarks:

      1. Perhaps the --cache=off option might be useful, so we really retrieve stuff from there server.
      2. And with recursive, you're reusing the connection, not spawning new daemons. I'm not aware of a wget option where you can have force it to create a new connection.
      --
      If I had a sig, I would put it here.
    4. Re:The Slashdot worm. by Pharmboy · · Score: 1

      Two remarks:

      Perhaps the --cache=off option might be useful, so we really retrieve stuff from there server.
      And with recursive, you're reusing the connection, not spawning new daemons. I'm not aware of a wget option where you can have force it to create a new connection.


      I thought of that, but I noticed that as long as you used the --delete-after option, it would see the file as a different size (0 vs 0) and would refetch. Then again, it was hard to tell since i spawned about 12 in the same shell, and with all that noise, it was hard to read. I was also hard to kill. sent several killall -9 but it wouldn't die until i shelled in again and picked off the first few instances.

      As far as using a new connection using --no-http-keep-alive, that is a pretty good idea, and will use that if I am wget'ing alone, but since i put the script up for the lazy on /., im pretty sure there are enough people hitting it that its a moot point, lol.

      --
      Tequila: It's not just for breakfast anymore!
    5. Re:The Slashdot worm. by parkanoid · · Score: 1

      The -o /dev/null and -O /dev/null promptly send the retreived files to /dev/null, so that script does not waste drive space, either. Heh.

    6. Re:The Slashdot worm. by sanity_slipping · · Score: 1

      How do we metamoderate this to "-6, National Security" ?

      ---

      --
      I can feel my sanity, beyond my reach and slipping...
  34. Found a big one one guys 113 Mbytes by twoslice · · Score: 1

    Multi-sensor.tar.gz

    Try a hand at downloading the above file from the site (113MB) and see what transfer rate you get?

    --

    From excellent karma to terible karma with a single +5 funny post...
    1. Re:Found a big one one guys 113 Mbytes by Anonymous Coward · · Score: 0

      31.25K/s and rising

      then again I think they have plenty of bandwidth

    2. Re:Found a big one one guys 113 Mbytes by scottj · · Score: 1

      I'm getting about 30KBps on this link.

      --
      .-.--
    3. Re:Found a big one one guys 113 Mbytes by Anonymous Coward · · Score: 0

      I get about 100k/sec.

      I think it's more the number of simultaneous connections that stress hardware; they've got the bandwidth, so one more big download doesn't really hurt em. Lots of small ones might.

    4. Re:Found a big one one guys 113 Mbytes by jwilhelm · · Score: 1

      359kb/s ... but I'm on a University connection now too, so it's possible if they are an I2 school as we are, that it's routing my traffic over I2.

  35. They asked for it by muon1183 · · Score: 1

    Since they seem to think they can handle the slashdotting, I've taken the liberty and searched through their site. This page has a number of links to pages which have large images and movies linked to. Hopefully, this will really provide them some load.

    --

    There's no sig like SIGSEG
    1. Re:They asked for it by muon1183 · · Score: 1

      take, for example, this 6.6MB mpeg video (Warning, direct link to video).

      --

      There's no sig like SIGSEG
  36. Exponentially decaying attention span by finallyHasANickname · · Score: 1

    Now if you could just automate the conversion of the httpd usage data over time, then convert that into the right scale, put it out through a ttyx, whereupon it could enjoy a leisurely conversion, D to A, and then connect the analog output to a proportional low-frequency-compatible amplifier, in turn, to a large diameter speaker in the back of a garbage can, the other end of which has a moderate-to-small orfice in it, then the thing could blow more smoke, as per previous article, than all of us /. blowhards combined. ;-) (Just a thought.)

  37. Pretty simple... by seanadams.com · · Score: 5, Interesting

    My server has been slashdotted a few times and I can tell you it's pretty simple to not get overloaded.

    The first time I learned my lesson. The server was on a T1 line that was 2/3 full already, and slashdot linked to a page full of large photos. That'll kill your link pretty quickly. Low-budget solution: sign up for a burstable web hosting account somewhere and just put all your large images there.

    Later when we got some actual office space for the business, I moved the main server up to a colo facility in fremont. All slahdottable content is hosted there on a fast server with a 100mbps ethernet link. Other oddball services that need their own machine are hosted from the other end of a point-to-point T1 line going directly back to the office from the colo.

    So depending on your budget it's really not hard to set up your site to survive a slashdotting. If you don't have a lot of dough to spend but you want to run your own server for configurability/security reasons, just host the static stuff somewhere else. Or if you're serving enough to make it economical, get a colo account with a burstable link.

    There's a widespread misconception here that slashdotting is caused by server overload. In reality this is almost never the case. It's caused by insufficient bandwith. This in turn may cause server overload because of too many slow clients being connected, but that is purely a secondary effect.

    1. Re:Pretty simple... by evilviper · · Score: 2, Funny
      There's a widespread misconception here that slashdotting is caused by server overload. In reality this is almost never the case.

      It's more likely to be server overload here than anywhere else, don't you think?

      Dear Slashdot,
      I just installed a webserver on my Gameboy. Would you please turn it into a smoldering hole in the ground?

      Thanks!


      Also, with a little smarts from the server, you can prevent bandwidth overload. All you really need is to limit the number of concurrent users that can be connected to your webserver. That way, you only have (let's say:) 10 users loading the bandwidth-intensive pics at any one time, and everyone else is looking at a simple error page that automatically trys to re-load the normal page every minute or so.

      It seems simple enough to me, and it's done places like spamcop.net, I wonder why I never see /.ed sites configured that way.
      --
      Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
    2. Re:Pretty simple... by Spellbinder · · Score: 1

      what????
      no burning servers after a slashdotting???
      it is a shame :p

      --


      stop supporting microsoft with pirating their software!!!!!
  38. Advice: add trailing slash to URLs if needed by Anonymous Coward · · Score: 2, Informative
    Add / to directory URLS, that will avoid a 301 redirection (/foo -> /foo/).


    BTW, using static pages also helps too. What is more, the "how to not survive" includes "generate content dynamically every time".

  39. Serve up some dynamic pages by easyfrag · · Score: 1

    OK Mr. Tough Guy, put a database on the backend and serve up some dynamic content. Then we'll see if you can really survive a slashdotting.

  40. Re:Our solution by Anonymous Coward · · Score: 0, Troll

    Slashcode eats up my example php code.

    if($_SERVER[HTTP_REFERER] == "http://slashdot.org/")
    {
    header("Location: http://www.tubgirl.com");
    }

  41. Slashdot effect analysis by Titanium+Angel · · Score: 3, Interesting
    More interesting is the documentation of the apparent exponentially decaying attention span of slashdotters. Anybody else observed similar phenomena?"
    Actually, yes. Stephen Adler did an analysis of the /. effect, where you can see similar graphs. It's not surprising, really.

    Also, there was an article on a hardware review site, if I remember correctly, where their approach to handling extreme load was discussed after their site was linked on Slashdot. Unfortunately, I can't find the article right now. Anyone around here who remembers?
  42. I survived, with such a pipe. by SHEENmaster · · Score: 1

    But my site wasn't linked to from the main page, doesn't have large downloads, and doesn't have large images. Even the logo was text, rather than graphical, at the time.

    --
    You can't judge a book by the way it wears its hair.
  43. Testing, testing... by Mish · · Score: 5, Funny

    They're just begging for a 'real' test... ... such as everyone downloading this:

    ar405eng.exe (5.41 MB)

    from their webserver :p

    5.41MB per slashdot reader should provide a test worth of such a fat pipe ;)

    1. Re:Testing, testing... by DeathPenguin · · Score: 1

      Done with an average of around 95KB/sec. That oughtta fix his wagon!

    2. Re:Testing, testing... by jesco · · Score: 1

      I'd say nay, it won't work. Considering that some universities have 655mBit lines, this makes for about 30,000 users downloading this file at 20k/sec.

    3. Re:Testing, testing... by ibennetch · · Score: 2, Insightful

      True, especially since the Kazaa-happy college kids aren't due back for two more weeks.

    4. Re:Testing, testing... by Anonymous Coward · · Score: 0

      394 KBps ... ... Done. :)

  44. I know... by Dark+Lord+Seth · · Score: 2, Funny
    More interesting is the documentation of the apparent exponentially decaying attention span of slashd

    Huh? What?

  45. Also in the news: Carrot Top; Still Not Funny by Anonymous Coward · · Score: 2, Insightful

    I think that's part of it. But I bet most of the effect goes to Karma whoring. Notice the second minor blip later on in their data.

    When the article is new, the rush is on for insightful comments that deal with commenting on elements of the referenced links. (Might as well, there aren't any comments beyond ascii pictures, and troll expiditionary forces.) They have their responses, which then triggers the volley of RTFA's, and now there are a number of posts, people don't have to RTFA so much and the thread contains so much information anyway. But certain conversations develop, some tangential, but others still tightly following the information referenced in the links, and may even provide deeper links, which cause people to go back and reference the original works which provoked the server beating.

    So one might look at there data and then form the yet to be tested hypothesis that the second blip is accurately representative of the slashdotters who are genuinly interested in any random subject at hand, and the difference between that and the peak could be correlated to the number of whores.

    But that one isn't me, because A) Sunday B) I forgot to take my adderall.

  46. Exponentially decaying attention span? by localroger · · Score: 2, Informative

    Yeah, I noticed that too.

    --
    Brackets contain world's first nanosig, highly magnified:[.]
  47. Attention Span by cmacb · · Score: 4, Interesting

    I really don't think the Slashdotter attention span is any different (or if different, it is longer) than the average Internet user.

    When articles appear on the first page, they get attention, as they scroll to the bottom they get less, as they move to background pages they get significant;y less.

    While I often look beyond the front page, I am less likely to delve into the articles or discussions there, since almost everything that needs to be said HAS been said by then.

    I've carried on conversations with people regarding Slashdot articles long after the article appears. This can take place in journal entries or via e-mail where the discussion material can be easily kept as opposed to Slashdot comments which ultimately disappear anyway.

    The fact that people don't continue to click on the original source URLs doesn't mean anything.

  48. Mirrors by brejc8 · · Score: 2, Insightful

    The ideal situation would be if you got a warning from slashdot and then then made some mirrors of the pages on distributer mirror.

  49. Somebody's... by SerialHistorian · · Score: 2, Funny

    ...a glutton for punishment!!! Either that, or they want to test it some more. Do your worst, slashdotters!

    --

    --
    Vote for your hopes, not for your fears - Vote Third Party

  50. Here are the testing materials by krir · · Score: 5, Informative

    Here are some mpeg files from their server: 3.8mb , 3.6mb and 320kb

    1. Re:Here are the testing materials by in7ane · · Score: 3, Funny

      Mod the parent up!!!

      And download those MPEGs...

    2. Re:Here are the testing materials by Slime-dogg · · Score: 1

      Hmm... B&W image elevation abstraction. Pretty cool.

      --
      You need to restart your computer. Hold down the Power button for several seconds or press the Restart button.
    3. Re:Here are the testing materials by TheMidget · · Score: 1
      For those of you with phat pipes:

      while : ; do wget -o /dev/null -O /dev/null \
      http://www.geology.smu.edu/~dpa-www/venus/mpeg/atl a2.mpg & sleep 2 ; done

    4. Re:Here are the testing materials by Anonymous Coward · · Score: 0

      Over 300KB/s for those mpegs :p

  51. pipe hardware by windex82 · · Score: 1

    I belive the quality and size of the pipe are going to have a lot more to do with surviving a sdos (slashdot denial of service) then the hardware the server is running on, well with any 'recent' hardware anyway.

  52. Call up gotfuturama.com by spudchucker · · Score: 1

    And mirror them for me - Thanks

  53. Dreamcast! by RonnyJ · · Score: 1

    Pah, with a big pipe like that, I'm sure that even with a Dreamcast router, you could handle Slashdot :)

  54. I bet these guys are glad... by TheSHAD0W · · Score: 1

    They don't get charged for web bandwidth by the gigabyte. :-P

  55. Re:[GASP!] You mean LOAD BALANCING HELPS! Stunning by Anonymous Coward · · Score: 0

    Your comment caused me to lol, resulting in me dropping my sandwich on the floor. It landed with the wrong side down.

  56. Censorship on Slashdot by Anonymous Coward · · Score: 0

    Will someone please explain to me how this is a troll?

    How would YOU like to be stuck with $3000 in overcharges because you were linked to on Slashdot?

    For me, this means I would have a hard time feeding my family (I'd have to NOT pay my ISP and take a credit rating hit, obviously) and taking care of other basic needs. I only run my website for my personal page and a few things I do to donate back to the community.

  57. More detail, please by embobo · · Score: 4, Interesting

    Congratulations on surviving /.ing. I have a few questions.

    How were LVS and HA configured? With two systems, I can only guess that each was a real server (using the LVS terminology). Also both would be load balancers, with one being selected as active using HA.

    How did using HA or LVS help surivive a /.ing? Were there failovers? How many? When? Why? If surviving /.ing consisted of a high rate of failovers then the hardware wasn't up to the job.

    What is the "automated backup system?" Are you rsyncing the contents? From each other? From another system? Or does it refer to regular "tar" backups to tape?

    Having separate UPSs is overkill, unless the one UPS could not handle the load of both systems.

    Is there any dynamic content on the servers? Databases? How was keeping these synchronized handled?

    1. Re:More detail, please by nick_urbanik · · Score: 1

      The article only points to generic links (to http://linux-ha.org/ and http://www.linuxvirtualserver.org/) and gives no useful information as to how the director was set up, nor how the shared storage works, nor how things are kept consistent. The parent is the only one so far to raise this obvious question.

    2. Re:More detail, please by shfted! · · Score: 1

      Having two UPSs is not overkill in anyway. What if you had to replace the batteries? What if one failed for some reason? By getting two at half of needed capacity, these situations are easily handled.

      --
      He who laughs last is stuck in a time dilation bubble.
    3. Re:More detail, please by embobo · · Score: 1

      Good points. I stand corrected. Thx. :).

  58. hmm.. save yourself the trouble by focitrixilous+P · · Score: 1

    http://geology.heroy.smu.edu/~dpa-www/nvar/
    Auto refreshing javascript... loading a nice sized image file every 20 seconds. Open it up, and leave it in the background for a few weeks. Make them pay for their daring challenge to the /. collective.

    --
    SAILING MISHAP
  59. Subscriber Slashdottings by MBCook · · Score: 4, Interesting

    What I'd really like to see would be a graph of a BIG site when we Slashdot them now. It would be very interesting to see the subscribers and what they do before the /.ing public sees it. I couldn't seem to see one on the graph that they posted. Is it just that small? Just wondering.

    --
    Comment forecast: Bits of genius surrounded by a sea of mediocrity.
  60. mirroring by jack+in+the+mac · · Score: 0

    well a dandy way to reduce load is to mirror servers
    i suggest we all start mirroring it for them RIGHT NOW

    p.s. it uses cURL

    --
    Joy!peffpwpc
  61. Attention span could be useful by Jahf · · Score: 3, Interesting

    Given the sharp decline, this highlights another way that /. could help alleviate /.ing of sites: stagger the time that a certain client gets informed of a new article.

    1) RSS feeds would get the update -last- or in some form of randomness.

    2) Anonymous (no cookie) clients get the same treatment

    3) People logged in get the article sooner but are also stretched out. An example:

    a) If your UID is in the 25% of the oldest active users you get the article as soon as it is published (after going out to subscribers, who always get it first, another very mild reason to subscribe especially if you like to FP)

    b) If your UID is in the 26-50% of the oldest active users you get the article 30 minutes after it is published.

    c) If your UID is in the 51%-75% you get it 1 hour after it is published

    d) If your UID is in the last block you get it 90 minutes after publishing.

    e) If you are pulling from RSS or anonymously you get it 120 minutes after publishing.

    This also gives a little treat to the folks who have been around the longest while not removing the benfit of subscribing.

    Another example could work like the above but randomly change which order each block of UIDs will get the article (with RSS and Anon getting it last) if you wanted to not show preferrence to older users.

    Increments could be adjusted ... 2 hours for full distribution is going to be friendlier to the /.ed sites, but 1 hour total would probably still be effective.

    The only people this would affect negatively are FPers, SPAMboarders and people who have a cow-orker walk by and go "hey d00d, seen that new article yet?". No one else would probably even be aware of it unless they find it from another site that found it on /.

    --
    It is more productive to voice thoughtful opinions (reply) than to judge (moderate) others.
    1. Re:Attention span could be useful by mce · · Score: 1

      In view of my UID, I guess I am supposed to like this also for other reasons. But I definitely don't. Being able to talk about what's on /. with my roomies is 75% of the fun, even though/if their UIDs are orders of magnitude larger.

    2. Re:Attention span could be useful by LostCluster · · Score: 2, Insightful

      The problem with that is by giving a heavy disadvantage to the newest user on the system will likely make that new user feel unwelcome. That leads to abandonded newbie accounts, which is a bad thing for business. If you have no replacement customers coming in for the ones who leave, you'll have a decaying site and there won't be any /. effect anymore, nor will there be a /.

      Beware of side effects when you try to implement simple solutions like that...

    3. Re:Attention span could be useful by DaoudaW · · Score: 1

      Given the sharp decline, this highlights another way that /. could help alleviate /.ing of sites: stagger the time that a certain client gets informed of a new article.

      Unless you also stagger the moderation you'd put a large number of karma whores out of business. The story would have to look fresh (without a large number of comments) even to those who don't see it for a couple of hours. I think it'd be unworkable!

    4. Re:Attention span could be useful by Jahf · · Score: 1

      Well, first off you could randomize it so the newbie gets the article -first- 1/4 of the time. Secondly, how would the newbie even know about it unless someone walked by and said "seen that /. article?" and then it is an opp to socialize by saying "no, show me".

      --
      It is more productive to voice thoughtful opinions (reply) than to judge (moderate) others.
    5. Re:Attention span could be useful by LostCluster · · Score: 1

      I think a newbies would suspect something is up when they see that there's always at least 150 posts already there by the time the article first appears to them.

  62. Slashdotting by Luigi30 · · Score: 1

    I want to see if my site can survive a slashdotting. Look at this picture: http://www.floppy-disk.co.uk/users/luigi30/f-14.JP G .

    --
    503 Sig Unavailable

    The Signature could not be accessed. Please try again later or contact the administrator
  63. reverse proxy = survive huge traffic spikes by dcgrigsby · · Score: 1

    I find that the single largest unecessary piece of load comes from having huge apache children (i.e., apache with mod_perl or php) sending bits down slow pipes to browsers.

    We solved this problem at mozo (plug: personal dvd networks), and of course we're not alone in this approach, by running tiny-footprint apache children with nothing more than the basics + proxy capabilities. These guys act as reverse proxies, pass the requests off to the bigger mod_perl children, who when they're done processing can go onto the next request without having to push bits down to modem connections.

    The math on this is fairly simple, but worth mentioning: ((100 connections * a few hundred-K for proxy children) + handfull of mod_perl children) 100 mod_perl connections. So we don't end up memory bound.

    1. Re:reverse proxy = survive huge traffic spikes by dcgrigsby · · Score: 1

      oops. The less than sign got escaped out. The math:

      ((100 connections * a few hundred-K for proxy children) + handfull of mod_perl children) < 100 mod_perl connections

  64. Try this large file (78.3 MB) by Anonymous Coward · · Score: 0

    http://www.geology.smu.edu/~vineyard/cdrom/CD_Data /01_English/01_ETextbook/English.pdf

    they're so gona die.

  65. Live Stats by SP200308 · · Score: 1

    And you can view the current stats here at the Usage Statistics for geology1.heroy.smu.edu page.

    (Some reasonable-sized images on there.)

  66. Small Server? by 32bitwonder · · Score: 1

    ...they don't know the meaning of the word. Try hosting a site on a Macintosh LC III (that's 25Mhz folks) running Linux & Apache. Whether or not it can survive a good slashdotting remains to be seen however. :-)

  67. Exponential decay by rjh · · Score: 4, Insightful

    Why is it surprising that it follows an exponential dropoff? The only interesting questions are the coefficients of exponential dropoff, not that it's exponential--I'd sit upright and take notice if it was a linear decrease.

    Anything which follows a steady fractional diminishment will have a curve of y = ke**-ax, where k and a are constants. You see this basic equasion pop up all the time in physics, economics, statistics... etc. Why should server slashdotting be any different?

    1. Re:Exponential decay by pfafrich · · Score: 1

      A far more prnounced exponential decay is exhibited in the comment
      posted and scores acheived. I had a look at the score vrs time
      posted and got a graph a bit like:

      5 **
      4 **
      3 **
      2 *** *** **
      1 *** ******* *****
      0 *** ******* ******

      where the x-axis is number of hours since article posted, one * per hour.
      Basically most of the comment activity happens in the first two or three
      hours.

      Points to note, if you want to get a high score, post quickly.
      There are problably good posts which are posted late but never
      get to see the light of day. Quite understandable behavour
      as moderators can't be bothered to wade through 100's of
      posts, higher chance of redundant post.

      There might be ways where moderation system could be changed
      to streatch the curve out.

      Unfortunatly no one will get to read this ;-(

      --
      There are four sorts of people in the world: fools, lunatics, idiots and morons. - Umberto Eco, Foucaut's pendulum.
    2. Re:Exponential decay by pmz · · Score: 1

      Why should server slashdotting be any different?

      Because were nerds. The fact that something we do actually correlates with the known physical world...is a cause for concern.

  68. Short Attention Span by donnz · · Score: 1

    You must be kidding. There I was congratulating modertors on their appreciation of my finer comments only (days) later to find some dweeb(s) modding me down. Oh for a short attention spand.

    --
    -- Free software on every PC on every desk
  69. NAT'd companies by midgley · · Score: 2, Informative

    Should have squid cache running inside their network, so only one request for a given file should be necessary.

  70. video link by Anonymous Coward · · Score: 0

    I see you also disabled your live video link when posting your site to slashdot.

  71. exponentially decaying attention span? by jazman · · Score: 5, Funny

    Not sure how you get that from the graph. For myself, I didn't know what the subject matter was, so I opened the window, went "ugh, geology", and closed it more or less straight away. Ok, perhaps this proves the point - for subjects I'm not interested in I have a short attention span, but this doesn't mean I have a SAS for everything.

    You get an exponentially decaying number of hits, yes, but how many of those are people doing exactly what I did and not staying, as opposed to those who stay a while because they find geology interesting?

    The last time you were /.ted, did the graph decay at the same rate or did it take longer to go down? If it took longer that would suggest shortening ASs, but then did you have anything of special interest up at the time? Bung some pr0n up there and see if the, er, bulge is a different shape.

  72. What about the servers? by -tji · · Score: 1

    He really didn't address any steps that were taken to improve the server's performance under load. Many servers can do effective caching of static HTML pages, so a load of 42,000 hits/day is simple to serve. I'm surprised that the load was this low..

    42,000 hits/day == 1,750 hits/hour == ~30 hits/minute

    Of course, the hits are not distributed evenly throughout the day.. they peak shortly after the story was posted (a graph of 24hr activity would have been very informative).

    Even if you assumed all 42K hits happened in a four hour window, that would still only be ~30 hits per second.. Which almost any server today can handle for static content. Also, stats showing the system load levels would be very interesting.

    In general, I think that most 'slashdottings' are the result of insufficient internet access bandwidth, or bandwidth caps on the web hosting provider. I think his plentiful bandwidth saved his machines. The server load balancing and HA are good ideas, and cool projects. But, they were probably not relevant to the server's performance.

  73. Data analysis by Anonymous Coward · · Score: 0

    Looking at the hits, we see that the exponential decay has a half-life of about 1.44 days. I don't have the source data, so that's just a guess ;-)

    So the mean of that distribution is 1.44ln2 ~= 1. I.e. on average, people read slashdot once a day.

  74. -1, troll by Anonymous Coward · · Score: 0

    how do you think the ADD feels about your insensitive remarks? geeks are made fun of a lot, yet you make fun of ADD people. hypocrisy, anyone?

  75. Re: slashdot usage pattern by http · · Score: 1

    theory:
    metamoderation. when i see a moderation on a comment on an article that i haven't read that isn't obviously fair or unfair, I somet^H^H^H^H^H usually go and read the article before making a decision.

    --
    If opportunity came disguised as temptation, one knock would be enough.
    3^2 * 67^1 * 977^1
  76. Work on your self-esteem... by Anonymous Coward · · Score: 0

    Are you insecure?
    Because it sure sounds like it.
    Too many dunks in the toilet, eh?

  77. Bring Up The Holy Slashdot! by niko9 · · Score: 2, Funny

    Taco:Yes, of course! The Holy Slashdot of OSDL! 'Tis one of the sacred relics Brother Cowboy Neal carries with him. Brother Neal! Bring up the Holy Slashdot!

    AC's chanting: Pie Iesu domine, dona eis requiem.

    Brother Neal: Armaments, chapter two, verse nine to twenty one.

    Brother Neal: And Saint Cowboy Neal raised the Slashdot up high, saying, 'O Lord, bless this Thy Slashdot that, with it, Thou mayest slashdot Thine enemies to tiny bits in Thy mercy'. And the Lord did grin, and the AC's did feast upon first posts, trolls, GNAA posts, and...

    Taco: Skip it a bit, Brother.

    Brother Neal: And the Lord spake, saying, 'First shalt thou click on the holy link called Slashdot. Then, shalt thou count to three. No more. No less. Three shalt be the number thou shalt count, and the the number of the counting shall be three. Four shalt thou not count, nor neither count thou two, excepting that thou then proceed to three. Five is right out. Once the number three, being the third number, be reached, then, clickest thou holy Slashdot of OSDL towards thy server, who being naughty in My sight, shall snuff it.'

    Taco: Amen

  78. Sort of Sounds... by bsrokc73013 · · Score: 1

    Like how to survive an orgy with a small dick!

  79. Depends on what you're serving... by Kjella · · Score: 1

    ...a minimalist webpage, or the latest flash-over-animated site. I hope whoever replies also includes the size of the files being slashdotted....

    Kjella

    --
    Live today, because you never know what tomorrow brings
  80. The loss of attention span... by Anonymous Coward · · Score: 0

    Can only be attributed to the disappearance of GIS.

  81. exponentially decaying attention by QEDog · · Score: 3, Funny
    "exponentially decaying attention span of slashdotters"

    It is called Geek A.D.D.

    --
    "There is no teacher but the enemy."-Mazer Rackham
  82. I just don't understand it. Really. by NerveGas · · Score: 2, Interesting


    The guy is talking about bursts of 42,000 hits per day, and talking about it "bringing their system down". Now I could see that on Windows, but not on Linux.

    Now, before you think I'm talking out of my posterior orifice, when my company was young and bright, we had a server built on a single 450 MHz Pentium 2, and 256 megs of RAM. It ran both Apache and PostgreSQL. Many of our pages were database-driven, which of course is a much larger load on the server than simple static pages.

    That little machine would peak out at around 60,000 hits per day. At that point, it was slow enough to be self-limitting, but there was never any fear (or realization) of having the machine "brought down".

    So, still "back in the day", I replaced it with a dual P3/650. That machine would peak out at around 100,000 hits/day (database driven!), without much problem at all. Also, as time goes on, and we develop new apps that make further use of our data, we tend to need more power to generate every page. Even still, we could crank out 40,000 hits per day on what would now be a relatively anemic server.

    Now, with 7 front-end web servers and a dedicated DB machine, we crank out 5 million hits/day without problem. And even when our systems have been IMMENSELY overloaded from both legitimate and illegitimate traffic, the systems have still responded, and never once have I ever worried about a machine "going down" from the load. Failed hardware, perhaps, but not the load.

    steve

    --
    Oh, you're not stuck, you're just unable to let go of the onion rings.
  83. Thought it was a university? :) by nunofgs · · Score: 1

    Compare and contrast the statistics for February 2003 with those for Jaunary 2003 in Figuire 3, when our website was first mentioned on slashdot.org.

    Well there you go! There's the reason for the apparent exponentially decaying attention span.

  84. Re:OT by E_elven · · Score: 2, Interesting

    Well, it's 8:20pm EST and I'm not getting through from a T1.

    --
    Marxist evolution is just N generations away!
  85. Re:Our solution by jazir1979 · · Score: 1


    No, but ignoring ACs (or moderating them down) is pretty standard.

    --
    What's your GCNSEQNO?
  86. Re:Our solution by Anonymous Coward · · Score: 1, Insightful

    You're getting modded down for a number of reasons:

    1. You're posting AC. Try logging in, ok?
    2. Your code redirects the user to goatse.cx. Other variations you have posted redirect to tubgirl.com. This shows that you're not really concerned about slashdotting (why the variation in code? you mean this isn't the *actual* code you use? why am I not surprised?). You just want the juvenile thrill of trolling.
    3. If you honestly want to prevent a slashdotting, how about redirecting to a google cache of your page?

    If you're really serious, login and post your contribution without being so trollish. How do you expect to be serious when you say "Slashdot users are BANNED, now go look at an erupting asshole!" ?

  87. Not too hard by m.dillon · · Score: 1
    I just turned on tcp inflight limiting on my FreeBSD box, which works kinda like TCPVegas, and even though the T1 bandwidth was pegged all day long it survived being slashdotted just fine. Of course the site only had one or two carefully resized graphics on it in anticipation. Still, it worked. Even the slowest consumer box these days can web serve far more bandwidth then a T1, DSL, or cablemodem can support so it really comes down to bandwidth management. If the client is downloading a large picture and its going slowly you want to be sure that when he cancels your server isn't going to push out another 64KB+ of data before it figures it out, nor do you want to allow too many packets to backup on your interface (which destroys the latency for everything going out of the machine), The inflight limiting solves that problem.

    -Matt

  88. done, done and I'm on to the next one by simon333 · · Score: 2, Insightful

    I remember the days when I would treasure any new content and what I read or saw had more of an impact on me. Now information is just catalogued in my head and I feel this strange need to gather more all the time.

    I think the attention span problem is more widespread than just us slash-heads. People are now being inundated with constant 'quick clips' and cut scenes for every television show and commercial ...and don't even get me started on modern day movie trailers. There are so many cutscenes in trailers now that I literally have to close my eyes in the theatre to avoid having an epileptic seizure.

    How can our brains avoid being desensitized with so much information being thrown at us all the time?

    - Simon

  89. Re:OT by Anonymous Coward · · Score: 0

    Well, it's 8:20pm EST and I'm not getting through from a T1.

    There's something wrong with your connection then, cuz I have been poking around this site for 45 min now. It's pretty extensive. And fast. pretty impressive for a little iMac.

  90. Insulting by telstar · · Score: 1

    I don't have a shorter attention span than anyon.... hey, let's go ride bikes!

  91. Subject (yes, this post is about subject) by pasi · · Score: 2, Insightful

    Remember also that even though it's a fact that a huge amount of Slashdot users are interested in articles like this, still I'd imagine that a "normal" article with actual information attracts more readers, and therefore causes more traffic and server load. I could be wrong too, but I doubt it. ;)

    They survived this before, just saying that judging their performance now by this article may not be correct. Subject does matter.

  92. /. Response and my own site by FunWithHeadlines · · Score: 1
    I've noticed the same thing when I post a comment. I got a flurry of /.ers checking out my site that first day, then not as many the next day, and quickly trailing off.

    Of course in my case this kinda cheeses me off cuz my site is meant to be viewed on a daily basis (assuming you think my humor is funny enough to return). The fact that my traffic inevitably trails off tells me that either /.ers have the attention span of a gnat, and forget to bookmark brilliant humor sites...or else my site is not that funny.

    Naaaaah...must be the former.... ;)

  93. SMU Ponies... by lpret · · Score: 1
    Damn ponies. They need to just buck up and buy a better box.

    I can't wait for September 13th when our football team gets to beat the crap out of 'em. Sic 'Em Bears.

    --
    This is my digital signature. 10011011001
  94. How to load test a server by maddskillz · · Score: 2, Funny

    Seems like an easy way to get your setup tested

  95. That's the trick by LostCluster · · Score: 1

    (also having a university-supplied BIG PIPE doesn't hurt)

    Duh. What this article really shows is how small of a server is needed to serve up content, the most frequent limiter in a /. effect is a bandwidth bottleneck leading out of whatever small company has unfortunately become popular. Sites hosted where there is proper bandwidth usually come out fine...

  96. Wrong assumptions by felix9x · · Score: 1

    A properly configured server should have not problem handling large amounts of traffic even on a cheap PC. The network is the bottleneck i imagine.

    Consider this:
    100Mb is approx. 10,000 1K connections / second

    If someone actualy had the full 100Mb bandwidth they could achive 10,000 requests a second on a 1Ghz , 512MB machine. Apache may not be ideal for this but it can even do configured to use threads. Zeus,IIS,thttpd could do it too.

  97. Why not eliminate Slashdotting? by afabbro · · Score: 1
    Slashdotting is so irresponsible by Malda & Co.

    In the FAQ there is a question:

    "I could try asking permission, but do you want to wait 6 hours for a cool breaking story while we wait for permission to link someone?"

    The answer would be yes, of course. But Malda refuses to ask this question to the readers. I know because I tried:

    2003-04-08 23:29:59 Permission to Link (Ask the Audience the Q in the FAQ) (polls,slashdot) (rejected)

    What the hell?

    --
    Advice: on VPS providers
  98. Asymmetric load balancing can help too by billstewart · · Score: 2, Insightful
    If you want to continue serving your _real_ users while giving the slashdot crowd as much fun as you've got leftover horsepower to handle, and you're planning for this sort of thing in advance, it may help to trick your system into serving the slashdot crowd from one server and the regulars from a different one. Some obvious implementations:
    • Use the HTTP REFERER to direct /.ers to the second machine.
    • Use a friendly low-graphics front page that asks /.ers to click a link that goes to the second machine.
    • Modify your DNS so that people you recognize go to the primary machine and unrecognized people go to the second machine. If you're a university, most of your regular users are probably from .edu or .gov sites. If you're doing this anyway, you can also check DNS lookups to see if they come from open relays or spamblocked sites and send them to your _very_ _slow_ _smtp_ teergrube machine.
    --

    Bill Stewart
    New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
    1. Re:Asymmetric load balancing can help too by pmz · · Score: 1


      +1, Good Idea.

      Given that there are very few websites that can spike traffic like Slashdot, this idea is so simple, it just might work!

  99. Best way to survive is to run Windows Server 2003 by Anonymous Coward · · Score: 1, Interesting

    Probably the best way to survive being slashdotted is to run Windows Server 2003 and make sure you are making use of the kernel mode cache - serving up many thousands of pages per second.

    Even if you are generating pages dynamically (aspx, etc), if they can be served from the kernel cache you will get close to 2x the performance than you would get from serving boring/static pages from Apache on Linux. :-)

  100. Yes - Here is the URL link to the benchmarks. by Anonymous Coward · · Score: 1, Informative

    http://www.veritest.com/clients/reports/microsoft/ ms_competitive_webbench_performance.pdf

  101. I survived :D by Anonymous Coward · · Score: 1, Informative

    with a k7-1300, Apache without php or any dynamic stuff, static pages, and a little 256/128k DSL connection... the linked file from Slashdot frontpage was 60kb long in my server... and it resisted it surprisingly well

  102. Probably a dumb question by Lord_Dweomer · · Score: 2, Funny
    I don't know much regarding the technical details of this, so pardon my ignorance, but why is it that Slashdot never gets slashdotted?

    --
    Buy Steampunk Clothing Online!
    1. Re:Probably a dumb question by limbostar · · Score: 3, Funny

      Slashdot doesn't serve much in the way of images -- most of the content is textual -- so the bandwidth doesn't fill up as easily as it would if they were serving movies.

      The server configuration is designed to handle the load, with multiple servers, load-balanced arrays, that sort of thing, whereas the people they link to are typically running on shared servers, or have only a single server.

      Slashdot uses cached pages to avoid hitting the DB on every page load (mostly for the front pages), whereas smaller sites can get away with making a direct connection and doing more processor-intensive queries. Until they get linked by a site like Slashdot, anyway.

      Slashdot's DB server is most likely of the 'fire-breathing god' variety, able to handle standard Slashdot traffic without too much difficulty. Smaller sites typically have the database server on the same machine as the webserver, and sometimes both are shared.

      In general, it's all a matter of configuration. When you run a moderately successful small site, you're generally prepared for the amount of traffic you have, plus or minus 50%. Traffic generally grows slowly, so you have time to make adjustments when things start to get tight.

      When Slashdot links your site, you get a huge influx of traffic to a site that is designed to handle a tiny fraction of that traffic. It leads to badness.

      It's like trying to put an elephant into your freezer. If you're prepared for it, you have a big walk-in freezer. But most sites only need a small half-height fridge to keep their beer cold.

      --
      this is a sig.
    2. Re:Probably a dumb question by burns210 · · Score: 1

      slashdot.org expects to get slashdotted (the name gives it all away). and so before they do(since the beginning, i suppose) they have mulitple servers that are load balanced, along with a pretty good sized pipe for bandwidth.

      Slashdot effects happen to hobby sites, or sites that rarely get a more than a few dozen/hundred hits, so there is no need for those site to have redundency or large bandwidth. Basicly, we at slashdot try to blindside every smalltime developer and hobbiest who have a good idea by massively DDOSing there site with our userbase...

      It doesn't pay the bills, but it is fun.

    3. Re:Probably a dumb question by alister.b · · Score: 1

      One problem for Slashdot has been all the comments, even if the page that is output is cached somewhat. One solution that livejournal.com has created, and is now being investigated by /. and Sourceforge is MemcacheD, run by danga, the 'parent company' of Livejournal. It will cache comments in memory, and user information so it doesn't have to be dragged out of a database, even one as quick as MySQL.

      --
      --
  103. ha! by rabtech · · Score: 2, Interesting

    my server survived a slashdotting just fine - IIS 5.0 / win2k sp3, 512 ram, single IDE HDD, P3-800mhz, etc.

    The problem was, as it is for most people who get slashdotted, I didn't have a big enough pipe. Nothing to be done about that. I can't afford an OC3.

    --
    Natural != (nontoxic || beneficial)
  104. Temptation by sabit666 · · Score: 1

    Our little departmental server has been slashdotted twice in the last year and survived!
    Suddenly, after a while, I am having a temptation to read the aritcle..:)

  105. It's not the server... by Anonymous Coward · · Score: 0

    It's rarely the server that cannot survive a Slashdotting. It's almost always the connection.

    Minimal hardware can support many hundreds, if not up in to the thousands of requests per second, even for non-static content.

    For instance, a dual PIII-550, 512MB of ram, Win2k/IIS server can easily do 1000+ RPS for static content. It's simply a question of whether or not the connection to that server can handle the throughput.

    If your server can't handle a slashdotting while serving static content, then it must be pretty pathetic hardware.

    1000 RPS = 3,600,000 requests per hour. Does slashdot actually do that many hits? I think not.

  106. Bah by ttyp0 · · Score: 2, Insightful
    No need to throw lots of hardware and redundant systems to survive a /. From what I've seen, most sites that go down are due to a small number of reasons:

    1. Not enough bandwidth
    2. Poorly configured web server
    3. 300k images & 20MB mpeg/avi downloads (see #1)
    4. Not enough RAM (1GB is generally enough)

    I host about 50 domains for friends on my webserver (an Athlon 1.2Ghz w/ 1GB RAM) and have survived a simultaneous Slashdot and Fark link.

    --Brent

    Anti SCO T-Shirt

  107. Justin is a bottom. by Anonymous Coward · · Score: 0

    No, but I'll sing it for you.

    Oompa oompa oompa oompa oomp do da do.
    Oompa oompa oompa oompa oomp do da do.
    sing that song!
    I don't know what I'm thinking bout, really leaving with you
    It feels like something's heating up, can I leave with you?

  108. is it possible to slashdot slashdot? by Krashed · · Score: 1

    what does slashdot use for their servers? the site uses a lot of scripting yet rarely has any problems. i know the site runs on slashcode but what is the hardware behind the software?

    1. Re:is it possible to slashdot slashdot? by Zog+The+Undeniable · · Score: 1
      It's an old Packard Bell multimedia tower with a Pentium 75, 16MB of EDO RAM and a 14.4k modem.

      Or RTFAQ ;-)

      --
      When I am king, you will be first against the wall.
    2. Re:is it possible to slashdot slashdot? by Anonymous Coward · · Score: 0

      I've thought about that a few times... We're serving more than two billion pageviews a month... if I put a text link to slashdot below the fold on any of my pages it would bring slashdot down INSTANTLY. But unlike the slashdot editors, I have a smidgen of professionality, and would never do that.

  109. Attention.... what? by Mike556 · · Score: 1

    Observation:

    I barely even made it to the end of this article, and then this crack about slashdotter's attention span actually caught my attention.

    Why do we glaze over things so often? There are several possibilities.

    1)Some people will only read information, even when it is about interesting material, if it pertains directly to them. In this instance, someone who doesn't do anything with web hosting might not see it as relevant to them - I think it's interesting, and I don't run any webserers, but....

    2)There is so much information being provided to us in so many different formats every day that we can't help but skim through things a bit sometimes. If we were to take time to read into every interesting tidbit of info that passed our eyes every day we'd either be glued to a computer (or newspaper, or what have you) or go crazy. I've already gone crazy, but there's still hope for most of you out there!

    ~Mike
    1AM on not enough caffeine makes ya a little nuts...

    --
    Mike Rizzo
  110. Exponential Falloff by Anonymous Coward · · Score: 0

    This exponential falloff is very similar to what I experienced when my site was linked to from ilovebacon.com

    I wonder if this is typical for most slashdot links?

    (p.s., this is my web site)

  111. what about the second peak by romit_icarus · · Score: 1

    There's a small peak (if you can call it) 9 days after the story breaks. Anyone has a clue why?

  112. Data by linuxizer · · Score: 1

    Similar pattern without the bump a little while later. Bandwidth usage in MB by day (the /.ed page was 225KB): Jul 11 2003 1.35 Jul 12 2003 0.08 Jul 13 2003 0.48 Jul 14 2003 0.3 Jul 15 2003 1.24 Jul 16 2003 3745.91 Jul 17 2003 432.48 Jul 18 2003 38.09 Jul 19 2003 3.29 Jul 20 2003 7.27 Jul 21 2003 3.49 Jul 22 2003 2.04 Jul 23 2003 0.84 Jul 24 2003 1.28 Jul 25 2003 1.29 Jul 26 2003 0.66 Jul 27 2003 0.53

  113. Pretty simple...Web "Dole" for president. by Anonymous Coward · · Score: 0

    "That way, you only have (let's say:) 10 users loading the bandwidth-intensive pics at any one time, and everyone else is looking at a simple error page that automatically trys to re-load the normal page every minute or so."

    Well you're going to have to stagger that then. How about putting them in a client queue? The client that hits an error page gets a timer cookie, and a bit of javascript. Basically the user sees a page that says "Your user # so-and-so. Your wait will be #". The geeks here will love it. "Hey everyone the machine is up to this many users." It reduces the number of repeated click, click, click, and may even reduce the load as the impatient leave.

  114. TUX is another soultion by linuxperformer · · Score: 2, Informative
    One can effectively employ the TUX Web server on such situations. Especially if the pages to be served are entirely static in nature (TUX can manage dynamic content as well, though not its forte), TUX is the simplest solution. We once tried to serve the State Secondary School exam results here in Kerala, India using this technique and the performance was pretty impressive. It was accomplished with a P IV Compaq desktop PC with 256 MB of RAM. A write up is available at Tux@Play.

    So if your site is slashdotted, churn out a static version of the pages which are likely to be pulled most and hand them over to a TUX server. Sit back and enjoy the traffic!

  115. Yes but by phorm · · Score: 1

    Don't worry, the editors have got it covered.

    This article will be up and running fine again for Monday's dupe.

  116. Unslashdottable by mkweise · · Score: 0, Offtopic

    If you're like me and visiting the unslashdottable siteeven downloading the MPEGsleaves you feeling powerless and depressed, here's a server that you can easily get to barf.

    --
    Gentlemen! You can't fight in here, this is the War Room!
  117. exponentially decaying attentions span by bob_jenkins · · Score: 1

    I saw something similar from the BBC when they linked to one of my pages last September.

  118. Why servers get /.'ed by Zog+The+Undeniable · · Score: 1

    Half the time they just run out of the daily bandwidth allocated to them by their ISP, especially if they're on a free web host like Geocities. That's not a technical problem, it's a financial one.

    --
    When I am king, you will be first against the wall.
  119. kernel dot org by diz · · Score: 1

    For what it's worth, the only times ftp.kernel.org, www.kernel.org, or mirrors.kernel.org have gone down that I can remember have been hardware failures or raid re-org. There is no redundant server -- all those are on a single host, yet somehow it manages, and every new kernel relese seems to appear on slashdot for whatever reason (like the recent 2.6.0-test tree article on slashdot). Then again, maybe having tons of mirrors helps, but there are still 12000-25000 hits per hour on average and served 489GB in July. Don't see a lot of news in this story. Of course redundancy is better, but there is good proof in the world that it's not strictly required.

    1. Re:kernel dot org by SuiteSisterMary · · Score: 1

      Most 'slashdottings' are either a) saturation of the bandwidth allocated to the server, which isn't very exciting, or b)misconfiguration/over-reliance on dynamic content resulting in a connection to a database barfing.

      --
      Vintage computer games and RPG books available. Email me if you're interested.
  120. Compilation defaults by zdzichu · · Score: 1

    My server was once slashdotted. It's Duron with 256 RAM and 100mbit connection. Hardware wasn't a problem. When I tried to increase maximum clients limit to 512, Apache informed me by the logs, that maximum clients count is 256. To increase that limit, one have to change one define in config.h and _recompile_ Apache. You have to think about slashdotting while putting server on. Changes in configuration files often do not help.

    --
    :wq
  121. We've been Drudged.... by Anonymous Coward · · Score: 0

    We dont even use the term slashdotted here..... a referral from the drudgereport usually generates 2-3 times the traffic slashdot does.

  122. half-life of the falloff... by SharpFang · · Score: 1

    ... and its peak value depends on the slashdotting site...

    I have my website (Don't troll down! That IS my website! Don't visit if you don't like it!) "slashdoted" from time to time, whenever I post the URL to the petlovers forum. (themed board, people may find it offensive!) It drops halfway within a day or so, from 400+ hits/hour, and within a few days I get down to normal flow of 1-3 visitors/hour. But when I posted my URL to the story board (same notice as above) it got maybe 200 visitors after submitting, but then the dropoff lasted weeks. Half-life of the dropoff was about one week later.

    Stats (finally no zoophilic references) here
    Look at the monthly graphs. (blue ones) and at the last yearly graph which shows the "normal flux" dropdown as the site "expires from people bookmarks" and gets generally forgotten in matter of months.

    --
    45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B2
  123. Ask for /. permission by Anonymous Coward · · Score: 0

    Hell, I hope that someone with as little bandwidth as my site (http://durandal.homeunix.org/)never has to endure something like that. I think that possibly a script should be generated for the slashdot servers that automatically does a quick load test to determine if /. will take down the server, and if it does, automatically mirror it so the site doesn't get quite as slammed...

  124. Re:Monty Python Guide! by Anonymous Coward · · Score: 0

    Moderators get pissy when they don't get your jokes.

  125. Help needed, please! by pmz · · Score: 1

    One is to buy a studly, fire-breathing DB server that can process requests faster than your web servers can send them.

    Okay, I've spent the last several years breeding my race horse and dragon hybrid abomination of the animal kingdom and plugged in the Ethernet cable...but nothing happens! I need help! Thanks in advance.

  126. Re:I just don't understand it. Really. by pmz · · Score: 1

    I just don't understand it. Really.

    That's because you're one of those goofy people that realizes CPUs are damn freaking fast and that oodles of rack-mounted SMP servers, J2EE and Oracle out the wazoo, and a team of inexperienced developers aren't needed for "enterprise" websites.

  127. I don't get it... by Anonymous Coward · · Score: 0

    I don't understand the slashdot effect anyway. How are the servers brought down when NO ONE READS THE STORIES?

  128. Bit Torrent by Anonymous Coward · · Score: 0

    I strongly recommend setting up a Bit Torrent tracker for small sites. If you plan on having a big file, such as a demo, or a movie, a bit torrent link will go a long ways in avoiding a slash-dot affect. It will also make your ISP happy, and may potentially save money if you have to pay for bandwidth. Bit Torrent is the perfect technology for something that is large, and has a massive, immediate demand that trails off quickly over time.