Slashdot Mirror


Traffic Shaping on DSL?

jackla asks: "I'm now looking for software to do traffic-control on my Windows XP box. I am connected with DSL and my upstream is capped at 96kbit/s (down is 1.5Mbit/s) - this means that high(>70kbit/s) upstream utilisation KILLS my downstream: it just drops down to about 400kbit/s and stays there unless there's more upstream space. That said, I read alot about the Linux shaping solution (wondershaper or something) which sounds exactly right, except I need something that works for Windows. What I want to do is prioritize upstream ACKs (for example) so that my downstream isn't affected by upstream use. If anyone heard of a peace of software that can do this, I would love to hear about it." It would be nice if something like this existed cheaply for Windows. I am unaware of such, but maybe a few of you have ideas. Could such a traffic shaper be built using low powered computers? If so, how would you build and configure it so it would maintain compatibility for the single Windows machine, behind it? (Think: homebuilt traffic-shapping appliance)

368 comments

  1. Comment removed by account_deleted · · Score: 0, Funny

    Comment removed based on user account deletion

  2. What are you running? by secondsun · · Score: 5, Interesting

    Are you a server, business or home user? I understand that you want as much bandwidth as possible, but if you are just a home user, 400 kbit down stream is not bad at all.

    To answer your question directly, my solution would be to buy a cheap box (like say, the Mandrake boxes from Wal-Mart) and use it as your traffic shaper. Linux products for this are much cheaper than any (useable) solution you can find for windows.

    --
    There is nothing wrong with being gay. It's getting caught where the trouble lies.
    1. Re:What are you running? by DrVxD · · Score: 4, Insightful

      > 400 kbit down stream is not bad at all.
      Unless you're paying for 1.5Mbit (which it sounds like he is), in which case it sucks. 400k isn't bad, but it's only about a quarter of 1.5M. That's a BIG difference if you're transferring a large file (e.g. a movie), and is even more apparent if you're trying to (e.g.) stream video.

      --
      Not everything that can be measured matters; Not everything that matters can be measured.
    2. Re:What are you running? by Col.+Panic · · Score: 1

      More like buy a cheap box like a PII or even Pentium (like $100). One beauty of Linux is you don't need a P4 to run it, especially as a router.

    3. Re:What are you running? by Snoopy77 · · Score: 2, Informative

      If you are going to go with a linux box to do the traffic shaping just pick up an old Pentium/Pentium II from somewhere. You don't need a monitor or extra gadgets, just a box with appropriate networking capabilities. You could even borrow the mouse and keyboard from your computer to help get it up and running. Once it is all configured you should be able to just tuck it away in a nice cool place and forget about it.

      --
      "She's a West Texas girl, just like me" - G.W Bush Iraqis
    4. Re:What are you running? by CoolVibe · · Score: 2, Informative
      Maybe he should look into switching ISP's. The cool thing with *DSL is that you actually have a choice about who your ISP is. Take the sitation where I live in .nl for example, where cable internet is restricted to one ISP that can totally screw you over and leave you groaning and paying the bill. Signing up for DSL is the perfect way to give those scamming crappy cable ISP's the middle finger and to tell them to stick their cable where the sun don't shine. But that's just the situation here in the low lands. I am digressing...

      If there's an other ISP that gives you high bandwidth with xDSL that does capping in a reasonable way (i.e. I'm capped at 256 kbit upstream while having 1 Mbit down) I say switch.

    5. Re:What are you running? by rmadrid · · Score: 1

      ??? I'll try to answer the initial question, which occurs at network layers lower than IP. ADSL is usually supplied ethernet 10Base-T with ATM. Traffic shaping can be applied at the ATM(WAN/Copper) part of the modem, but this is up to the service provider. Sure, you can shape your ethernet (LAN) component using a box (AppVantage Packeteer etc.etc.) but the functionality is already there on most ADSL boxes.

    6. Re:What are you running? by acoustix · · Score: 2
      Unless you're paying for 1.5Mbit (which it sounds like he is), in which case it sucks. 400k isn't bad, but it's only about a quarter of 1.5M.

      I don't know of ANY DSL providor that garauntees 1.5Mbit (you probably meant mbit). All of the ads that I see say "up to 1.5mbps". Depending on his distance from the CO it might be a good speed.

      --
      "A plan fiendishly clever in its intricacies"- Homer Simpson
    7. Re:What are you running? by ColaMan · · Score: 2

      I don't know of ANY DSL providor that garauntees 1.5Mbit (you probably meant mbit). All of the ads that I see say "up to 1.5mbps". Depending on his distance from the CO it might be a good speed.

      Telstra in australia will not give you an aDSL line if when tested it can not reach 1.5Mbps.
      Which is a bit of a pain, as no 1.5Mbps, no aDSL.

      Moves are afoot by InterNode (another .au ISP to get Telstra to change this to something lower, so that people can still get service even if they can't get 1.5Mbps). As all aDSL work is done by telstra (they pretty much own the last mile of copper in .au) they effectively can do what they like.

      BTW by poking around my alcatel modem , it tells me that my line is good for max of around 6.8Mbps downstream to me. I know someone mentioned once that in Hong Kong they could get 6Mbit / 1Mbit adsl connections for relatively cheap. Bastards.

      (Oh, btw it should be big M , for Mega, little m is milli)

      --

      You are in a twisty maze of processor lines, all alike.
      There is a lot of hype here.
    8. Re:What are you running? by faaaz · · Score: 1

      Uhm, big M = Mega, small m = milli therefore 1.5mbit = 1.5 millibit, and I don't think that's even possible.

      --
      we come in peace / shoot to kill
    9. Re:What are you running? by Anonymous Coward · · Score: 0

      Heh... it sounds like you need a PacketShaper appliance.

    10. Re:What are you running? by packeteer · · Score: 2, Informative

      Speaseasy is my ISP and althoug its not high upload its everything else... you get fast enough down and they are what a true ISP should be... they cost a little extra but *GASP* they give linux support... call em at 3:00AM and they have a linux techie waiting for you with no wait time... even during peak hours its only maybe 5 mins to get a person on the phone... personally for me internet is more than your "up" and "down" speeds... im tired of people wanting raw kbit speed when thats not what they even really use... low latency is most of the time a bigger issue as a high down really only helps thsoe divx files come down... when your latency is low your games run smooth and the pages load faster... is 512kbit going to feel slower than 1.5mbit when its a 10k page???

      --
      unzip; strip; touch; finger; mount; fsck; more; yes; unmount; sleep
    11. Re:What are you running? by tylerdave · · Score: 1

      Speakeasy is great. Unilke other ISPs they understand that you may want to run a server from home and therefore don't have any wording against that in their TOS.

    12. Re:What are you running? by Anonymous Coward · · Score: 0

      Then maybe he should complain to his ISP for offering 1.5Mbit down and 70Kbps up! That's absolutely insane. Why not replace the whole thing with a television set and settle back into the old consumer being delivered information mode?
      I have 256Kb/s up on my 1.5Mbit down DSL and it's horrible. Whoever invented ADSL needs to be pistol whipped. All it accomplishes is turning broadband customers into consumers rather than participating nodes on the network. The Internet is successful because of the vast diverseness of information out there and not because 3 or 4 huge multinational media conglomerates deliver all the web traffic to everyone like they do for television. Do NOT let them turn the Internet into television by keeping these fucked up asymetrical speed patterns people. Complain at ALL times that you demand they bring you SDSL or symetric bandwidth!!!

    13. Re:What are you running? by TheMostBob · · Score: 1

      Sure it is - 666 seconds per bit....bandwidth of the Beast. :-)

      --
      -- Bob
    14. Re:What are you running? by essdodson · · Score: 1

      None of this thread is relevant to the question. What he's experiencing is a signifigant drop in performance when his upstream nears the cap. I experience the same thing. If you can't ack packets then the host will not send anymore and it becomes a problem.

      --
      scott
    15. Re:What are you running? by jafuser · · Score: 3, Funny
      1.5 millibit/s equals 666+(2/3) sec/bit? I'll take a 110 baud modem over that ;-)

      1.5 millibit/sec - I wonder how many Libraries of Congress per microfortnight that'd be? hmmm... Assuming the most often figure that 1LOC = 10TB, that's

      806.4 microfortnights/bit
      = 8,866,461,766,385,664 microfortnights/LOC
      = 1.127845612317619342652578202505e-16 LOCs/microfortnight

      BTW, I prefer LOC/attoparsec over the more frequently used LOC/hectare for storage density measurements.

      --
      Please consider making an automatic monthly recurring donation to the EFF
    16. Re:What are you running? by Megane · · Score: 2
      It's not his ISP. Until a recent upgrade to faster DSL (I haven't had it long enough to get a feel for it yet), I had 1.5M/128K DSL, and full-speed uploads using Hotline would completely hose other connections. Before that I lived where I could only get 384K/128K, and full speed downloads could completely choke the connection too.

      One of my objectives now is to look for servers (ftp, http, etc.) which can inherently limit stream speed, but the "prioritize ACKs" suggestion by the submitter sounds like something I can do with ipchains/ipfilter on my router server after some quality time with TFM.

      --
      #naabhaprzrag, #sverubfr-000, #agi-fcbafberq, negvpyr[pynff*=' negvpyr-ary-'] { qvfcynl: abar !vzcbegnag; }
    17. Re:What are you running? by Anonymous Coward · · Score: 0

      Get a cheap box off of ebay. I got a pentium 166 that is plenty good for a router/fw/traffic shapper. I added a compaq dual port 10/100 pci card and some extra ram, it is great. I have less than 50bux in the whole deal, and its a whole lot cheaper than a box from walmart. It does not take a whole lot of cpu to do this, so a new box of any kind would be overkill. Heck, even a good 486 would work

    18. Re:What are you running? by sargon · · Score: 1
      That will not help. The issue is partly with the provider and partly with the provider of the local loop, the ILEC, which, typically in the U.S., is an RBOC. ILECs provide only what they are required to provide by contract. If the service you purchase from another provider does not guarantee bandwidth, you do not get that 1.5Mb. If you sign up for guaranteed bandwidth but your provider does not contract with the ILEC for that service, you do not get guaranteed bandwidth.

      If you read the very-fine print which comes with the TOS, you will notice that most ISPs do NOT guarantee the bandwidth for the typical home-user DSL service---unless you sign up for a "guaranteed bandwidth" service, which, of course, is extra.

      "Guaranteed bandwidth" service is an extra service from all ILECs. They pass that cost on to all providers/resellers who get service from them. That is why you can get "1.5MB DSL service" for $49.95, but a "guaranteed bandwidth" service for 512K is $79.95.

      Keep in mind that the ILECs don't want to be in this service in the first place. They were forced into opening access to competitors by the 1996 Telecommunications Reform Act. That hurt their revenues, since that $2,000 T-1 was no longer something many companies would consider in the face of a $100 DSL circuit from a CLEC/DSL provider. And the home user has always been a headache, something to be avoided.

      Disclaimer: I have worked for three telcos.

    19. Re:What are you running? by Anonymous Coward · · Score: 0

      wow, how did I know the top ten replies would be the cookie-cutter "use linux" replies?

      Tow that Linux Party line!

    20. Re:What are you running? by jackla · · Score: 1

      I'm from Israel actually and the caps are: Down:1.5 * 2^20 * bit Up:96 * 2^10 * bit Jackla

    21. Re:What are you running? by Archfeld · · Score: 2

      Aye low latency and a FLAT router path to the backbone.

      Do a tracert and see how many hops you make...

      Try looking at an SDSL connection and solve the asymetric problem :)

      --
      errr....umm...*whooosh* *whoosh* Is this thing on ?
    22. Re:What are you running? by Anonymous Coward · · Score: 0

      I hate people like this.. just spreading FUD. Read a book or something, and then talk.

    23. Re:What are you running? by zoombat · · Score: 1
      I don't know of ANY DSL providor that garauntees 1.5Mbit (you probably meant mbit). All of the ads that I see say "up to 1.5mbps".

      I think your confusing MB/Mb with Mbit/mbit:

      MB = Mega Byte
      Mb = Mega bit

      They get confused a lot. Most people *I've* seen, say "Mbps"; "mbps" might get listed if the advertiser isn't paying attention to detail or doesn't know what the heck they're talking about...

    24. Re:What are you running? by packeteer · · Score: 1

      Speakeasy brags about their tracert results... its 2 hops to the backbone... that simple... their network is streamlined and efficient like a REAL ISP real be... Spekaeasy SDSL packages are sweet if you can pay that much... with rpmfind services hosted on their network you can get all your redhat/drake/whatever distro rpms at blazing speed...

      --
      unzip; strip; touch; finger; mount; fsck; more; yes; unmount; sleep
    25. Re:What are you running? by packeteer · · Score: 1

      they actually encourage home servers and sharing bandwidth... they give multiple static IP's and untill they run out EVERYONE who signs up get a static IP even if you pay less for a dynamic... they give extra bandwidth extra ip's and extra support... great stuff... personally i dont run servers from home other than personal ones because speakeasy runs all the rpmfind/Q3/DOD/CS/RTCW servers you couls want...

      --
      unzip; strip; touch; finger; mount; fsck; more; yes; unmount; sleep
    26. Re:What are you running? by critter_hunter · · Score: 1

      I don't know about the rest of the world, but here switching ISP for DSL is totally useless since the other ISPs just buy the lines from Bell. So you always get the same service with the same dl/ul caps no matter who you're with. The problem is that both cable and telephone are pretty much monopolies around here, so there's no way to get a good priced high speed connection.

      Hell, last month the DSL started limiting upload and download quantities AND went 5$ up at the same time. That's when I decided I didn't need high speed enough to keep supporting the fuckers.

      --
      Karma: Could be worse (could be raining)
    27. Re:What are you running? by Archfeld · · Score: 2

      thanks..I am on Megapath SDSL with covad now, but Megapath service has started to deteriorate severely in the last month and they refuse to address the problem. I am gonna be leaving them with both connects.

      --
      errr....umm...*whooosh* *whoosh* Is this thing on ?
  3. hate to sound like a slashdot weenee by npietraniec · · Score: 0, Troll

    I hate to sound like a slashdot weenee, but use Linux. It's better suited for such things.

    Ok, you can all flame me now.

    1. Re:hate to sound like a slashdot weenee by Anonymous Coward · · Score: 0

      no doubt
      this dude is either very brave or very stupid for posting his email with a windows ask /.

    2. Re:hate to sound like a slashdot weenee by SirSlud · · Score: 1, Offtopic

      Or innocent. Honestly, I'm doing a poll, why AC?

      --
      "Old man yells at systemd"
  4. I don't get it... by neksys · · Score: 2, Insightful

    I have DSL through Telus in Canada with the same upstream cap, and high upstream utilization means a very, very minimal hit to my downstream speed. Perhaps it is more of a hardware issue than anything else? Or rather, something that you should be asking your DSL provider about instead of the general Slashdot community? *shrug*

    1. Re:I don't get it... by PFAK · · Score: 0

      Thats just a tad funny. I have telus and it reaps my bandwidth when I upload. My pings go sky high, and my download goes to like 20kb/s. Im kinda getting sick with Telus.. their services is hitting rock bottom. Even though my parents both work for telus, I cant stand their service :-/ pretty sad.

      --

      Free means no restrictions, ironic the FSF's GPL forces restrictions, isn't it? What's your definition of free?
    2. Re:I don't get it... by edrugtrader · · Score: 1, Redundant

      i also have a similar cap and run my website (down for a week weeks while i try and get tyan off their asses) off the line. this means i average a loaded 128kbits going out, but i can easily download at blazing fast speeds. i think your problem lies in hardware as well.

      maybe your provider directly hacked the modem they gave you, or your ethernet card or router isn't good at handling up and down traffic at once.

      --
      MARIJUANA, SHROOMS, X: ONLINE?! - E
    3. Re:I don't get it... by Anonymous Coward · · Score: 0

      Telus ADSL is 1.5M/512k. This poor guy has less than 1/5 you do.

    4. Re:I don't get it... by AnimeFreak · · Score: 2

      I think it is an area by area basis because I live near the Surrey-Newton co-loc and I can tell you DSL service is pretty good. However, if you go out to the Whalley co-loc, I have heard DSL is pretty shitty there.

    5. Re:I don't get it... by TFloore · · Score: 3, Interesting

      Don't know if it is his provider or his hardware, but my cablemodem connection does the same thing. And it seems to be dependent on the cablemodem hardware.

      When I first got my cablemodem connection, it was with an older hybrid cablemodem. Got about 1.8mbit/sec downstream, and about 500kbit/sec upstream, and I could use all of that upstream and it didn't affect my downstream at all.

      About a year after signing up for the cablemodem, my ISP "upgraded" their network, and I got a new cablemodem. DOCSIS 1.0 (1.1?) 3Com USRobotics Cable Modem CMX. It maxed out at about 2.2mbit/sec downstream, but only about 256kbit/sec upstream, and now, when I use all of the upstream, my downstream drops to about 350kbit/sec. (I say "maxed" because I think they dropped the max downstream to about 1.5mbit/sec 6 months ago, a year after I got the new cablemodem.) The computer connected to the cablemodem did not change at all, so this was purely from the cablemodem change.

      So I would also be interested in something that can prioritize the packet ACKs, since the cablemodem doesn't seem to do it itself anymore.

      I do tend to regard this as a hardware issue with the cablemodem. From reading this, you can probably understand why.

      --
      This is my sig. There are many like it but this one is... Oops. Frank, I've got your sig again! Where's mine?
    6. Re:I don't get it... by PFAK · · Score: 0

      Yeah. Im in the Coquitlam-ish area. Its supposed to be good, but even people closer get slow. I mean, and its always dieing on me. They have been degrading, I used to get like 192kb/s, now I only get 90kb/s on average.. Also they blocked port 80 (incoming) which kinda ticks me off.

      --

      Free means no restrictions, ironic the FSF's GPL forces restrictions, isn't it? What's your definition of free?
    7. Re:I don't get it... by TheDick · · Score: 1
      When I first got my cablemodem connection, it was with an older hybrid cablemodem. Got about 1.8mbit/sec downstream, and about 500kbit/sec upstream, and I could use all of that upstream and it didn't affect my downstream at all.

      No fucking way, Hybrids use Telco return, and are usually 14.4 modems (sometimes 28.8) Perhaps your old modem was a Terayon or something.....

      --

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

      I have had dsl since it was offered in my area, (i'm in canada as well), and I know the pains of a slow downstream if someone is downloading off of me (downloading off of me - ie utilizing my upstream). Perhaps the poster should look more at the application layer of things, what exactly is taking away from your upstream. FTP? Kazaa? Web Server? Each of those programs offers bandwidth limiting in some form.. Kazaa you can specify how much bandwidth can be used by people downloading from you, FTP and Web servers as well. I'm sure if you look into an alternative way of sharing your data on the net, you'll find some application that will do the bandwidth limiting for you.

      Just a thought.

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

      Your upload cap will be between 500k - 850k, not the 96k he's capped at.

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

      Are many of us from BC Canada, I just saw two posts from the Lower Mainland with no reference of BC at all....

      Cheers

    11. Re:I don't get it... by jackla · · Score: 1

      I am using Alcatel SpeedTouchHome(sorry for the caps :) it does the PPPoA connection to the ISP and I connect to it via ethernet.

    12. Re:I don't get it... by Robert+The+Coward · · Score: 1

      As some who use to support those crappy Hybrid Cable Modem From RCN I can tell you there are 3 Types 1 with Telephone return path built in at 14.4 2nd was with an Extern 33.6 Modem due to phone lines ran mostly at 28.8 and 3rd Style was 2-way with went both ways thought the cable line. Hybrids had major problems most where the fact that we well replacing them every other month and sending them back for repairs all the time. Altought some custmers loved them with the nears static IP address. As for the change in bandwidth his area is probly using more overall bandwidth so he is getting a smaller pice of the DS3 that RCN puts in it area.

  5. Wow! by Anonymous Coward · · Score: 0

    I think this is the first Ask Slashdot I've read in the last 2 months where I haven't been as lucky as I feel.

  6. Router-on-a-disk by DrVxD · · Score: 3, Informative

    LRP's project might be of some use. Yes, it means getting a cheap box to run Linux, but then you can use all of that real neat networking software that's available for Linux boxes but isn't available for Windoze boxes.

    --
    Not everything that can be measured matters; Not everything that matters can be measured.
    1. Re:Router-on-a-disk by Anonymous Coward · · Score: 0

      LRP IS OBSOLETE!

      Try LEAF http://leaf.sourceforge.net/

    2. Re:Router-on-a-disk by Anonymous Coward · · Score: 0

      lrp's a little dated. go to http://leaf.sourceforge.net and check out one of the more current ones... i use dachstein on an old pentium 133 and it works great.

  7. Re:Cheapest solution by Anonymous Coward · · Score: 0

    Yuppers, this is the best solution when dealing with MS product. Someone buy this guy a beer or martini or whatever.

  8. Huh? by DreamingReal · · Score: 2
    Sorry if this is a stupid question - Why would 70kb/sec upstream utilization reduce downstream bandwidth by nearly 1Mb/sec? That makes absolutely no sense to me, but then again, I'm not a Network Engineer. Can anyone more knowledgable explain this?

    --
    We want some answers and all that we get
    Some kind of shit about a terrorist threat

    - Ministry
    1. Re:Huh? by Huogo · · Score: 1

      I can't explain it, but the same thing happens to me. Its like, if I'm using 90% of my upload bandwidth (128 kbit), I only get 10% of my download bandwidth (1.5 mbit). Its very annoying. I would be intersted in something like this for windows as well (I find it too annoying to reboot whenever I want to play a game)

    2. Re:Huh? by 1010011010 · · Score: 3, Informative

      TCP has to send back replies ("ACK"s, or acknowledgements) for each of the packets it receives, so that the sender knows if it needs to retransmit a packet. This is part of how transfer control is achieved by the protocol. So, if there's less space for ACKs going upstream, then the tcp traffic in the other direction can get slowed down.

      --
      Napster-to-go says "Fill and refill your compatible MP3 player", which is a lie. It's not MP3. It's WMA with DRM.
    3. Re:Huh? by DrVxD · · Score: 2

      The 70k (or whatever) upstream limit is based on actual bits transmitted across the network - including protocol bits. Since the 1M/sec downstream traffic has a certain amount of handshaking back to the source of the data, that handshaking has to travel back as part of the 70k upstream. If the handshakes don't get, then the downstream can slow or resend (net effect is a slowdown of actual throughput).
      Does that make sense? I know what I mean, but I don't think I'm communicating it very well :-(

      --
      Not everything that can be measured matters; Not everything that matters can be measured.
    4. Re:Huh? by Anonymous Coward · · Score: 3, Informative

      All DSL modems have a queue for outgoing packets. The queue will be a fifo queue, so once packets enter the queue there is no way to remove them.

      If you are uploading, then the queue on the router will be full of packets...

      If you are also downloading, then the ACK packets acknowledging the downloaded bytes will have to wait on the queue on the DSL router.

      The fact that these ACK packets are getting delayed will slow down subsequent data packets being sent, and the downstream transfer rate will drop as a result.

      By limiting the rate at which packets are sent to the DSL router, we can remove the queue from the router, and move it back onto a linux machine, where we can then prioritise the traffic as we see fit. ACK packets are allowed to skip to the head of the queue, so download speed can be maintained.

      Hope that explains things

    5. Re:Huh? by Anonymous Coward · · Score: 0

      A small upstream compared to a large downstream is a huge hinderance. When you use all your upstream, the ACK packets sent in response to a successfully received packet are delayed quite a bit. This results in a severely limited downstream, as you can not immediately tell the remote host that you got the last packet and are ready for the next. This can even lead to retransmission of the packets if the TTL on the remote host is set too low, effectively using more bandwidth for the both of you. I hope I did an adequate job of explaining this problem.

    6. Re:Huh? by BurritoWarrior · · Score: 3, Insightful

      Couldn't he reduce the number of ACK's by increasing his window size?

    7. Re:Huh? by marcushnk · · Score: 2, Informative

      Try this.
      20% of TCP trafic is the "reply" stream.
      so if you cut 20% of the reply you slow the downstream..
      Make any more sense?

      --
      "Consider how lucky you are that life has been good to you so far. Alternatively, if life hasn't been good to you so far
    8. Re:Huh? by DrVxD · · Score: 1

      Much more lucid than my explanation :)

      --
      Not everything that can be measured matters; Not everything that matters can be measured.
    9. Re:Huh? by Anonymous Coward · · Score: 3, Insightful

      I'm not a TCP guru, but I think this would potentially cause just as much of a problem. If, for example, he has a larger window, then when he has packet loss there will be a tremendous penalty. Instead of perhaps a number of small lost packets, he loses a number of _large_ packets. I could be wrong.. but I think this is one of those tradeoff situations.. catch-22 and all that.

    10. Re:Huh? by Anonymous Coward · · Score: 0

      Well AFAIK the window size is dynamic in that it will grow according to bandwith although yes there is a maximum window size. Large window sizes are only really useful where you have high bandwidth along with high latency, such as in satellite comms.

      I have a suspision that most TCP/IP implementations aren't optimized well for asymetric links (although I could be wrong).

    11. Re:Huh? by seanadams.com · · Score: 2

      Sorry if this is a stupid question - Why would 70kb/sec upstream utilization reduce downstream bandwidth by nearly 1Mb/sec? That makes absolutely no sense to me, but then again, I'm not a Network Engineer. Can anyone more knowledgable explain this?

      IAANE... here's why: TCP is always transferring data in both directions, even if you're only downloading. When your upstream link is full, you can't send acknowledgements of the data you're receiving downstream. These lost "ACKs" tell the transmitting side that somewhere along the path there is a saturated link, but it doesn't know in which direction they're being dropped. The transmitter will back off his data rate to try and reduce the packet loss, and this is what causes your slower throughput.

    12. Re:Huh? by DreamingReal · · Score: 1
      Perfect explanation. Now I understand. Thanks!

      --
      We want some answers and all that we get
      Some kind of shit about a terrorist threat

      - Ministry
    13. Re:Huh? by Anonymous Coward · · Score: 2, Informative

      You are correct. The problem that needs to be solved is how to insure his ack messages make it out, proioritized above any other traffic being sent. As for the example of streaming video though, that is not the case as video runs on udp, which has no ack, so this only becomes an issue with large file downloads...

    14. Re:Huh? by idealego · · Score: 1

      It's not a technical limitation (well not with my cable company anyways) it's something the cable company does to screw people who run programs like kazaa or any server programs for that matter.

      I've got both cable and dsl here, I've had cable for a several years and the cable didn't used to have this problem. All of a sudden it started happening where if you were uploading to people your download speed was terrible.

      My dsl connection doesn't suffer from this at all though, I can do max up/down at the same time no problems.

      This is using shaw cable and telus dsl in victoria BC Canada for anyone wondering.

    15. Re:Huh? by teknopurge · · Score: 0

      yeah, i know im a bitch, but this NOTHING to do with TCP sparky. it's the underlying DSL technology. I will bet a cold beer that this guy has ADSL, which along with all the other caveats it has, dynamically changes your amount of upstream/downstream bandwidth. and please dont cry, it's in your TOS. That's what you get for the 49.95/mo special. The solution: SDSL. A much stronger technology that uses a symetric transfer mechanism, as opposed to ADSL. why do i pay 109.99/mo for SDSL, 4 static ip's, and GAURENTEED LEVELS OF SERVICE?(uptime, bandwidth, etc) so i never have to cry like this....

      teknopurge

    16. Re:Huh? by PhreakOfTime · · Score: 1

      I have noticed similar results here in the chicagoland metro area. I used to have service through AT&T broadband(joke), then for numerous reasons I switched to DSL. The main reason was AT&T would not allow 'web servers' to run, and I figured why do I want the company Im paying for service decide what I can and cannot run?
      So the switch to DSL. I can u/l and d/l my max limit with no problem on DSL all day if I want. The posts claiming ACK packets will slow transfers down, while correct, are horribly inaccurate. If Im transfering bi-directionally then the miniscule packet size of an ACK is dwarfed by my MTU packet size. Sure there may be a .05% drop in 'actual' data transfer, but I see no NOTICABLE speed change in a d/l.
      And to repeat my experience with the 2 technologies and what each one said to me:

      Cable - "All we do is provide web access and mail service, anything else is not in your contract"

      DSL - "All we provide is the pipe, any network services are your responsibility"

      For the comparable price, Ill take the non-motherly approach to my internet service...

    17. Re:Huh? by leo_fischer · · Score: 2, Interesting

      I believe that it is due to ADSL's asymmetric nature. An ADSL connection has an upstream and downstream speed, such as 1.5MBps vs 128kbps. I think the connection is NOT duplex, so when you are sending upstream, you cannot be recieving downstream. So if you upload at 64kbps, you are effectively using HALF your bandwidth, limiting your downstream speed to 750kbps. i.e. if you are transmitting half the time, you can only recieve half the time. So if you use too much of your upstream bandwidth, you can bring the downstream to a crawl.

    18. Re:Huh? by Buck2 · · Score: 1

      I like how you got -1 for explaining the problem correctly. If you were to ask my DSL provider, as I have, about this issue, they would respond the same way, except with a little more tact.

      I guess you need that (tact) to be modded fairly around here.

      --

      As my father lik@(munch munch)... ....
    19. Re:Huh? by adamjaskie · · Score: 1

      While that is quite true, I have had quite a lot of success increcing my window size. I have not seen very many, if any, problems with packet loss from my cable modem. I find that increcing the window size on all my computers dramatically improves the time it takes to load pages from the internet. That said, I suggest he purchaces a cheap box (say, a high end Pentium, Pentium Pro, or low end Pentium II) and install a very customizable distro, such as Slackware or Debian, without X or any of that crap. Then, set it up as an NAT box, and put the traffic shaping software on it. Thats what I do (although I dont use traffic shaping software) and it seems to work just fine.

      --
      /usr/games/fortune
    20. Re:Huh? by Spazmania · · Score: 1

      TCP has a "window size". This window is how many bytes of data it can send ahead of an acknowledgement that the data has been received. The maximum the protocol allows is 64kbytes. If the full upstream queue delays those ACK packets by, say, one second then the maximum possible downstream speed for that TCP stream will be 64kbytes/sec or 512kbps.

      This has major implications in high-speed cross-country networks where speed of light delays calculated against the window size cap the transmission speed well under the media's maximum capacity. Its also a significant problem running TCP over geosyncronous satellite links where the round trip speed-of-light delay is half a second.

      As I recall, MS Windows further caps the TCP window size to 8kbytes. Increasing it won't help since this'll just clog the upstream faster.

      If you're too badly clogged, packet loss becomes an issue. The TCP protocol implements congestion control by halving the window size every time a packet is lost. Guess what impact this has on your speed.

      --
      Moderating "-1, Disagree" is simple censorship. Have the guts to post your opinion.
    21. Re:Huh? by monkeydo · · Score: 1

      Not only that, but the sender will resend all of the unacknowledged packets making the problem even worse. Drop enough acks and you lose your session altogether. The solution? UDP! What, you don't want to use TFTP for all of your ISO's?

      --
      Si vis pacem, para bellum
      The only thing more annoying than a Libertarian is an (un|mis)informed Libertarian
    22. Re:Huh? by Dahan · · Score: 2
      I like how you got -1 for explaining the problem correctly.

      No, check the moderation stats for his comment--his comments start at -1 because his karma sucks. And his karma sucks because he's a troll... he didn't explain the problem correctly at all.

    23. Re:Huh? by Zero+Sum · · Score: 3, Informative
      If, for example, he has a larger window, then when he has packet loss there will be a tremendous penalty. Instead of perhaps a number of small lost packets, he loses a number of _large_ packets.

      Increasing the window size will not change the packet size. The packets requireing retransmission will retransmitted irrespective of the window size.

      There are two things he could do that might improve things, increase window size which means that the sender will not require an ACK so often or increase packet size so not as many ACKs are needed in the first place. Both may help but don't confuse them.

      --

      Zero Sum (don't amount to much). [root@localhost]

    24. Re:Huh? by Anonymous Coward · · Score: 0

      Being employed for said ISP, I'll try to explain it to you with out violating our NDAs.

      What your experiencing is most likely due to upstream limitation on the modem. This will delay ACK traffic in a queue on the modem. You can set your TCP Recieve Window size higher which will help as less ACKs will be sent.

      But something like a traffic shaper would definately help I think, I'm thinking of setting one up myself.

      Also, Terayon encrypts their traffic which when you upload means alot of CPU horsepower on the modem.

      Also what may have happend is your modem was reset from a no upstream limit to a specified limit.

      If the goal was to stop Kazza/ Morpheus / Gnutella traffic it would be more intelligent to put QoS on the core routers on specific ports.

      I have been told that most of the slow down in download speed is due to the load encryption places on the modem.

      One strange thing I have noticed is that when pinging someone on a terayon modem that increasing the packet rate over 1Kbytes / sec seems to drastically increase the amount of time it takes to transmit. It seems to move in a linear fashion indicating to me that it is infact due to encryption.

      This is why I think that a traffic shaper could definately help because it would remove the 400 ms delay to encrypt a large packet. And spit the small ACK packet out almost immediately.

    25. Re:Huh? by Buck2 · · Score: 1

      I've never figured out how to check the moderation stats except in "Parent".

      I apologize for blaming a moderator for marking him down when it didn't happen.

      I still think he got the explanation right. ADSL is probably the issue.

      --

      As my father lik@(munch munch)... ....
    26. Re:Huh? by Dahan · · Score: 2
      I've never figured out how to check the moderation stats except in "Parent".

      Click the (#3914122) in the header. You'll see he has one Underrated moderation and is currently at Score: 0, so he must've started at -1.

      I still think he got the explanation right. ADSL is probably the issue.

      Well, maybe he got it "right" in a trivial sense... but not in any useful sense.

      Paraphrase of poster: "I have a DSL line where the upstream and downstream speeds are asymmetrical. Whenever I upload, it kills my downstream bandwidth because the ACKs aren't getting through fast enough. What can I do to fix this?"

      Paraphrase of teknopurge's reply: "HAHAHAA!!1 The probelm isnt TCP dumbass! teh probem is that U SUXX0RZ cuz u have ADSL!! I M so l33t and have SDSL. Rawk!"

      I think it's pretty obvious the poster has ADSL; he basically says so. And sure, SDSL doesn't have this problem. That wasn't the question though... and the problem has everything to do with TCP and its behavior on asymmetrical links... there have even been a few papers on the subject.

    27. Re:Huh? by Buck2 · · Score: 1

      Thanks for the tip on finding the moderation history. As you've probably already guessed, he was at -1 when I responded.

      Well, maybe he got it "right" in a trivial sense... but not in any useful sense.

      In my lonely and foundationally incorrect universe he was modded down twice for presentation. As you've pointed out he wasn't modded at all at the moment I responded. Luckily for me I was then snapped back into the real world.

      Paraphrase of teknopurge's reply: "HAHAHAA!!1 The probelm isnt TCP dumbass! teh probem is that U SUXX0RZ cuz u have ADSL!! I M so l33t and have SDSL. Rawk!"

      I find this humorous. I would have modified it as +1 funny and +1 informative. *shrug*

      I found the same information that the original poster wanted by googling and contacting my ISP when I had the same question (SDSL _can_ solve this problem). I think that if someone has gone through the same effort at some time in their life, which it appears teknopurge did, that they should be able to vent every once in a while.

      How would you have modded me if I would have wrote:

      I just ate a pile of plastic letters and I'm going to shit out the solution to your problem:

      ERGGGHHHH!!!

      *squeeze* S ...

      *squeeze* D ...

      HNNNNHH!!!!

      *squeeze* S ...
      *squeeze* L

      AAAHHHH!!!

      I feel better now and so will you.

      Informative (considering it solves the problem) AND inflammatory (ala teknopurge). Does it deserve -2? ... probably not.

      --

      As my father lik@(munch munch)... ....
    28. Re:Huh? by Dahan · · Score: 1
      How would you have modded me if I would have wrote: [...]

      Well, I think I've been $rtbl banned from moderating or something, seeing that I've never been given mod points since that big /. controversy about editors having unlimited mod points that they use to -1 entire threads they want to suppress...

      but supposing I did moderate that post, I'd probably give it an Offtopic (I wouldn't take it below 0 though). The topic is "Traffic Shaping on DSL"; whether switching to SDSL would also fix the problem isn't relevant (unless it happened to be the only solution, which isn't the case here). And it might not even be a possible solution either... I have ADSL at home, but can't get SDSL. I can probably get a T1, but that's too pricey for my needs. Many people have cable modems, which are also asymmetrical, and can't get any form of DSL at all.

    29. Re:Huh? by Buck2 · · Score: 1

      Well, I think I've been $rtbl banned from moderating or something, seeing that I've never been given mod points since that big /. controversy about editors having unlimited mod points that they use to -1 entire threads they want to suppress...

      I get moderation rights pretty infrequently too, nowadays. I wasn't aware of a point wherein it might have changed, though. You're full of information. :)

      You also have many good points about the relevance of both my hypothetical and tekno(whatever)'s post. I think we're just going to have to disagree about that, though. I still believe that making a comment about a solution that obviates the whole problem cannot be labelled "offtopic". It might be "redundant" ... maybe. But who's to say that it's not valuable information? You're assuming that a casual browser of the conversations knows that this problem is limited to ADSL, and that the comments made by tekno(whatever) are so irrelevant as to be noise.

      I dunno.

      --

      As my father lik@(munch munch)... ....
  9. Absolutely hilarious by Anonymous Coward · · Score: 0

    Me and a bunch of my friends (XyberK00l, h4cKj00, and linUx0r) thought that was the funniest thing we'd ever seen. Then we started doing gay shit, y'know, just because.

  10. XP QoS by Anonymous Coward · · Score: 0

    XP already has QoS built in. How to configure it is beyond me. I don't think that you actually can, it's only there for applications that support it or something. Maybe someone else can shed some light on this topic.

  11. Build a Linux box, use QoS and cbq.init by fialar · · Score: 4, Informative
    My recommendation, keep that Windows box behind the firewall box (let the fw box NAT for it). Run Linux or BSD on the firewall box.

    If you're using Linux on the firewall machine, make sure you enable QoS and ALL the modules in it. Then grab cbq.init and set up the traffic shaping rules. The script file is well documented.

    -F-

    1. Re:Build a Linux box, use QoS and cbq.init by Anonymous Coward · · Score: 0

      What's QoS and how do you enable it and it's modules? Got a link?

    2. Re:Build a Linux box, use QoS and cbq.init by Wesley+Felter · · Score: 5, Informative
    3. Re:Build a Linux box, use QoS and cbq.init by Anonymous Coward · · Score: 0

      Uggh, CBQ is totally outdated, and hard to set up. Follow the instructions in Advanced Routing HOWTO. Set up qdiscs with TBF's going into SFQ. It'll let you prioritize acks (it bursts them), and all other traffic going up will slow down when they occur.

  12. Packet Scheduler by tbaggy · · Score: 1, Informative

    Instead of sarcastic comments about "run linux" "format c:" blah blah, check out this kbase entry from MS regarding the packet scheduler. It might be useful to you.

    1. Re:Packet Scheduler by Anonymous Coward · · Score: 0

      Okay, I've recently been forced to deal with the beast, normally, I'd use Solaris. Anyway, the QoS Packet Scheduler in XP is apparently installed and enabled by default in XP.

    2. Re:Packet Scheduler by bogie · · Score: 5, Informative

      Unfortunatly this won't help him. The QOS packet service is designed primarily for audio/video applications and would require that every router he connected through strictly obeyed the 802.1p protocol. That is just not going ot happen.It also requires a heck of a lot more then just clicking that little check box on the XP client. You can just consider that an on/off switch, the real work is done at the server and endpoint router levels.

      Have a look at this link for some futher explaination. http://infocenter.cramsession.com/techlibrary/geth tml.asp?ID=1674

      --
      If you wanna get rich, you know that payback is a bitch
    3. Re:Packet Scheduler by bogie · · Score: 2

      I forgot to add-

      In this case "run linux" or more appropriatly, run an additional machine that runs linux would be in all likelyhood the right answer for this situation. Other options are probably way too expensive.

      --
      If you wanna get rich, you know that payback is a bitch
    4. Re:Packet Scheduler by fermion · · Score: 2, Insightful
      So how is this any better? The question is traffic-control on my Windows XP box... not, as you say, linux box, or even the windows 2000 box in this KB article.

      There has been an extremely large number of trolls lately, and, as these answers suggest, very little useful information. This may in fact be a hard question, but must we compensate ignorance with stupidity?

      --
      "She's a scientist and a lesbian. She's not going to let it slide." Orphan Black
    5. Re:Packet Scheduler by phanki · · Score: 1

      It was a wonderful link given by you. I guess the guy wanted something like this. I still have few questions unanswered though. The person wanted to know how to keep the ACKs for the upstream going without affecting his downstream. Does the QoS scheduler provided by Win 2k provide an option to fine grain what you want. And any idea what is the algorithm that win 2k uses ?

    6. Re:Packet Scheduler by Ent · · Score: 0

      It would provide the control that you want by like the parent mentioned (although only half correct) the upstream router would have to be configured to accept this. Probably something this guys ISP wouldnt be willing to do.

    7. Re:Packet Scheduler by Anonymous Coward · · Score: 0

      No exactly, at least on XP. The QOS scheduler also effects traffic on your local machine. Windows will calculate your total bandwidth by profiling over time. It then allocates 20% of this bandwidth to each connection that doesn't request QOS. It allocates bandwidth over normal TCPIP by slowing down ACKs to remote sends. Sometimes this is nice because you can listen to streaming radio or TV while downloading a big file without it totally clobbering your "signal".

      So, you can benifit from XP QOS without needing any special upstream routers.

    8. Re:Packet Scheduler by Froqen · · Score: 1

      ugh.

      Psched is primarily about three things, marking, shaping and priority. Marking happens both at the layer 2 (802.1p) and layer 3 (DSCP). 802.1p is okay, but nowhere near a requirement for things to work. Most people who get value out of psched right now are just using the shaping aspects of it, or the nifty features done even when unconfigured. See http://support.microsoft.com/search/preview.aspx?s cid=kb;en-us;Q316666 . To use the shaping part see http://msdn.microsoft.com/library/default.asp?url= /library/en-us/qos/tcfunc_01et.asp .

      Again: the existing stuff in windows doesn't need/care about 802.1p, does some nice things for people on a modem and people who use ICS without any extra code/configuration, and with code is mostly usefull for shaping outbound traffic.

      P.S. If you think that psched is doing anything with 20% of your bandwidth, pls just don't even bother posting.

  13. Kick the leechers off your ftp server.... by whoppo · · Score: 1

    and you won't have a problem with downstream traffic :)

    --
    chown -R us /base
  14. The Open Source Meritocracy delivers by MrChuck · · Score: 0, Flamebait
    "News for Nerds, stuff that matters"
    not "information for the sheep of Microsoft"

    OpenBSD and NetBSD use ALTQ from KAME. Basic traffic shaping tool that can be used by the largest installs or the smallest.

    FreeBSD uses its "dummynet" interface that can also replace those $50,000 (US) "WAN IN A CAN" devices.

    If you configure it, it can add a 1000 mile lag and drop 5% of the packets to simulate a T1 between Ontario and Kansas City.

    Linux has stuff as well. I don't know it.

    Windows: Well, given that the source code is open to everyone and that the code base has always been based on a meritocracy: the only code that goes in are features that people care about, I'd expect it really soon. Hold your breath for it, ok?P?

    1. Re:The Open Source Meritocracy delivers by zoydoid · · Score: 1

      hey chuckie, WIN2000 and WINXP have the QoS Packet scheduler built-in.

    2. Re:The Open Source Meritocracy delivers by CoolVibe · · Score: 2
      Yeah, but they are only for audio and video streaming. Not really useful if you run your own webserver due to protocol issues.

      The best thing he can do is either:

      • switch to a better ISP
      • Use another box to run a unix based OS with a better suited QoS/packet scheduler/shaper combination on it and have that provide the routing/natting for his home network.
      The second approach might set him back a couple of bucks, but he'll get so much in return.The benefits outweigh the drawbacks.
    3. Re:The Open Source Meritocracy delivers by Anonymous Coward · · Score: 0

      People keep saying that.... But I dunna think it works the way you think it works. Read up on it, it's useless for this problem.

    4. Re:The Open Source Meritocracy delivers by teikan · · Score: 1

      No, they implement fair queueing in which each TCP stream can be (virtually) assigned its own output queue and equal share of uplink bandwidth.
      So if an upload stream and a download stream are filling the uplink, packets in the upload stream output queue are dropped first and those in the download stream output queue (mostly ACKs and small in number) are likely to survive.
      I am happily using ALTQ/WFQ on my NetBSD router connected to the cable modem in order to avoid this problem.

  15. Capping by BrianGa · · Score: 1

    What are you using your upload bandwidth for? Many windows programs, such as ServU FTP server, allow you to cap your upload speeds.

    Otherwise, perhaps something like this would be better for you.

  16. troll? by Anonymous Coward · · Score: 0

    "I'm now looking for software to do traffic-control on my Windows XP box"

    Sounds like a troll to me.

  17. DSL Reports... by Jeffv323 · · Score: 5, Interesting

    ... has a Tweak Test that tests some connection settings such as your RWIN. I had the exact same problem as you and it turned out that my RWIN was set wrong and once I fixed it, the problem pretty much went away. Try it and I bet it helps.

    --
    I'm a minister!
  18. Linux.... by I_am_Rambi · · Score: 2, Informative

    If you know the Linux program, why don't you just use Linux for the network controller? Have the Linux box be used as a gateway, running ipchains and the wondershaper. Linux is not that hard. Linux can also be run on an early machine. I am running Red Hat 7.2 on an AMD 133 with 48 mb of ram. I am running in command line, which is hard for a beginner to learn, but it is running dhcpd and Samba nicely. I would recommend Linux because security is easier to setup and maintain than Windows. Windows you will never know what ports are open or who is watching. Linux you can close ports alittle easier, much better for server (or routers/gateways). Why Live on Windows? Most of the programs you need for network admin are free.

    Linux, because Windows XP is eXPerimental!

  19. its time by slugo3 · · Score: 0, Troll

    ok, I think its time for windows.slashdot.org

  20. Re:Go cable, baby by Anonymous Coward · · Score: 0

    Actually, here in UK land, ADSL is better. Not because of the technology, but because of the people who have control over it (and their attempts at attracting customers). A normal ADSL gives 512down and 256up, but normal cable gives 512 down and 128up. Anyone using gnutella or hosting a little 'friends only' website would go for ADSL. If you want to spend more money, ADSL still gives better network performance.

    It is kind of sad that the cable guys, who have superior technology on their side have their bandwidth limited. Fastest possible for the home user on cable is still 1Mbps down.. but for ADSL it is 2Mbps down.

  21. first thing to find out is.. by trybywrench · · Score: 4, Informative

    what is going upstream from your pc. You can download an eval. version of Iris ( a sniffer ) and see the traffic originating from your pc destined to outside your network.

    You might find something crazy going on because a non-serving pc should be pretty quite. You will see broadcasts and ACK's but thats normal. If your computer is spewing traffic and you can't find the source your NIC could be off in the weeds or you may have been hacked (not uncommon with windows and DSL/Cable). I have 38 IP's in hosts.deny because of detected port scanning on my DSL.

    --
    I came to the datacenter drunk with a fake ID, don't you want to be just like me?
    1. Re:first thing to find out is.. by DrVxD · · Score: 2

      I'd take a wild guess that he's running some sort of P2P client (e.g. KaZaA, Direct Connect, Gnutella etc.) They can soak up a lot of upstream bandwidth (so I've heard :-)

      --
      Not everything that can be measured matters; Not everything that matters can be measured.
    2. Re:first thing to find out is.. by Kerbz · · Score: 2, Informative

      Well, of all of the recent P2P clients out there (that I have looked at), they all have bandwidth throttling built-in. Any decent ftp server has throttling built-in as well. FYI: Napster did not have bandwidth throttling, which I bet is the number one reason why many people didn't let it run 24/7.

      Moving on...

      After reading some of the comments here, I'm surprised to see so many people posting comments whom are confused by this person's problem.

      Experiencing slow web-browsing or slow downloads when upstream is saturated, despite having significantly adequate downstream, is easily explainable. The big picture is that servers are not going to send you more pieces of the file/webpage until you confirm that you received the last piece it sent to you. If your upstream is saturated, you cannot send that confirmation back as quickly, and the server is then simply waiting for you to "ask" for more

    3. Re:first thing to find out is.. by Anonymous Coward · · Score: 0

      There is also a registry hack to renew the Iris eval period. I wrote a batch file to run whenever it exprires so it never does. Not that I'm suggesting that sort of thing ...

    4. Re:first thing to find out is.. by Change · · Score: 1

      Also take a look at Ethereal, which is free, and quite nifty.

    5. Re:first thing to find out is.. by hansroy · · Score: 1

      Legitimate uploads can do the same. Whenever I update my website, such as putting a new photo album up, I just walk away. My downstream completely disappears, to the point 56k would be fast.

  22. Already There by BurritoWarrior · · Score: 3, Informative

    I believe XP comes with the "QoS Packet Scheduler" and has it installed by default. I believe it can do what you ask.

    1. Re:Already There by Anonymous Coward · · Score: 0

      Nope, QOS (Quality Of Service) is dependent on the TYPE of service, usually protocols that are highly latency sensitive like voice over IP or video. Usually these run on top of UDP, not TCP. He indicated the problem occurs when transferring files, likely with TCP protocols like FTP or HTTP.

    2. Re:Already There by Anonymous Coward · · Score: 0

      From what I understand, if your application does not support QoS and/or you do not have an Admission Control Service (ACS) server (like Win2k server) running for that packet scheduler to request bandwidth from (extremely simplified...) QoS does nothing.

      View what information I have about QoS here:

      http://www.blkviper.com/WinXP/service411.htm

      The particular service in question is called "QoS RSVP"

      -Black Viper

  23. Interesting concept by Anonymous Coward · · Score: 0

    this concept seems to be interesting, it is based on packet sizes, a simple linux-based router could do this.

  24. ECN and SACK...? by Anonymous Coward · · Score: 0

    Explicit Congestion Notification and Selective Acknowledgment would go a long way in abating this problem. Anyone know if it's supported in Windows XP?

  25. Dummynet by seanadams.com · · Score: 4, Informative

    Could such a traffic shaper be built using low powered computers?

    Fire up that 90MHz Pentium, install FreeBSD, and build a kernel with bridging and dummynet enabled. Dummynet is an awesome network simulator. Just set up a couple of ipfw rules for the types of traffic you want to limit, and then set the bandwidth parameters in dummynet. It's very easy to do basic stuff like you're describing, but you can do all kinds of other things with dummynet... latecy, loss, queue limits, simulating multiple hops and multipath links with different latencies. There are no tools of this caliber (let alone free!) for Windows.

    Next question?

  26. OpenBSD and ALTQ by ChrisMWage · · Score: 1

    OpenBSD (and a few of the other BSDs, I believe) has great support for ALTQ (alternate queuing) which does precisely what you want. (It's what I use it for).

    Go here for ALTQ.

    And here is my altq.conf, if you're curious.

    --
    --Chris http://chris.quietlife.net/
  27. I'm afraid.. by mindstrm · · Score: 2

    I don't see what you are getting at.

    Rather, I think I see what you are getting at, but I think the assumption is flawed.

    Full use of your upstream should NOT be crippling your downstream so much, that's not how it works. TCP should adjust accordingly.

    1. Re:I'm afraid.. by Anonymous Coward · · Score: 0

      > TCP should adjust accordingly.

      And it does: missed ACKs caused by upstream congestion cause the server to reset the TCP window size and transfer less data, slowing things down (just like it's supposed to do - of course this causes more ACKs, so...) The net affect to the user is that the download speed is greatly reduced. The actual bandwidth is not affected, it just looks like it is.

    2. Re:I'm afraid.. by Anonymous Coward · · Score: 0

      Know what the 'A' in ASDL stand for? :p

  28. use linux. by Anonymous Coward · · Score: 0

    Make a MASQ and pass your eXPerimental trafic
    trought it. then setup you shaping on linux.
    simple, efficien and improve security.

  29. Just go to your local shaper by taya0001 · · Score: 0, Offtopic

    I had surfboard made by this guy. He will shape it anyway you ask. He could prob fix your problem

  30. xp doesn't allow you to format when it's running by silicon1 · · Score: 0, Troll

    I think it's obvious that this cannot be done due to the sole fact that Windows XP doesn't allow you to format the drive that it is running on while it is running.

  31. Comment removed by account_deleted · · Score: 1

    Comment removed based on user account deletion

  32. Goddamned idiot ZELOTS!!!! by marcushnk · · Score: 1, Funny

    Can you people not read?? That guy has stated he understands that Linux can do it, BUT HE WANTS A WINDOWS SOLOUTION!!!!

    Get that through your dumbarsed bloody zelot minds.
    I love Linux, I love what it stands for and I love what it can do for me.
    But I have specific needs that linux cannot (easily) do. Also I'm better at troubleshooting win2k than I am at ANY linux disto.
    This guy has a requirement for a NON LINUX soloution. DONT FUCKIN TELL HIM TO FORMAT C:!!!

    grrrr /rant over
    And BTW, Although I don't have a windows soloution for you, what about the packeteer Packetshapers? They're and external unit, and VERY easy to configure, also if the power goes out they may stop shaping but will continue to feed.

    --
    "Consider how lucky you are that life has been good to you so far. Alternatively, if life hasn't been good to you so far
    1. Re:Goddamned idiot ZELOTS!!!! by thechuckbenz · · Score: 1
      The original poster specifically mentioned "Could such a traffic shaper be built using low powered computers? If so, how would you build and configure it so it would maintain compatibility for the single Windows machine, behind it?"

      So (most of) the recommendations of a linux box are for the shaper itself.

    2. Re:Goddamned idiot ZELOTS!!!! by marcushnk · · Score: 1

      Listen carefully you anon coward.
      Find your self a speeding bus, and perform an experiment in kinetics with it.

      Point 1 : Nope its not, but its not Linux based either, and he never said he didn't want an external unit. He also never said that it was because he found Linux hard to use, but incase that was the reason I offered an easy solution.

      Point 2 : you didn't shoot anything but your own foot, fucknuckle. besides I'd be surprised if you could hit the broadside of a bard with your intellect.

      Point 3 : Since when was this a Linux/BSD enthusiast site. This is a site for people who are interested in the less mundane more tech side of life. It just so happens that a great deal of those people are Linux/BSD enthusiast as well.
      Obviously its a little hard for you to realise this but its an intelligence thing. Go back to surfing stileproject you dweeb.

      Point 4 : *sigh* I give up.
      I remember now the old saying... never argue with a Moron, as they'll beat you down to their level and beat with experience.. /me out

      --
      "Consider how lucky you are that life has been good to you so far. Alternatively, if life hasn't been good to you so far
    3. Re:Goddamned idiot ZELOTS!!!! by Anonymous Coward · · Score: 0

      Why can't you people realize that a single OS should not, and in many cases can not, do everything. Linux/UNIX shouldn't be used for everything and Windows shouldn't either. There is the old saying "use the right tool for the job.", Windows just isn't the right tool for this job.

      The sad part is that Windows already has QoS built in, but it doesn't work exactly how people want it to work.

    4. Re:Goddamned idiot ZELOTS!!!! by Anonymous Coward · · Score: 0

      just use FreeBSD, problem solved.

    5. Re:Goddamned idiot ZELOTS!!!! by Alric · · Score: 0, Flamebait

      why don't you a) calm down & reread his submission and b)learn to spell a word if you're going to be using it so much?

      He has no "non linux" requirement. He just doesn't want to run linux on his main machine. As somebody else pointed out, he even asks if a cheap firewall machine could work.

      And the word is spelled ZEALOT.

      Here's a quarter...go buy an attitude that doesn't so closely resemble a jackass.

    6. Re:Goddamned idiot ZELOTS!!!! by Technician · · Score: 2

      Um, what I'm getting from this thread is, what he wants doesn't exist. Advice I always give anybody asking my opinion on what to buy is; define what you need to do and buy the hardware/software to do it. For some people that means a Windows machine for games, others it means Linux for a robust web server, others it means a MAC for desktop publishing (don't flame, Windows DTP has come a long way). I never recommend a brand, hard drive size, memory size, platform, etc. As you noted there is lots of platform specific solutions being recommended. Weed those out and see if there is any workable Windows solutions. If not, he may need to find a platform that will run a suitable application for his needs. In this department there seems to be lots of endorsements of a reliable solution that works.

      --
      The truth shall set you free!
    7. Re:Goddamned idiot ZELOTS!!!! by marcushnk · · Score: 0, Offtopic

      *sigh*
      I guess I'm a sucker for punishment.
      Yes I guess my grammar and spelling is a little poor at times, I tend to suffer from "Mind moves faster than fingers" Syndrome.
      Your grammar is no better, but please note I'm Australian so our rules on Grammar MAY differ to yours.
      Yes I'm (semi) old Gen /.'er, but this site is NOT a Linux enthusiast site. Its a site for (generally) tech news of different relevance, and the people who are interested in that sort of information, tend to be Linux enthusiasts..
      Convince me otherwise and I'll recant.

      I was pissed because he asked for WINDOWS based and not specifically LINUX based.
      I thought of throwing in the packeteer because he never said anything about a plug and play external solution.

      And for the record.. I had an idea what ^H^H^H^H meant, I just wanted to be sure, I asked a question and got an answer.. Why should I be ashamed for that?

      --
      "Consider how lucky you are that life has been good to you so far. Alternatively, if life hasn't been good to you so far
    8. Re:Goddamned idiot ZELOTS!!!! by essdodson · · Score: 1

      I'm really glad someone said it. This is a complete disgrace. You guys are fools. If someone asks how to peel and orange you don't give them an apple corer. Blah blah, I don't want to hear about how great Linux is, noone asked "How do I setup Linux packet shaping", the question is explicitly targetted at a Windows solution. Get over your Linux world domination kick and realize that some people just don't want Linux.

      --
      scott
    9. Re:Goddamned idiot ZELOTS!!!! by Anonymous Coward · · Score: 0

      Bastard. I didn't want a calm response to my flame.

      Ah well, what can ya do?

      So you're an Aussie, eh? Yeah.. Your grammer would be a bit different, plus every Aussie I've ever met say things way to fast (like Saturday.. when the Aussie boys I know say it, it sounds almost like sunday unless you listen closely enough. They screw me up all the time with that one.).

      And yes, my grammer and spelling stink too, and I'm quite subject to the typos.

      But I've got to say, so far in this thread it's shown that the most cost effective workable solution would be a linux or freebsd box between the windows computer and the modem. So far I have seen nothing mentioned for windows that can do what he asked. Most people are just suggesting the best tools for the job at hand. It would have helped if people would have directed him to specific HOW-TOs and documentation, but still, I think people suggesting linux, bsd, and the product you suggested were pushing the original asker in the right direction. Other people pushing QoS with Win2K's packet schedualer are the ones steering him wrong as it does not function in this manner.

      Oh well, good day.

    10. Re:Goddamned idiot ZELOTS!!!! by festers · · Score: 1

      Um, there is no Windows tool for this, you stupid tool. Sometimes Windows is the right thing for the job. Sometimes it's Linux. In this case, it's Linux. You can shut your face now.

      --


      -------
      "Every artist is a cannibal, every poet is a thief."
    11. Re:Goddamned idiot ZELOTS!!!! by Mattsson · · Score: 1

      The packeteers are also *very* expansive for a private person to use on his dsl-line. :-)
      If I'm not mistaken, the least powerful version scales up to 45MBit/s and costs about $20000. :-/

      --
      /.Mattsson - My native language is not English, so please don't whine over linguistic errors. (That's lame anyway...)
    12. Re:Goddamned idiot ZELOTS!!!! by MiTEG · · Score: 2
      I think it's more like if someone asks you

      What are the best oranges to grow in California?

      and you respond with

      You should move to Florida, you can grow much better navel oranges there.

      --
      The future isn't what it used to be.
    13. Re:Goddamned idiot ZELOTS!!!! by AndyMcL · · Score: 1


      Packetsh*tter is pretty good, but Sitara is better. I have done comparison tests and Sitara wins (www.sitaranetworks.com). It is easier to set up, does TCP rate shapping in addition to cbq, you can do Layer 3-7 QoS, and has better session fairness.

      They are both WAY out of a home users price range.
      I think that the cheapest one's are around $3-5k.

      That is just my 2 cents.

      AndyMcL

    14. Re:Goddamned idiot ZELOTS!!!! by subrider · · Score: 1

      I guess he shouldnt have posted his question in a place for Linux dedicated people then. He should have gone to Microcrap for an answer

  33. Just go to your local shaper by taya0001 · · Score: 0, Offtopic

    I had surfboard made by this guy. He will shape it anyway you ask. He could prob fix your problem



  34. OpenBSD not gonna work by krokodil · · Score: 2

    Having same problem I've attempted to use
    ALTQ with openBSD. It did not work since
    they do not work well on slow speeds (below
    8Kb/sec).

  35. I was wondering wtf happened to my Kazaa by Anonymous Coward · · Score: 0

    Out of nowhere, I was unable to get decent downloads. I've been dishing out about 3 GB/4 days...throttled down the outgoing bandwidth and I've got excellent speeds once again. Thanks for the (indirect) tip.

  36. QOS on Win2K/XP by Anonymous Coward · · Score: 0

    The QOS support on Win2000/XP is an API for applications to request certain amounts of bandwidth from other network devices. There's no mechanism where you can control the queue of outgoing packets yourself.

    The idea is that your VideoPhone application will tell your ISP that the data's important and should be given guaranteed bandwidth.

    It won't let you shape the traffic yourself :(

  37. Traffic Shaping Appliance by MonMotha · · Score: 2, Informative

    Linux has rather impressive QoS facilities (so many options that it has an entire submenu in menuconfig). What you'll want is some floppy router distribution (mine's not quite ready yet, but I hear LRP has most of these options) and a decently powerful machine (100-200MHz pentium with 32-64MB should PLENTY for a DSL, but remember, you're not just routing/NATting).

    There are tons of preconfigured things out there, but you might want to read up on tc (the traffic control manipulator, part of the iproute2 distribution), ip (from iproute2) and iptables (to help classify packets) before you dive in. The kernel ships with most of what you'll need (including the common CBQ scheduler), but there is a really cool scheduler known as HTB that is more accurate because it's resolution is traffic based, not time based. If you want to shape inbound traffic destined to teh router itself, you'll also need the IMQ patch.

    Hope this helps. If you want more info, EFNet #iptables, look for KurD, the human router. He plays with this stuff all day at his job.

    --MonMotha

  38. I'm designing DSL equipment for this by AaronW · · Score: 2

    When I was designing support for discarding packets in a BRAS (Broadband Remote Access Server), I put in support to prioritize control packets (i.e. tcp ACK over data packets to try and make this a lot better. A BRAS is a box that sits at the other end of a DSL modem and terminates PPP and PPPoE sessions (and more). This would only affect traffic going to the end user since the upstream traffic is usually shaped in the modem. I put this in specifically to deal with the problem you describe. I ran into the same problem when @Home reduced the upstream bandwidth to 128Kbps from 1Mbps. I did some shaping on the software running on my computer which helped a lot, but most users don't have that flexibility.

    The box I'm working on supports tens of thousands of simultaneous DSL users and can shape and buffer each user's traffic independently, going so far as being able to shape individual traffic flows to and from a user. It's also designed so that a user can change the amount of bandwidth they want on the fly and letting the ISP choose how to charge for this. This also allows for things like downloading video on demand, where the pipe from a video server to the subscriber can guarantee bandwidth for that flow.

    As for shaping traffic in Windows? I havn't a clue. I don't do Windows.

    --
    This post is encrypted twice with ROT-13. Documenting or attempting to crack this encryption is illegal.
    1. Re:I'm designing DSL equipment for this by DrPascal · · Score: 1

      This post provided absolutely no merit or value to the discussion. Did I miss your point, or were you just patting yourself on the back?

      "I did this really cool thing. Sucks for you!"

      --
      DrPascal: Not the language, the mathematician.
    2. Re:I'm designing DSL equipment for this by Anonymous Coward · · Score: 0

      I appreciated it, anyway.

  39. Hardware solutions by SiO2 · · Score: 1

    I understand that you're looking for a software solution, but just for a bit of extra information consider looking at hardware solutions for packet shaping. They tend to be cost prohibitive for the home user, but you might want to check out the solutions from Packeteer: http://www.packeteer.com/.

    At the small, private university where I'm the network admin, we use one of these boxes on VLAN2, which is the dorm network. We use it largely to choke outbound file sharing traffic and keep track of who is being naughty and who is being nice.

    I know it's not what you're looking for, but I thought the Packeteer packet shaping boxes might be of interest to the /. community.

    Later.

    SiO2

    1. Re:Hardware solutions by DrVxD · · Score: 2

      Might be nice as a "off-the-shelf" solution (if you have a rack), but - as you say - probably cost-prohibitive. A Linux box would be a much cheaper option for a "homebuilt traffic-shapping appliance".

      --
      Not everything that can be measured matters; Not everything that matters can be measured.
  40. Re:fp? by Anonymous Coward · · Score: 0

    Yeah, hi. I use Windows and I think you Linux guys are really neat. You have such kewl programz. Do you think that one of you could tell me where I can find some kewl programz that might work with windows stuff? I want some thing that looks like that gr33n window in XP. That's s00p3r kewl.

  41. its meant to be... by skydude_20 · · Score: 1

    . If anyone heard of a peace of software that can do this,

    Hey man, Peace Love & Linux, you subconsiously want to be doing this with Linux

    --
    Jesus saves souls and redeems them for valuable cash prizes
  42. Nope. by mindstrm · · Score: 4, Informative

    Unless I'm mistaken, that won't hlep.

    What the QoS scheduler does in windows is permit applications who specifically request a certain qos to keep it. Í don't think it's a robus, configurable queuing system.

  43. WTF is this? by Com2Kid · · Score: 0, Offtopic

    Ok so I know this is horribly off topic and I expect to be moderated as such, but, err, uh

    WTF IS THIS?

    (replies to com2kid@attbi.com please. Nobody 'round me knows WTF it is and that is the third one I have found flying around my house!!)

    1. Re:WTF is this? by Com2Kid · · Score: 0, Offtopic

      or is that a wasp?


      Notice, connector to stinger is thin, like one little strand. Ain't like any wasp I have ever seen, and I have asked others....

      I saw a transparent shell like structure around one stuck up against the wall, thoguht it was dead, next day it was flying around. Creeps me out. You don't have a sense of scale in that picture, the darn things are /HUGE*/.

      *Huge being defined by a person who lives in the Western part of Washington State and ain't used to anything bigger then his pinky fingernail wandering around. :-D

      That is actualy a slightly smaller one in that picture though, the other ones where bigger . . . .

    2. Re:WTF is this? by Drachemorder · · Score: 0, Offtopic

      It's a dirt dauber. We have tons of 'em in Mississippi. They're harmless. I didn't know there were any in Washington. Maybe I accidentally took a few of them with me when I visited the state a few years ago :-D

    3. Re:WTF is this? by JayLawrence · · Score: 0, Offtopic

      Looks like a paper wasp. This
      link has some wasp identification. Whatever happened to google searches?

    4. Re:WTF is this? by alyandon · · Score: 0, Offtopic

      It's a dirt dauber. We have tons of 'em in Mississippi. They're harmless. I didn't know there were any in Washington. Maybe I accidentally took a few of them with me when I visited the state a few years ago :-D

      I dare you to poke a 12" stick into a fully occupied nest and see how "harmless" they react... :)

    5. Re:WTF is this? by mrbuttle · · Score: 0, Offtopic

      That appears to be a mud dauber, or mud wasp.

    6. Re:WTF is this? by Com2Kid · · Score: 0, Offtopic

      This may give a better idea what I say when I mean it has no waist.

      Shiiiit my Karma's gonna burn. :(

    7. Re:WTF is this? by Com2Kid · · Score: 0, Offtopic

      Spent around 5 hours on google, ain't seen nothing like this yet.

      I just got done going through a listing of paper wasps, does not appear to be so.

    8. Re:WTF is this? by Com2Kid · · Score: 0, Offtopic

      I am starting to REALLY hate these mild ass winters we have been having lately, shit is starting to move in the area that ain't supposed to be here, FUCK.

      I am used to having NOTHING around here that is the least bit harmless OR scary looking. Well except for under the house in the crawl space where some pretty freaky shit lives that ain't ever seen the light of day, but nobody /ever/ goes down there so that isn't much of a problem. :-D

      (two people in house hold not myself, alergic to stings, one deathly so, thus why I am interested in finding out WTF these things are)

      The lookups I did on google don't seem to show a waist as thin and narrow as this one has though. . . aaah! Also no mention of the weird ass tube thing around in, ::shivers::

    9. Re:WTF is this? by Com2Kid · · Score: 0, Offtopic

      Know of any state wide removal services? Actualy down through oregon would be nice, oregon is the traditional "barrier against creepy crawly things" for washington state. . . .

      (I mean WTF are these things doing in Washington state? They weren't here last year, and fuck, I have asked people who are over 70 years old living in Washington and they haven't ever heard of the things either, WTF are they doing /here/?)

  44. Bandwidth Limiting HOWTO by dimator · · Score: 5, Informative

    Bandwidth Limiting HOWTO might be of some assitance? Or not... either way. It just caught my eye on linuxdoc.

    --
    python -c "x='python -c %sx=%s; print x%%(chr(34),repr(x),chr(34))%s'; print x%(chr(34),repr(x),chr(34))"
  45. Shocking by Anonymous Coward · · Score: 0

    A whole page full of "run UNIX" comments. This is one of the times that your blind zealotry is not needed.

  46. Das religi�se Wesen by ellisDtrails · · Score: 1

    The guy asks a legimate questions -- a more technical and interesting from a DYI perspective than I've seen on Slashdot in a awhile, and everyone creams him. A herd mentality. Instead of steering him to a sensible solution, which is mostly on a Linux platform anyway, he is cast away. Left to drown in the cruel world of google searches and pop-up ads. -- ellis d. trails "Madness is rare in individuals-- but in groups, parties, nations, and ages it is the rule." -- Nietzsche

  47. Comment removed by account_deleted · · Score: 0, Offtopic

    Comment removed based on user account deletion

  48. Re:Go cable, baby by hummer357 · · Score: 1

    ehm... well... not actually...

    over here in Belgium adsl is 1024kb down 128kb up, and cable is 10Mb (yes -- TEN megabit!) downstream, and 128k up.
    my personal download speed record is 4 streams for a total of over 1100Kbytes/sec -- on cable of course ;-)

    and that all for a good 42 (or dollars) per month...

  49. Interesting by Anonymous Coward · · Score: 0

    Interesting. I'm moving to the Netherlands soon- I wonder if the situation there is similar?

  50. Um... grammar 101 by Anonymous Coward · · Score: 0

    If anyone heard of a peace of software that can do this, I would love to hear about it. Wow... stick with Windows. If you can't get past basic grammar, don't try linux.

  51. �:+0 by Anonymous Coward · · Score: 0

    Probably around $50 to 42 now. Can't wait 'til Bush gets the Boot.

  52. Lots of people complain about this.... by billstewart · · Score: 5, Informative
    There are a couple of problems that can lead to this symptom, and asymmetric connections are more likely to suffer from it. Basically, if your upstream bandwidth is swamped with file transfers, it's hard for the ACKs for packets you receive on your downstream to get through, so the downstream TCP sessions get throttled. Many DSL providers oversubscribe their links (though the problem is usually worse downstream than upstream)(cable can have this problem also), so that 128kbps upstream you thought you had might only be 38kbps if lots of people are uploading at the same time - so your upstream packets may be queued at the DSLAM, and your little ACK packets have to wait for a windowful of big file transfer packets to clear out, or they may be queued in _your_ DSL router. An 8KB TCP window is 80k bits, so even if you're getting your whole 128kbps upstream, that's over half a second of data that could be sitting ahead of your ACK, dogging your downstream flow rate. It's especially bad if your upstream traffic has a big window size, while your downstream is using slowstart and has never gotten a very big window.

    So yes, traffic shaping can be your friend. Unfortunately, it may be hard to know what to tune it too, depending on where the bottleneck is.

    --

    Bill Stewart
    New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
    1. Re:Lots of people complain about this.... by pediddle · · Score: 1

      I've never had this problem with my cable modem, but I used to have it absolutely horribly back why I had a 56k modem. I couldn't use any type of filesharing program and surf the web at the same time, and I whenever I would download a file my ping times would go up to about 5000ms. This seemed to be even worse in Linux than in Windows. It seems like PPPD (the Linux dialup program) should automatically activate some kind of traffic shaping, giving ACKs higher priority or something (I don't know that much about how it all works). Modem access with Linux is just about unusable when trying to do two things at once, the way it stands today.

    2. Re:Lots of people complain about this.... by roc_machine · · Score: 1

      An 8KB TCP window is 80k bits, so even if you're getting your whole 128kbps upstream, that's over half a second of data that could be sitting ahead of your ACK, dogging your downstream flow rate. It's especially bad if your upstream traffic has a big window size, while your downstream is using slowstart and has never gotten a very big window.
      If I'm not mistaken, this can be correct by adjusting the Recieve Window on a Windows machine, which defines the size of the recieving computer's packet buffer. You can find out how to adjust that here.

      An 8K buffer is too tiny for a highspeed connection. If a packet is lost or corrupted, the reciever has to notify the sender of this and in the mean time, buffering the incoming packets until the replacement packet arrives. Of course, I don't have my 4th year comp sci network book in front of me, so I could be off a bit (or a lot) here.

      An extremely large recieve window (buffer) is not great for performance either. Does anyone know why that is the case? I think the same principle applies but I'm not sure.
    3. Re:Lots of people complain about this.... by WhiteKnight07 · · Score: 4, Informative

      Give Cable Nut a try. It lets you play with all that stuff and well, a butload more stuff having to do with how the TCP/IP stack behaves. It's for windows only but it maybe just what this guy needs. Although its not true traffic shaping it might help solve the problem.

      --


      We're going to make information free Mr. Anderson, whether you like it, or not.
    4. Re:Lots of people complain about this.... by Antity · · Score: 1

      There are a couple of problems that can lead to this symptom, and asymmetric connections are more likely to suffer from it. Basically, if your upstream bandwidth is swamped with file transfers, it's hard for the ACKs for packets you receive on your downstream to get through, so the downstream TCP sessions get throttled.

      This is a design issue with TCP, since TCP was simply not designed for asymmetric connections. So the protocol assumes that if upstream capacity is exhausted, then downstream is, too.

      --
      42. Easy. What is 32 + 8 + 2?
    5. Re:Lots of people complain about this.... by Anonymous Coward · · Score: 0

      That's because it's a MODEM!

      One download will max out your connection. So of course things get congested when you try to multitask.

      That's the nature of narrowband.

    6. Re:Lots of people complain about this.... by LordNightwalker · · Score: 1

      TCP was not especially designed for symmetric connections either. Besides, what makes you think that upstream and downstream exhaustion have anything to do with each other?

      Even in a symmetric environment it's perfectly possible to clog up your upstream BW and still have lots of downstream available (think webservers hosting large files).

      The protocol doesn't assume anything; the protocol sends packets and waits for acknowledgement. That's all it does. It doesn't know nor care about symmetry in down or upstream capacity.

      --
      Install windows on my workstation? You crazy? Got any idea how much I paid for the damn thing?
  53. Common Problem by EoRaptor · · Score: 1


    This actually done by the service provider deliberately. Happens a lot with cable providers as well. This is a more subtle attempt to punish those who actually use their upstream, though the very low upstreams we are seeing all over the place is another facet of this.

    I know you think I'm trolling or wrong, but try it yourself, especially if you are an ex-@home user, who had they entire network set up like this. Uploads will kill downloads, every time. The exact method varies from equipement to equipement, but the most common method is to have the modem itself forward ACK's at a lower priority than other packets. When the upstream is idle, tyhis is fine, but upload something, and your downstream goes to heck.

    For the conspiracy lovers out there, this goes hand in hand with the aforementioned disparity in upstream and downstream rates. The *only* reason this is setup like this is to prevent you, the end user, from providing content that might compete with the ISP's partners. There is no technical reason why the upstream has to be slower, it's purely a business decision on the part of the ISP.

    1. Re:Common Problem by Anonymous Coward · · Score: 0

      What a load of BS. This isn't some conspiracy set upon users by ISPs, it's built into TCP/IP, and the larger the asymetry in the bandwidth the more pronounced the effect.

      If your outbound ACKs have to compete with your other outbound traffic on a saturated oubound link, ALL your incoming traffic has to wait for those ACKs. That's why the incoming slows down. It makes no difference whatsoever how large your incoming pipe is, if you can not send ACKs back out, your download will crawl.

      All you have to do is prioritize the outbound ACKs ahead of other traffic and the problem is almost completely eliminated.

      You do have to surrender a tiny bit of your bandwidth to get the best result, but the difference the Wonder Shaper type things can make is startling.

    2. Re:Common Problem by dossen · · Score: 1

      There is no technical reason why the upstream has to be slower, it's purely a business decision on the part of the ISP.

      While it does not force upstreams to be as small as reported, the frequency division schemes used for xDSL means that high downstream bandwidth results in low speed upstream. There simply isn't enough bandwidth on the copper-pair for several megabits in both direction. The question then becomes, upstream or downstream?

      And that aside, what's wrong with offering a huge pipe downstream? Anybody with half a brain and a little knowledge of tcp/ip knows that this will consume part of your upstream for ACK's. Why blame the isp for not providing a huge upstream, instead of enjoying the huge downstream?
      I don't know the actual offers in the states, but around here (Denmark) the isps offer several different downstream/upstream packages, so one can just choose the right one. And the price naturally reflects ones choice. If you need the extra upstream you simply have to pay for it.

  54. When did Ask Slashdot turn into Tech Support? by gsmit07 · · Score: 1

    As for the answer to the question, I suggest exploring the QoS settings in W2K server.

    Commentary: There is a fine line between being a support center for Joe-Bob and a discussion driven community forum. The average /.'er is savy enough to search deja/google for these types of questions, and at least exhaust those avenues of research before posting an AS question. I get several pertinant results with an un-refined deja search - with some honing I bet I could find the answer to this question in 5 minutes.

    AS need not be a place for know-it-all computer types to rag on newbies. Well thought out questions are not the monopoly of us (self-claimed) know-it-alls.

    1. Re:When did Ask Slashdot turn into Tech Support? by deek · · Score: 2, Insightful

      There is a fine line between being a support center for Joe-Bob and a discussion driven community forum.

      You're correct, there IS a fine line between support and discussion. With support, the same issues appear again and again. With discussion, an issue is talked about once, and never appears again unless there's new information.

      Personally, I think that Traffic Shaping is a fine topic for discussion. If it also answers someones technical question, then all the better! But if this topic appears again in the next few months, though, then that has stepped over the line.

    2. Re:When did Ask Slashdot turn into Tech Support? by fbain · · Score: 1

      you're and asshole, gsmit07. Just go back to where you came from and be quiet.

    3. Re:When did Ask Slashdot turn into Tech Support? by gsmit07 · · Score: 1

      Thanks for proving my point. When did the comments section turn into a place for children to use the profanity they've learned in the schoolyard?

  55. Blame the HARDWARE by Webexcess · · Score: 1

    Some DSL modems *cough*alcatel*cough* have been designed to buffer data (1MB or more) when it reaches the maximum allotted bandwidth instead of simply dropping it on the floor like a good device should. Buffering the data seriously confuses the TCP flow control of your OS because it sees huge bandwidth going through one second, followed by no bandwidth the next, and so on.. These modems also throttle the downstream when the upstream is maxed, which is what you're probably seeing.

    If you traffic shape your upstream to just under the cap you shouldn't see this sort of stupidity (or you could update the firmware yourself, though your ISP probably wouldn't like that)

  56. The proper way to do this by 0x0d0a · · Score: 2

    Okay, here's the issue. I've set this up myself, and there are good HOWTOs on it, and I've done a bit of testing. First, the problem is not prioritization. The packets are leaving the XP box in perfect order. The problem is that the cable modem has a ridiculously large upload queue, which fills up. Packets then take a hell of a long time to get through. You just want to cap the outgoing data rate *at your computer* to just below the cable modem limit, which keeps the buffer from filling up. You can play with prioritizing various packets if you want, but it doesn't do much of anything for me. The data rate cap massively improves download rate.

    Honestly, I'd just get an old box and set up a nice Linux router/mail server/whatnot, which will give you more flexibility and if you decide to add more machines to the network, not require you to have your workstation up 24/7.

  57. Linksys by ruvreve · · Score: 2

    This could be an interesting feature to add to their already EXTREMELY USEFUL "Cable/Dsl Routers"

    Linksys managers listening??? Get your engineers working on it.

    Linksys recruiters listening??? I have several other ideas that would be nice to incorporate and since I'm a college student I'd rather not give them away for free :)

    1. Re:Linksys by Anonymous Coward · · Score: 0

      Uhh, Linksys already makes a router with Qos features.

    2. Re:Linksys by Zalgon+26+McGee · · Score: 2

      Linksys has a broken PPPoE client in their routers - they do not ensure that all packets are greater than the Ethernet minimum size. See http://www.istop.com/linksyssucks.html for the details.

      Personally, I'm running an SMC Barricade 7004ABR, and love it.

      --

      ---

      Book(n): Utensil used to pass time while waiting for the TV repairman

    3. Re:Linksys by Anonymous Coward · · Score: 0

      One really useful feature is to monitor the up/down cap on the router.

      My ISP find a bug in Ethernet frame fragmentation not meeting minimum frame size.

  58. BBIAgent by t0qer · · Score: 5, Informative

    I've been running BBIagent
    http://www.bbiagent.net
    for about the last year or so. I share my DSL with a few neighbors via some cat5 we strung up over the last year, and my counterstrikin was startin to hurt real bad whenever one of the teenager kids was downloading from kazza

    Before I was using a linksys router. My current BBIagent setup is a packard bell p60 with 32 megs of ram, and 2 old 16 bit isa 3com nics. The performance increase is stunning.

    You can specify what kind of traffic to create priority for, forward, block ports. There are a few options for hack detection blocking. The nicest features of bbiagent is.

    1. Single floppy distro which is configured on the fly via the bbiagent website
    2. Java app for admining the sucker, very well laid out.
    3. Realtime control over your network connection without losing it. On my linkstink whenever I would change a rule or a forward the whole thing would reboot itself, bbiagent doesnt do that.
    4. Open source!
    5. Linux Based!

    I can go on and on about it, but if you're really looking to control what and how packets are handled, I would really recomend giving bbiagent a try.

    --toq

    1. Re:BBIAgent by Marasmus · · Score: 2

      Whoa, that is COOL! Just checked out and played with BBIagent.... and that is the freakin' best little web-managed firewall system I've seen to date. Wow.

      Though I won't use it myself (i've become pretty skilled with the BSD pf), I know a ton of people who will absolutely love this. Absolutely awesome. Thank you for posting this!

      --
      .... um, i lost you after "0110100001101001".
    2. Re:BBIAgent by Anonymous Coward · · Score: 0
      Yeah, that's just great, until the power goes out or you shut the router box off to save a little energy. BBIagent, in its current form, saves *precious little* of its configuration data between reboots. It'll remember how to connect alright, but all your port forwarding, packet shaping, and security/access control lists go poof on a reboot.

      I mailed the developers and they said the next version will probably have more support for saving everything between boots. I might switch if that happens. I currently use FREESCO, a great little floppy- or hard drive-based router distribution.

      No, it doesn't do traffic shaping or have really fancy packet dropping support (uses 2.0.38 kernel), but it does have a rockin' HTML/CGI admin interface, comes with DHCP, DNS, HTTP, LP, telnet, time, dial-in, and a few other servers I'm forgetting, all on a floppy. It kind of bugs me that you can't access the filesystem in BBIagent, but then you don't really have a need if it's not expandable. FREESCO also has a self-downloading, self-installing package system for expanding it with everything from Samba (*very* useful if you want a little NAS with your routing) to Apache to nmap.

      That said, I really, really, really wish I could do traffic shaping with it. =) Maybe they'll upgrade the kernel in the next version.

      Lime

    3. Re:BBIAgent by t0qer · · Score: 2

      Your freesco looks great, but where's the built in PPPOE support? I RTFM'd and didnt find a mention of it once!

  59. You aren't going to be able to fix this locally by tlambert · · Score: 5, Informative

    The reason your downstream rate gets limited by your upstream rate limit is because of your inability to send acknowledgements upstream for the downstream packets you are receiving because of the congestion on the upstream link.

    Once the advertised receive window is filled with packets, the remote side is not going to send any more data until it gets an ack from your computer.

    At the point that the rate is being limited, the way to fix this is to give preference to ACK packets -- i.e. packets without payload.

    The problem is that this has to be done at the point the rate is being limited, which is at your providers router, not at your router, and not at your Windows machine.

    Installing the Microsoft packet scheduler, or some UNIX box with a real netowrking stack that you can exercise fine control over packet priority, as some people have suggested, isn't really going to fix your problem, except by maybe 10% (the exact amount depends on your average outbound payload size relative to an "empty" ACK-only packet).

    This is because the problem is the queued data in the transmit buffer in the rate limiting device not containing your ACK's for inbound data.

    Really, the TCP/IP protocol wasn't built for asymmetric reates, without equally asymmetric data transfers.

    Effectively, you need to be able to control the transmission of data packets from your end based on knowledge of how full the input buffor on the outbound leg for the rate limiting device gets, so that you can throttle your data payload packets accordingly, to keep that buffer as empty as possible.

    The only way you can really do this is to put code in your stack to monitor the advertised window from the other side, and the locally classify outbound packets as to whether or not they contain data payload, or are merely acks. You basically have to avoid filling the outbound queue on the rate limiter above 50%.

    In an ideal world, the machine doing the rate limiting would do this for you. Some rate limiting machines for asynchronous connection do this, and it's not a problem (you can see the posts from the people who are rate limited, and don't understand your problem). But those machines are more expensive, and it's just as well for the provider if you feel pain as feedback for uploading, since it serves their purpose in providing you asymmetric service in the first place.

    The problem's a lot easier when you are trying to avoid filling the inbound receive buffer for a router with a speed differential on one side (e.g. the inbound receive buffer on a router connected to a dialup modem bank, with a customer on a modem wanting to do QoS based on protocol type, so their SSH didn't lock up when an FTP or large HTTP transfer started up). *That's* where things like "AltQ" and the Microsoft packet scheduling engine become useful... not here.

    -- Terry

    1. Re:You aren't going to be able to fix this locally by Malor · · Score: 5, Insightful

      This really isn't true. This can be fixed locally.

      Consider: his problem is that his upstream provider is deciding what packets to throw away, and is making bad decisions.

      If he installs a local rate limiter, such that the upstream bandwidth is never oversaturated, the provider will no longer be making decisions about packet drops... he will be making them locally, and can prioritize them however he likes. Since he decides what gets dropped, he can pass the ACK packets first.

      That said... doing this with Windows isn't easy. I'm not that familiar with it, but Routing and Remote Access Services *might* have something along this line. I'd suggest perusing the Microsoft web site to see if there's anything like that. (or someone else may post this info here on /.)

      At work, we're doing bandwidth management on a 40 megabit connection, using software from www.etinc.com. It runs on Linux or FreeBSD. I'm mentioning it here not so much for this question, but as a general-purpose recommendation for people trying to do bandwidth management on large connections. This software costs about $700 and is closed-source, but it works really well. The earlier version we have totally choked and died, but the 3.21 version does exactly what it says it will.

      I'm still not sure, offhand, how you'd use that software to solve this specific problem. You can prioritize and limit based on ports and source/destination, but I'm not sure offhand that you could use bwmgr to optimize based on packet payloads.

      Basically, what it REALLY sounds like is that the connection is TOO asymmetric and that he might consider an alternate provider, if available.

    2. Re:You aren't going to be able to fix this locally by hklingon · · Score: 1


      If you could prioritize ACKs then, by virtue of the protocol, Uploads (and their downstream acks) would be the first to be throttled. Also, TCP, by design, has the underlying assumption point-to-poing bandwidth is symmetric. It is not in this instance.

      Anyway, How can you use the QoS thing to give Acks a higher priority? You can install QoS in control panel on Win2k/WinXP, but you can't actually install the QoS management snap-in except on W2k Server. I *doubt* they'd throw in ack management into this QoS, but they may.

      Wendell

    3. Re:You aren't going to be able to fix this locally by olman · · Score: 2

      I have a traffic shaper running in my linux router box right here. Works like a charm. The important thing is that you can have two or more priority classes. Higher one for ACK packets and slower one for "bulk" traffic. So ACK packets still do not fill up your upstream buffer, but they'll always get preference over data packets. You can read all about it in the linux traffic shaping howto (http://lartc.org/).

      You might think this is a one-stop solution for running P2P-apps unfairly, i.e. with extra slow upstream quota. Unfortunately, edonkey etc keep on opening new upstream connections since they think there's still plenty of bandwith to spare.. 20-odd rate-limited upstream connections is bad for your karma :-)

    4. Re:You aren't going to be able to fix this locally by joe+user+jr · · Score: 5, Insightful
      Nope. It may well work to apply a fix at the local end. It all depends on what actual traffic is causing the problem

      The asymmetry of ADSL is a business not a technical "feature" - and it's based not on some wild conspiracy to stamp out freedom and file sharing (as in the response far above this one) but on the sound knowledge that most traffic just is asymmetric - typically you have 1500 bytes of webpage arriving downstream for every 40 byte acknowledgement you send upstream, at least in the use-model that ADSL was constructed on. So it would just be a plain old waste of money to provision for symmetric network use.

      Where I am, we have a network of about 80 users sharing a 2M down, 256k up, ADSL connection. This worked fine until some of our users discovered gnutella.

      Gnutella is a very bandwidth-hungry protocol, and tends to saturate the upstream bandwidth on ADSL. This resulted in dramatic loss of performance for our web users, for reasons explained above (namely that since acks get delayed, the nagel algorithm on the other end kicks in, reading this as general congestion, and slows the sending rate down dramatically.

      I fixed this by installing iptables and the tc ("traffic control") application on the linux box we use for a router. This works using "class based queueing" - you divide up your traffic into several classes, depending what their source and destination IP and ports are (or if they're related to other traffic, with particular ports and IPs). And then you give each class a bandwidth limit (hard or soft).

      The way we do this is to use the iptables (successor of ipchains) functionality to insert a "marker" into each relevant packet, and then have tc put them into the appropriate class based on that mark - this gives you all the selectivity (and clarity!) of picking packets that iptables offers.

      In our simple setup, reserving 64k or so of bandwidth purely for acks going back to web servers (ie going to port 80) and a few other types of traffic, and a bit of fine tuning, is enough to keep the connection very usable, and let people use gnutella on it as well (at a rate that's reduced a little.)

      In the face of constant gnutella traffic, this improved our web connectivity by about 900% rather than 10%. Since you only send a 40 byte ack for each 1500 bytes you receive, a ratio of about 37:1, reserving 64k for acks is enough to cover for 37*64 = (over 2M) downstream traffic.

      If you run, say, a local ftp server, you could isolate the traffic from that very easily by marking packets which originate from port 20 and 21 on its IP address (assuming the ftp server is well-behaved and sends its data-connection packets from port 20) and limiting them so that you save 30k or whatever upstream for your other traffic.

      None of this needs to be done at or beyond the provider's equipment. Because we limit the rate at which we send traffic to the provider, their equipment doesn't get its queue filled, ever! (unless they're not fulfilling their committed data rate, which we can't really control).

      So a local solution may be entirely possible - this will depend on just what traffic is clogging up the upstream.

      As for specific software recommendations, I don't know of anything that does this on windows, personally. I suspect it's likely to be payware, and will cost more than an old PC that you can run linux on. We're using a P133 with 32M and I have a feeling that it's slightly overspecified (at least on RAM - I think 16 or likely 8 M would do fine).

      --
      .sigs: Just Say No!
    5. Re:You aren't going to be able to fix this locally by Anonymous Coward · · Score: 0

      Here is a big advantage to iptables that I did not know existed. I have used ipchains quite a bit, but that was a couple years ago and I never bothered to read up on the features of iptables. Thanks for the info - I will have to build a box this weekend.

  60. Collisions? by LinuxWhore · · Score: 2, Interesting

    I dunno. When I hear stuff like this is think of what happens when you start running traffic through a full-duplex NIC on a half-duplex LAN. If a full-duplex machine on the network starts sending/recieving traffic from the net, it's gonna seriously affect your throughput. I always suggest that people check through all their machines for full-duplex OR 100Mbit when using a 10M hub or a low-end 10/100 switching hub.

    --

    I am MuchTall
  61. What do you say... by Cheetah86 · · Score: 0, Offtopic

    What do you say when your download speed gets crushed by uploading?
    ACK!

  62. The XP in Windows XP by prestomation · · Score: 1

    XP actually stands for eXtra Painful :)

    1. Re:The XP in Windows XP by Anonymous Coward · · Score: 0

      That or eXtra Profit. Just a beta they foisted out to get more cash.

  63. Traffic Shaping on DSL? by DeathFrog · · Score: 1

    You probably get a better deal by not buying a Linux box, downloading the Linux on Windows client at RedHat, finding the Bandwidth shaper for Linux, and Zang, Bob's your uncle. Good luck guy.

  64. Um, no by tech81 · · Score: 0

    I work for a very large, national US telco ISP (who shall remain nameless) and I can tell you that your upstream shouldn't affect your downsteam. Hell, they operate on very different frequencies. Upstream on 20Khz and downstream on about 1.1Mhz. If you are having that kind of trouble, I would contact your ISP or telco DSL line repair group for testing while you duplicate the issue.

    1. Re:Um, no by Anonymous Coward · · Score: 0

      mm well I see about 1000 ppl on irc every day and i've yet to talk to ANYONE on adsl that does NOT have this problem... ppl who use isp's and adsl providers from all over the globe.. seems to apply to cable as well..so if its no technical problem.. then its very very weird..

    2. Re:Um, no by tech81 · · Score: 0

      It could be an issue with MTU/RWIN then.

    3. Re:Um, no by Anonymous Coward · · Score: 0

      But it's not a DSL problem, it's a TCP/IP problem.

      When the ACKs cannot get out in a timely fasion, due to the uplink being saturated with other data, then the download has to slow down. ADSL is to blame because it is asymmetrical and TCP/IP was not designed to run on an asymmetrical link. All is not lost though. Some prioritizing on ACK packets can fix up the problem somewhat. Of course a greater upstream link would be better.

  65. Windows XP "box" by z7209 · · Score: 1

    Hehe, when did a Windows system get promoted to the status of "box"

    ;-)

  66. Doesnt this depend on xDSL type? ADSL .. SDSL.etc? by Anonymous Coward · · Score: 0

    I mean.. i thought the diff between adsl and sdsl was that sdsl lets u have the full use of watever bandwidth u have each way full and unaffected by the other stream.. being SYMMETRIC?.. Anyway it would NOT suprise me if it was the ISP doing this to kill sharing.. I got adsl and have tried stuff like Direct Connect but its really useless since theres no way of limiting OUTGOING speed in DC or at least reserve a couple of kb/sec so it ends up with other ppl dlling from you and yourself getting a few bytes/sec because the more connx outgoing u got the bigger % of the speed it uses.. if I just send something it will usually limit itself to 12kb/sec but if 3 people try and get a file from me it will go to 15-16kb/sec and i get under 1kb/sec downstream myself.. and they call Direct Connect.. TRUE SHARING? true sharing my ass.. so IS there a SOFTWARE solution to this? FOR WINDOWS...so dont gimme no linux shit, its not even remotely interesting until linux becomes a proper desktopOS for dumb users who doesnt wanna screw with commandline arguments and kernel recompilations.. I/O

  67. isa server? by Anonymous Coward · · Score: 0

    1 Windows Server Licence $1500AU
    1 MS-ISA Server $1000 AU
    x Hours of troubleshooting and web-lookups ($x)

    I think thats the quickest way of getting shapping working with a Micrsoft Solution.

    Sorry, Linux is cheaper and easier to use. It also doesnt have "Help/Is your software Legit" in Windows Explorer.

  68. Re:Go cable, baby by H3XA · · Score: 1

    In China I have the choice of 1984Kbps/384Kbps ADSL connection and 10Mbps or 100Mbps cable connections.

    ADSL and 10M cable is about $15USD unlimited per month and 100M cable is double that.

    :)

    - HeXa

  69. Just change the TCP maximum receive window by Animats · · Score: 5, Informative
    OK, let's look at what's going on here. The downlink is being underutilized when the uplink is saturated. That just looks to TCP like a link with a long round-trip time, which TCP can handle. But the Windows default configuration isn't set up for it.

    Try changing your local TCP buffering parameters instead to allow a bigger receive window. Set DefaultRcvWindow to something like 32768; the default is 8192, which is low for a DSL line, especially one that asymmetrical. With a bigger setting, you can have more data in flight, which makes the TCP connection more tolerant of delays in ACK return.

    Given the numbers you're reporting, your ACK delay isn't that severe. You're only losing 2/3 of your downstream bandwidth. So quadrupling the amount of data allowed in flight should overcome that problem. Prioritizing your uplink traffic should be unnecessary at this time.

    A real question to ask is "why are you trying to run a server on a 96Kb/s line". Buy hosting from somebody; it's cheap and they'll have far more bandwidth.

    1. Re:Just change the TCP maximum receive window by tlambert · · Score: 2

      The problem is outbound pool retention time.

      Making the receive window larger is going to delay the amount of time before an ACK is mandatory, not reduce the contention once an ACK becomes necessary.

      In other words, if I'm filling a bucket a 10 gallons a minute and emptying it at 1 gallong a minute, no matter how big you make the bucket, it's going to overflow eventually.

      In other words, your outbound ACKs are still going to have to stand in line behind your outbound data at the outbound rate limiting box.

      I will now point out that every outbound buffer between you and the rate limiter will also be full, and need to drain before your ACKs get through...

      -- Terry

    2. Re:Just change the TCP maximum receive window by Anonymous Coward · · Score: 0

      I think your analogy is flawed. By increasing the recieve window you are essentially reducing the amount of ACK's that must be sent per data recieved. True the packets will still be slow to reach the destination because of upstream congestion, but they will be sent in fewer numbers at longer intervals. The downstream isn't actually slowing down, the ack's are just so frequent that the waiting period becomes too long.

      Analogy: pour 10 gallons of water per minute into a bucket, then wait one minute before repeating.

      Solution: pour 20 gallons of water per minute into a bucket, then wait one minute before repeating.

  70. Fragmentation by Anonymous Coward · · Score: 0

    Set your MTU to 1452 (Win XP defaults to 1480 and the tweak programs and sites rec 1500) and you will get a lot less fragmentation.

    Also make sure the Win XP firewall is set properly and not fully blocking ICMP ECHO. IIRC a router will send an ICMP ECHO if the packet is fragmented, a steady stream of fragmented packets will send an steady stream of ICMP EACHOs back and suck up your incoming bandwidth

  71. Setting up Win2K/XP for broadband by rabtech · · Score: 5, Informative

    You can play with these settings to get the best performance, but in general these should help out some.

    Note that, at most, simply disable then re-enable the network adapter in question. No rebooting should be required to make any of this take effect.

    Keys: HKEY_LOCAL_MACHINE\System\CurrentControlSet\Servic es\Tcpip\Parameters

    GlobalMaxTCPWindowSize:DWORD = 256960
    This should be a multiple of MSS, which is generally MTU - 40. Best results - pick a multiple of MSS lower than 16-bit (65535) times a scale factor that's a power of two. In other words, pick any multiple of MSS as long as it's under 65535, them multiply that by any power of two to arrive at the TCP Window size (RWIN).

    Tcp1323Opts:DWORD = 1
    This enables RFC 1323 options, which allows for a TCP RWIN greater than 64k. If you don't do this, most of the other settings are bunk as they will be limited by the 64k RWIN value.

    EnablePMTUDiscovery:DWORD = 1
    Enables automatic discovery of the MTU for your line, with the MSS set appropriately. You can set this to "0" to force your own value (see below).

    TcpMaxDupAcks:DWORD = 2 (range from 1 to 3)
    Number of duplicate ACKs recv'd for the same seq number of sent data before fast retrans is triggered.

    NOW on to the MTU: it must be set on a per-interface basis. Find your TCP/IP interface associated with your NIC under Parameters\Interfaces\

    MTU:DWORD = 1500 (probably... depends on your provider.)

    On an unrelated note, you can force IE to hit up a web server with more connections than normal, which can help web pages load more concurrently.

    it's under HKEY_CURRENT_USER\Software\Microsoft\Windows\Curre ntVersion\Internet Settings\

    MaxConnectionsPerServer:DWORD = 20
    MaxConnectionsPer1_0Server:DWORD = 20

    First is HTTP1.1+ servers, the other is HTTP1.0 servers. Specifies the max # of connections IE will open to a single web server in the process of downloading a page.

    --
    Natural != (nontoxic || beneficial)
    1. Re:Setting up Win2K/XP for broadband by hta · · Score: 2

      EnablePMTUDiscovery:DWORD = 1
      Enables automatic discovery of the MTU for your line, with the MSS set appropriately.

      note: Path MTU discovery is about discovering the MTU of the PATH, not the line. If you disable path MTU discovery, and make a connection to somewhere that has a lower MTU than 1500 (used to be most of the Internet, now it's rather few places), your packets will be broken up by the intermediate routers, and performance will suffer.
      In some cases, due to stupid firewalls, path MTU discovery doesn't work, and the connection will just hang. That's the time you need to finagle this value.

    2. Re:Setting up Win2K/XP for broadband by essdodson · · Score: 1

      There are specifications as to how many connections you should initiate to a server. Allowing 20 connections to a server is way beyond specifications and is rather unfriendly. IE by default respects the widely accepted standard. Increasing it beyond this is not suggested.

      --
      scott
  72. Re:wtf is this on ask /.? by Anonymous Coward · · Score: 0

    heh, a Real OS, that is

  73. Would a traffc shapper help a 128/128K connection? by BrookHarty · · Score: 2

    Currently I have a 128K/128K connection, I wont let the family touch the idsl when Im playing games. But if there was a way I could limit them to like 24K MAX during my games, I could get away with it.

    Anyone know if this will help on a low speed dsl?

  74. Linux firewall is your best bet by dataentity · · Score: 1

    i would grab an old machine install linux and use it as a gateway. There is a wonderful program call wondershaper that which we use with great success in our offic. We have a similiar connnection and with wondershaper we have 10 people using the connection with no noticeable slow down.

  75. Use your support by orasio · · Score: 2, Interesting

    I believe you are using a commercial OS, so there must be a way they can help you. At least I think a great support system must be reason you are paying for your OS, isn't it?

    Good luck

    1. Re:Use your support by Anonymous Coward · · Score: 0

      ya, and if your using open source OS like linux, dont expect to get any real support, except for RTFM and holy wars in support channels about what distro is best.. Oh, and don't tell anyone you also have windows machines on the network as well, because that just opens a whole can of bullshit about what distro you should be installing on them.

  76. Could it be interleave vs. fast path? by Malc · · Score: 2

    My router (Netgear RT314) has a telnet screen that displays through (Tx & Rx). I haven't checked with my new 1184/160 G.Lite DSL connection, but with my old Nortel (unknown standard) 960/120 I saw this problem. When I was downloading at full speed (105kB/s), I saw 3-4kB/s upstream activity, which was in excess of 20% of the upstream bandwidth. No wonder that fast uploads choked concurrent downloads. I increased my RWIN, but I don't want it above 32K. Unfortunately, this didn't have much affect.

    I think the situation is made worse when the DSL line is set for interleave rather than fast path. On my line, the first hop latency with interleave path is about 70ms, and fast path it is about 10-15ms. I know this because my telco recently tried to switch me over to fast path, which resulted in 40% packet loss due to a higher communication error rates with the modem and the other end (DSLAM port?) at the CO.

    This excellent paper (PostScript format) describes some of this, and in particular, interleave path vs. fast path.

  77. ADSL is DUPLEX by El+Micko · · Score: 3, Informative

    ADSL is duplex.
    ADSL works by dividing the usable range of frequencies on your phone line into segments.

    Lets say that your analogue phone line can support transmission to the local exchange of frequencies from about 0 to 1100 Khz (or there abouts) over copper. The bottom 4Khz is reserved for voice. (I dont know if voice is compressed, I suspect there is some sort of companding in action to give a wider actual response but I dont know.. but the frequency range is adequate.) Then there is a band from about 30Khz to 140 Khz that is used to support the upstream channel, and then frequencies above that out to about 1100Khz are used to support downloads. There is a gaurd band between each band that eliminates crosstalk induced by imperfect transmission conditions.
    This arangement gives you 256Kbps Up and 1500kbps down. Filters are used to isolate the appropriate channels at either end for voice, upload, and download. But the point is that its duplex by design, singnals go both ways SIMULTANEOUSLY over the wire. Remember that ADSL transmission is analogue, thats why you have a MODEM (its an acronym, not a noun!)

    So as has already been explained by previous posters, the most likely problem is that as you flood the upstream channel, your ACK packets are being queued, the network devices upstream then start to throttle back your downstream feed, as the ACKs are taking too long to come back. This is done to minimise the number of packets that will be dropped and resent to your address. Of course you arent dropping any packets , but the upstream devices dont/cant know that.

    Anyway hope that helps. The July 2002 edition of Australian Personal Computer has a lovely graphic that explains how ADSL works and why you can use your phone and have a nice broadband connection at the same time. Unfortunately this article isnt on their site http://www.apcmag.com

    Cheers
    Micko

    1. Re:ADSL is DUPLEX by Dahan · · Score: 2
      Remember that ADSL transmission is analogue, thats why you have a MODEM (its an acronym, not a noun!)

      Yeah, I always found it amusing that ADSL is no more of a "digital subscriber line" than my POTS line running a 2400 bps modem. Now ISDN and SDSL lines, I'd consider digital with their 2B1Q modulation.

  78. Packeteer by Anonymous Coward · · Score: 0

    Excuse me if I'm wrong, but I thought the guy was looking for a inexpensive solution to his problem. Isn't Packeteer's PacketShaper about $50,000 per unit?

  79. Best Solution by shepd · · Score: 2, Informative

    You want it cheap on windows too?

    Your mission for this year:

    - Become proficient in C.
    - Port what you need from Linux to Windows (you suggested Linux already does everything you need).

    Problem solved.

    And sorry to say this, but 99% of the time this is the only way software becomes "cheap", is when someone who wants some software but can't/won't pay for it creates it (or ports it when available). Maybe you'll get lucky and it'll be ported for you, but I doubt it. Its just far easier to set up a linux router for this sort of thing.

    Unfortunately, windows is well known as one of the world's biggest "pay" OSes. You just won't find much free (of consequence) for it that didn't exist on another OS first -- even taping a simple phone conversation will cost far more in software than hardware (source: The Screen Savers).

    I wish you good luck on your search, though!

    --
    If you could be told what you can see or read, then it follows that you could be told what to say or think - BoC
  80. QoS, OpenBSD Router by coene · · Score: 1

    Search for QoS software, I'm sure you can find something. For the *BSD's, the package of choice is ALTQ.

    Most QoS software will let you do everything imagineable to traffic, classifying it and letting you limit and shape it in the ways you want. It takes a tad bit of CPU time, though.

    You might be best to throw an old 486/586 with Linux or BSD on it in front of your PC. OpenBSD is great for this, as it has pf (for firewall & nat) and ALTQ integrated. Check it out.

    1. Re:QoS, OpenBSD Router by Anonymous Coward · · Score: 0

      ALTQ isn't going to FIX THIS!

      altq (at least last time that i checked) cannot give prio's to ACK packages and PF doesn't support TAGging packages with the ACK flag.

      FreeBSD with DUMMYNET could to the job however, but then you will be stuck with ipf instead of pf.

  81. Re:Traffic Shaping Appliance-floppy Linux by Tmack · · Score: 2, Informative
    One floppy Linux distro that I've used with good results is Coyote Linux. Its based on the LRP, boots from one floppy and can do Masq'ing. You can even build a custom kernel to include the QOS stuff, so long as there is enough space on the disk. Once booted, it gives you a basic text menu for information and configuration. iirc it requires a Linux box to setup the distro and write the floppie, but that may have changed.

    I was running this on old 386 and 486 machines with 8MB ram, though I think the newest releases require a little more ram (more ram would also allow for more firewalling rules).

    T

    --
    Support TBI Research: http://www.raisinhope.org
  82. No, this isn't babble.. by hklingon · · Score: 1


    I'm not trolling but, honestly, this thread has an unusual number of people posting information like "dsl is half-duplex!" or "gee, mine isn't like that, what gives?" whatever. Sigh.

    I can't speak to the complexities of the Windows TCP stack, except to say its based on some BSD tcp stack. There are a lot of options you can set in the registry, and you can read more about it here. It is unlikely you will be able to do this "properly" (read: RFC ) on the W2k/WXP default TCP service. there *might* be a different TCP/IP service for XP Pro or server to address this, but its unlikely. No, QoS has *nothing* do do with this. Let me explain.. as few posts seem to get it...

    In TCP you have acks for the data you have recieved. Acks can go back to the sender within data packets (Its just a sequence number, really, sliding window.. protocol 5 I think... use Google if you're lost) or ACKs can go back by themselves. There ARE parameters in W2k for affecting how long W2k will Queue Acks while waiting for a data packet it can piggy-back the ack on. You can decrease this parameter, which may help (or it could backfire). You can also change the protocol window size, so that windows will wait longer*, but this could also backfire as it may overflow the queue in the DSL router (I suspect this in combination that Acks are just plain being choked out).
    The poster seems to understand exactly why their connection becomes throttled, even though most of the replies here don't seem to.... ...

    Now, there might be hope yet. If you use PPPoE, it is a bit more likely you can find a packet scheduler/prioritizer to give Ack packets higher priority. I believe there are a number of PPPoE packages that probably address this problem as part of their design (this sort of thing is somewhat more common than most of the other threads lead you to believe.) I believe WinPoET does, though I can find no information of this on their page. There is also a truly excellent, free, PPPoE package available at (google snapshot) here. though you might have to talk to the author to find out more about its ack-priority algorithm...

    If you don't have PPPoE, you might also look at Nat32 though it only thinly refrences TCP Ack traffic shaping for DSL/Cable asymmetric connections in its documentation.

    I'm not being a zealot, but the *best* thing is probably to use some crappy P133 linux or BSD machine as traffic shaper. You can set up port forwarding for when NAT can't figure out what box to send stuff to, to avoid the downsides of being behind a firewall. The reason for this is that you would probably have to have a special implementation of the TCP/IP service that knows about your upstream and downstream, and prioritizes accordingly. The Registry parameters I mentioned earlier might help, but it won't truly fix the problem.

    *This would be good if you get out of order packets. If network traffic is sporadic, this tends to help. So, windows can re order from a batch of, say, 30 packets vs. 15, windows will wait for a missed packet longer.. this missed packet (which hopefullly won't expire) is more likely to make it in because there is more time for it to do so.

    I apologize for over symplifying... but you get the idea.

    Wendell

  83. rhm traffic shaping on M$ by Anonymous Coward · · Score: 0

    hrm... seems that you have found another limitation of the software.. HEY how bout proxy 2.0 server or IAS, or Checkpiont..

    LOL

  84. OT: slimmp3 by jovlinger · · Score: 0, Offtopic

    How are things in mp3 land?

    Are you guys developing any new mp3 players? Other cool gadgets?

    I'm thinking a slimp3 with _wireless_ ethernet would be pretty cool. Also with spdif output.

    Even more fun would be something I hadn't expected.

  85. Borg? by lab16 · · Score: 0, Offtopic

    Why the bill/borg picture?
    Isn't the main theme about dsl/internet?

  86. Go to DSLreports and run the tune up tools by Anonymous Coward · · Score: 0

    Go to DSLReports and run the tweak test - it will give you an easy way of setting up a larger receive window on your machine which will give you a quick perfoemance boost.

  87. Getting the box is just the first step by LabRat · · Score: 1

    While it seems to me that a *nix-based solution would probably serve you the best (free, powerful, highly-customizable, etc etc), it also seems that most folks have also neglected to mention some strategies that could be used in setting up the traffic shaping. Probably the easiest way to look at things is that you are uploading data using a predictable source or destination port. Simply make that a *low* priority, and everything else will automagically have a higher priority. This is useful when one considers that not all traffic shaping packages have options of specifying traffic such as ACKs as shaping types (many of the simpler packages can only filter on port). Also, in order to maximize the throughput, you should set the overall upstream bandwidth on the shaping box to match the DSL capped value, so that buffers and such on the modem don't fool TCP into not issuing quench requests and such. This way, the shaping box is actually in full control, without interference from "artificial" factors upstream that might screw things up for you. This can be thought of as "matching the impedence" for maximum power throughput in an electrical circuit--same concept and I'm sure someone with more caffeine in their system than I could show the benefits mathematically.

    I personally use a FreeBSD machine (PPro 180MHz) as my NAT/router/firewall (and outgoing SMTP, internal IMAP, mp3 server, photo album, etc) with a number of windoze and *nix machines behind it. As such, it is *way* overpowered for what it does--so suggestions of pentium 90's are very feasible. For the low bandwidth requirements that you have, you could probably even use an old 486 with no problem if you were doing only net traffic functions with it. I've not had to actually do any kind of QoS or shaping with mine, but after thinking about the topic I might just give it a go for the fun of it :) Good luck with yours!

  88. Re:Go cable, baby by Anonymous Coward · · Score: 0

    Unlimited? Have you been to a Falun Gong site lately? Hmmm... how about a site for a free and independent Taiwan?

  89. Ask Slashdot... by bigfatlamer · · Score: 0, Offtopic

    How to fscking spell.

    Peace? Shapping? Jeebus Kriste!

    --
    There's one thing computing teaches you, and that's that there's no point to remembering everything.
    --Doug Copland
  90. how about an API? by BCoates · · Score: 2

    I have pretty much the same problem, and it looks from the posts so far that there isn't any existing solution to do queueing/shaping on XP... Does anyone know where to look for documentation for how you would write a program to do this? What API do windows software firewalls use to intercept packets? I've looked around MSDN a bit, but didn't find anything, is there an unofficial guide somewhere?

    --
    Benjamin Coates

  91. Ok... by angelkey · · Score: 0

    This is probably a troll that I am sinking my teeth into, but have you even tried Win2k or better yet, WinXP Pro? It's far more stable than any Linux windowing (which is what myself and (most) users want - you can shove command-line where the vi don't shine), runs 98% of the apps and games out there, and has hardware support like you would only dream of running Linux. Your comment is both misinformed and sadly mistaken.

    --
    "During times of universal deceit, telling the truth becomes a revolutionary act." - George Orwell, 1984
    1. Re:Ok... by FireBook · · Score: 1

      he's almost right on one score- winxp is the most stable _microsoft_ platform about- bear in mind that its not exactly hard to earn that accolade though :o) nowadays for office only machines i see no reason to use a windows box, its only for games that its really needed (wineX/wine just dont cut it yet)

      --
      My other OS is also FreeBSD
    2. Re:Ok... by Anonymous Coward · · Score: 0

      It was said on slashdot that windows xp is good for gamers. Ha, when I game in windows its in win98se, the windows os with the best support for games. Windows XP is just an "Me"ized version of win2k. To put it in other words "winxp is to win2k what winMe is to win98se", and winMe basicly was the stupid version of win98se. With that said,

      When my friends and I want to play quake3 we use linux because we have found linux will always give us better frame rates than windows. Better frame rates means smother game play and the extra split second we need to frag someone. We have found that anyone of us that uses and windows box to play will all most always loose. Moral of the story: If you want to play solitaire use windows, if you want to play real games use linux.

    3. Re:Ok... by Archfeld · · Score: 2

      I agree but there are very few games to play on my Linux machine and even FEWER people worthy of playing against. I use Linux for a dedicated server for UT, NWN, Quake, and as a firewall/email server, but I like to play NEW games, and I am NOT a coder, so waiting for someone talented enough to port the games really SUCKS :(
      I look forward to the day when games are coded for Linux before release. I payed for many Loki titles even though I'd played the games to death on my Windows machine just to try and help, poor investment strategy :(

      --
      errr....umm...*whooosh* *whoosh* Is this thing on ?
  92. So many pointless answers by gotak · · Score: 0

    First off don't ask questions like this on slashdot. All you get is BS mixed in with a bit of real facts.

    I personally have 3 computers in this house connected to the net. So i have an old Pentium 100 box doing NAT And traffic shaping. Works well.

    I run redhat on the P100 and I use iptables and iproute2 + tc(comes iwth iproute2) to do all the hardwork.

    Basically unless you are willing to fork out for 3rd party software I think there's no real answer to your question. If you have the money google should provide you with a few links and for free too!

    Otherwise get your self a beat up old box and read this site: http://lartc.org/

    Cheers!

  93. Re:Go cable, baby by H3XA · · Score: 1

    I can access everything I want to access..... being Australian tends to make me ignorant of those two issues you mentioned - though I still can read plenty of news stories on those issues from foreign press

    - HeXa

  94. Wondershaper by Anonymous Coward · · Score: 2, Informative

    The Wondershaper is really great. I use it on my ADSL capped at 650kbit upstream and ~475kbit downstream. Without it my ping times when uploading the full 650kbit is >500ms but with wondershaper it's 10-20ms (with a very small speed hit). Full download brings ping times around 100ms, and full speed uploads *never* hurt downloads at all.

  95. Packet Shaper for Mac OS X and BSD varients. by zquestz · · Score: 1

    throttled (http://intrarts.com/throttled.html) is currently available for Mac OS X and should compile fine on most bsd systems. It provides ack packet priority and bandwidth capping using divert sockets and ipfw. Not to mention, it is released under the GPL. Looks like Windows is the only platform that doesn't have a good solution for this problem.

    1. Re:Packet Shaper for Mac OS X and BSD varients. by Anonymous Coward · · Score: 0

      Windows is the only platform that doesn't have a good solution for this problem.

      And so, Windows starts to fall behind the competition. Actually Windows 2000 Server can do QoS if you shell out the big bucks, but I'm sure that's not an option for a lone computer on a home network.

  96. Run the numbers. by tlambert · · Score: 5, Insightful

    His upstream is 96Kbit, and his downstream is 1.5Mbit.

    In round numbers, this means his upstream is slightly less than 1/16th his downstream.

    Say his MTU is 1500 bytes. Let's say his ACK packets are 64 bytes (we were generous on the 1/16th, so 64 instead of 60 is not that far off).

    That basically means that, assuming no packets are lost, and that all ACK packets are precisely equally spaced, then 2/3rds of his upstream bandwidth needs to be dedicated to ACK traffic for the downstream bandwidth.

    Why do we have to assume equal spacing? Because he can't control anything other than his send order, because he can't control the rate limiting machine's discard.

    Is it possible to do this with a traffic shaper on the client machine? Yes, with a very sophisticated traffic shaper, which maintains stateful information (e.g. like a PIX firewall maintains per connection state information). It's possible because now we know the numbers for his connection -- we don't know the general numbers, though, for *any* assymetric link, so this isn't something you could make into an installable package, without the user having to resort to math/tuning tools.

    Even so... this only works if the traffic is connection oriented. So far, people have asked -- and he hasn't responded -- about the kind of traffic he's running.

    If the download traffic is RTSP or UDP or any protocol based on packets other than TCP packets, then there's no way to make preference choices on packets sent out. And then he's back in exactly the same boat he was in before.

    So it's not possible to say "install this", and it's not possible to say "install this, and set these tuning values based on your relative upstream and downstream speeds", unless you really limit the problem you are trying to solve.

    Doing that will probably not be satisfying, since the primary reason for a (mostly) unidirectional pipe is to push content to users, and most content streaming protocols are not based on TCP or other things which can be stateful.

    -- Terry

    1. Re:Run the numbers. by NKJensen · · Score: 1

      That basically means that, assuming no packets are lost, and that all ACK packets are precisely equally spaced, then 2/3rds of his upstream bandwidth needs to be dedicated to ACK traffic for the downstream bandwidth.

      If he limits the output passing via his traffic control to slightly less than the uplink bandwith, the output buffer will never fill.

      The traffic control gets to select which packets it will forward. If it forwards all ACK's this gives 2/3rds of bandwith leaving 1/3 to other (upload) traffic.

      I still don't see the problem? Are you assuming that the input buffer on the traffic control box will overflow and therefor loose ACK's?

      --
      -- From Denmark
    2. Re:Run the numbers. by tlambert · · Score: 2

      Limiting total outbound packets is not necessarily going to help him. Consider:

      o If the outbound packets are limited at random, rather than preferrentially limiting non-"ACK only" packets over ACK packets, then the outbound ACKs are going to be dropped, resulting in the same situation.

      o Even if there is preferential limiting -- which I think I've demonstrated is only possible with virtual circuit protocols -- the problem can still occur.

      At issue is the fact that you can not control the input buffer depth on the outbound rate limiter, unless it's a bridge, not a router. What's going to control it is whether or not the data outbound has been ACK'ed or not, and the normal interactions with your machine on the other side of the limiter's retransmit timers.

      The problem is all about pool retention time. The amount of time that a packet sits in RAM on the rate limiter is going to be your effective introduced latency. The only way you can know what this is going to be is empirically, because there's no way for you to know the buffer size.

      Actually, with all the people wanting to play with window size ("It *has* to work because I *want* it to work!" [stamps feet]), they would be much better off setting the transmit window to 1 packet. This would self-pace transmissions at the transmit rate. Of course, you would still be screwed if you had a lot of transmit connections open behind the rate limiter (e.g. you were running a web or game server, or you were running a peer-to-peer application).

      Actually, optimizing incoming bandwidth at the expense of outgoing bandwidth, even if you could do it successfully on the client side of things, rather than at the rate limiting server, would be the absolute worst possible thing you could do, if you wanted to run peer-to-peer, if both sides of peers were doing the same thing. 8-).

      If you want to look at it another way: you are actually talking about how to go about reserving however much up channel bandwidth is needed to support the down channel.

      -- Terry

    3. Re:Run the numbers. by Anonymous Coward · · Score: 0

      You're forgetting delayed ACKs. ACKs are not sent on every packet reveiced. ACKs are delayed to prevent this type of problem. In other words, I will ACK for the last five packets instead of each packet in succession.

  97. ADSL tweak guide... by frleong · · Score: 4, Informative

    For those interested in knowing how to tweak your ADSL, cable modem settings in Windows, the following link contains excellent and comprehensive information on how to achieve peak download speed: Navas Cable Modem/DSL Tuning GuideTM

    --
    ¦ ©® ±
  98. Some Trafic Shaping Solutions: by AftanGustur · · Score: 3, Informative


    You might want to have a look at the following projects:

    Traffic Control - Next Generation

    Differential Services

    GTC - A Graphical frontend the Linux kernel Traffic Control

    WRR and WIP

    And, yes, those are all Linux solutions, but that's simply because that' all I found available without paying 20.000 dollars.

    --
    echo '[q]sa[ln0=aln80~Psnlbx]16isb572CCB9AE9DB03273snlbxq' |dc
  99. The real cause of the problem, and an easy sol. by pashtunx · · Score: 1

    The real reason that your downstream transfers are becoming sluggish while you are uploading stuff is that the buffers on your cable/dsl modem are being flooded. Remember that you have a 10/100 mbit connection between your NIC and modem. So when you are uploading, you are in fact sending data to your modem faster than it can send off. What you need to do is keep your XP box behind your Linux Filter/Router and compile support for different packet queueing algorithms into your kernel. I use the TBF algorith which stands for Token Bucket Filter. You can use this to limit the rate at which your NIC sends data to your cable modem...thus eliminating the problem of flooded buffers on your modem. That fixes things up real nice.

  100. what kind of adsl modem are you using by aXi · · Score: 0

    If you are using an ethernet adsl modem you shouldn's have the mentioned problem. Yet if you are using a USB modem you might want to get a better driver or something.

  101. Simple solution by terminal.dk · · Score: 2, Interesting

    You don't need traffic shaping. Just make sure your TCP receive buffer size is 15 time larger than your TCP transmit buffer size, and the ACKs will be sent in a timely manner.

    Works perfectly for me. Disadvantage is, that this is a setting you must set on all machines.

    1. Re:Simple solution by jackla · · Score: 1

      Sorry for the ignorance, but do you set it in the interface props or the registry ?

  102. So much disinformation by Nicolas+MONNET · · Score: 4, Interesting

    The latency is -- most likely -- caused by the huge buffers in the modem. It *is* possible to improve the situation locally. It's got nothing to do with asymetric lines or somesuch.

    It's simple: what happens is that the upstream buffer in the DSL modem does'nt prioritize traffic at all, most likely it's just FIFO and big. So if the buffer is 128kB and you're serving a big file, your next Telnet packet is going to have to wait for these 128kB to go up before going itself.

    The solution: have a router that artificially limit the outgoing bandwidth to slightly less that the DSL line permits to make sur the modem's buffer never fills up. Then it's the router's buffers that are filling up; but your router is smarter and you can have it order packet. IE if you have 128kB worth of warez0r waiting to go up, it can decide to let that lone Telnet packet go first.

    Me I installed Wonder Shaper, works very well esp. when you've identified what causes the contention (just add the relevant ports to the junk traffic list), even if I completely saturate the link. There's one thing that doesn't work tho: I discovered that at times I had huge ping, again, even with wshaper. What happens (*I think*) is that my ISP is getting overloaded at times, and my actual bandwidth goes below what I set it to in Wshaper. I have to find a way to improve this.

  103. Two observations by olethrosdc · · Score: 1

    (The first observation is only valid for a TCP session where you data is exchanged between 2 machines - other traffic is ignored)

    While it is true that ACKs can be just empty packets, normally an ACK is piggy-backed on top of a real full packet! So there is no real way to 'prioritize' your ACKs.

    (The second observation covers the multiple connection scenario)

    In your case, let's imagine you are uploading data to server A and getting data from server B.
    You want to prioritize your ACKs to downloaded data... however, depending on the application protocol, you will have application-level ACKs. These will be seen as 'data' by TCP which will still generate an ACK for both server A and B.

    What this means is that *most* of the packets that you are sending to A will *also* have an acknowledgment biggybacked on them.

    Thus, there is no useful way to differenatiate between A and B, at least not on ACK information alone

    Conclusion:
    ACKs cannot by prioritized in a real-world scenario, nor can they be used to infer something about the type of traffic in a particular link.
    The best thing to do is to analyze the traffic in a particular link and adjust the priority of all packets for that link (both up and downstream) depending on user preferences.

    --

    I miss my rubber keyboard.(Homepage)

  104. He's not serving by Perdo · · Score: 2

    He's running an fserv on irc.. His problem is he can't download pr0n fast enough while sending enough to meet his quota.

    --

    If voting were effective, it would be illegal by now.

  105. Re:Go cable, baby by Mrs.Trellis · · Score: 0

    "I can access everything I want to access"

    I was in China a few months ago and though I was only using internet cafes I couldn't get BBC News nor CNN and other similar western 'news' sites. Has this changed or are there different rules for private connections?

  106. big receive window != reduced ACKs by tlambert · · Score: 3, Insightful

    "By increasing the recieve window you are essentially reducing the amount of ACK's that must be sent per data recieved."

    The problem with this theory is that you need to keep sending ACKs all the time to keep the window sliding. Yes, if you were to only send ACKs starting at 50% of the window size received, and you had a large enough receive window that the propagation delay for an ACK through a totally saturated link was less than 50% of the time necessary to receive the data being ACK'ed, and the rate limiter queued all packets, instead of just droping them, then you'd be right.

    But that's not what you do.

    The point of a windowed protocol is that you eat a single round trip latency over a very large data stream consisting of a large number of packets, and what you are saying is that it will act as if TCP/IP is a lock-step fixed window protocol. This just isn't true.

    So you compete for transmit space with the same number of ACK packets.

    The problem is still that you need 2/3 of the send bandwith just for ACKs on a saturated receive bandwidth.

    I would be really surprised if the send bandwidth limit wasn't set with *exactly this* in mind: large enough to handle full speed receives with an MTU of 1500 and an ACK packet size of 60 bytes, plus 50% (1.5Mb/96Kbit = 16, 1500/60 = 25, 25/16 = 150%).

    -- Terry

    1. Re:big receive window != reduced ACKs by Anonymous Coward · · Score: 0

      I've read a number of posts by you on this subject. I'm left with only one of two conclusions:

      1.) you really know what you're talking about,
      or
      2.) you're full of bullshit and make up for it with technobabble.

      'Cause damn, nothing you say makes sense. Whether that means I don't have the technical skills to evaluate your argument or you're just wrong I can't say. Either way you're not doing a very good job of explaining yourself, which makes your argument unconvincing.

      -rdrr2

  107. Re:Would a traffc shapper help a 128/128K connecti by Anonymous Coward · · Score: 0

    this is a case where ALTQ could help BIG time.

    if you let ALTQ give your gaming servers more priority then all other ip/port adresses, then the rest of the webbrowsing will get slower, while your game will run (more) smoothly.

    ALTQ could also limit the rest of the traffic to a max of 24KPBS, but that i think it's wiser too just give your game servers priority on the line.

    that way traffic rates are adjusted automaticly to your game servers needs.

    hope this helped.

  108. Get a Linux router by Pedrito · · Score: 3, Interesting

    Like others said, get a cheapo box throw two ethernet cards in it, load Linux, and use it as a router. I have this setup myself, along with 4 Windows machines behind it.

    Believe me, I sleep better at night, knowing that I have Linux between the Internet and my Windows boxes. There are a number of good firewall/proxy/router tools for Linux. You can then use the traffic shaping software, and more importantly, you don't have to worry about the constant security weaknesses found in Windows that make your machine an easy target for hackers.

  109. Traffic Shaping script. by jetmarc · · Score: 1

    I have made a script for traffic shaping with LINUX. It has the following advantages over many other scripts:

    1. it is not fair. It is designed to play Quake3 online, so Q3 traffic always has priority

    2. it requires only one NIC. most solutions require one NIC for LAN and another one for the WAN uplink

    3. it is able to serve SAMBA shares at high speed. most solutions cap all NIC traffic, and your file systems becomes dog slow. my solution caps LAN traffic at a different rate (in my case LAN=90mbit, DSL_in=450kbit, DSL_out=110kbit).

    You find a copy of the script in usenet:

    http://groups.google.com/groups?hl=en&lr=&ie=UTF -8 &oe=utf-8&selm=af3f5bb5.0203270727.3058629%40posti ng.google.com

    Marc

  110. Latency by Anonymous Coward · · Score: 0

    I have the same essential DSL bandwdith but my problem is that in-game (CS, Quake, etc...) pings go throw the roof when ever my upstream bandwith is saturated. And since it's only 96-128k up, a single user serving a file with gnutella or somesuch will kill my game. No to mention I can't offer as much as I'd like on my website

    I've been considering looking into QOS/Traffic Shaping to solve the problem. Has anyone actually done this?

  111. The full download speed costs upload speed, OK by NKJensen · · Score: 1

    Well, I'm getting confused by now:

    The guy wanted to optimize his download bandwidth and he didn't care about delays?

    To support a full download he must use most of his upload bandwith (2/3rds) for ACK's, right?

    Are you saying that it is not possible to have a local bandwith limiter do this, that is let the ACK's out first and let everything else wait?

    Or are you saying that doing so will not get him a better download speed?

    --
    -- From Denmark
    1. Re:The full download speed costs upload speed, OK by Anonymous Coward · · Score: 0

      "Are you saying that it is not possible to have a local bandwith limiter do this, that is let the ACK's out first and let everything else wait?"

      How do you make a bandwidth limiter preferentially permit outgoing packets that are associated with incoming packets containing data, rather than, say, the SYNs from incoming connections?

      The only way is for the limiter to statefully be aware of what's an incoming connection, and what's an outgoing connection, right?

      You have to assume that the protocol is a virtual circuit protocol (e.g. TCP), and you have to assume that data flows in the virtual circuit are going to be unidirectional (data in one direction, ACKs in the other), not bidirectional.

      -- Terry

    2. Re:The full download speed costs upload speed, OK by NKJensen · · Score: 1

      Well, I thought that ACK packets were easy to find in the input buffer of the bandwith limiter device.

      That is I would not have to care about anything else than the contents of the outgoing packages:

      If is't an ACK, send it first. If it's anything else, let it wait in the queue.

      A SYN packet may be dropped now and then using this scheme, but there must be many times more ACK's than SYN's I think.

      --
      -- From Denmark
  112. half duplex by 177777 · · Score: 1

    i'm currently testing several vendor's dslams and modems for a performance comparison. many modems out there have this nasty habit of being set to half duplex. this will absolutely kill bi-directional traffic flows. now for the bad news... many vendors have unlisted (not available by using ? or help) commands that can put your modem's interface into a full-duplex mode. If you find the command, try it; i think you will like it.

    --
    I haven't lost my mind, It's backed up around here somewhere...
  113. What I did.... by Anonymous Coward · · Score: 0

    The main problem for me is when somebody is uploading something from my Kazaa or WinMX. WinMX has the ability to limit upstream bandwith (very useful). Kazaa has the option in the setup but it's useless. Doesn't work.

    That's why I use WinMX, with other ones, the upstream gets loaded and I cannot download nothing. For other uses ... I don't know, but waht else would you want broadband for?

  114. tcpdump for windows by r6144 · · Score: 1
    Called Windump. It works pretty well (at least on Win98).

    P.S. Where to find the binarys? It really sucks to play with the DDK to compile the sources.

    1. Re:tcpdump for windows by funky+womble · · Score: 1

      Binaries for windump and analyser here (Analyser is a very nice gui for Windows packet capture).

  115. The problem is packet loss by Faulder · · Score: 1

    The problem is that he computer is sending more traffic than the upload cap on the modem and it is discarding some packets. When the outbound ACKs get dropped you get retransmission problems. Changing the window size and such will not solve the problem. He cannot prioritize which packets get dropped by the modem so he is correct in asking for traffic shaping so that he caps his upload to 30 or 40 kilobits below his cap, leaving room for acks to go out.

  116. XP PRO reserves 20% of bandwith by Atl_kevin · · Score: 1

    XP reserves 20% of bandwith for system use. you can change it with in the group policy editor. here's a link http://www.xp-erience.org/sections.php?op=viewarti cle&artid=139

    --
    All that is gold does not glitter, Not all those who wander are lost.
  117. Re:So... by _Neurotic · · Score: 1

    I'm running WinXP and I've also run Win2K. They're both stable, period. I have *never* had a blue screen with XP and can count on one hand the times I had one with Win2K.

    Face it you zealots, the stabilty argument went out the window (pun intended) with WinME. It's over now so quit your childish outbreaks.

  118. Overnet by Anonymous Coward · · Score: 1, Interesting

    The solution is easy. Quit using Direct Connect to leech your movies. Because there's no way to limit the upstream bandwidth within the program it always kills your downstream bandwidth.

    Switch to overnet. www.overnet.com It uses hash tables like eDonkey so that files can't be faked by (insert your favorite __AA here) and it's serverless like Gnutella.

    Check it out with the files at www.wic-net.org limit your upstream bandwidth to 5 or 10k less than your maximum, and enjoy.

  119. SnapGear routers have traffic shaping by Agent137 · · Score: 1

    For $200 or so, you can purchas a traffic shaping router. The SnapGear line (linux inside) supports it in an easy to configure way. Here is a review. SnapGear Review

  120. Netscreen 5 by Anonymous Coward · · Score: 0

    You can get a Netscreen 5 hardware router for $225 from http://www.computerstuff4you.com/. This is a solid state router that doesn't cost much more than the LinkSys or DLink routers, but is much more powerful, and one of the extra features is traffic shaping.

    Another option would be to use an old PC with FreeBSD and dummynet(4) for traffic shaping. The disadvantage is much higher power consumption and probably lower hardware reliability (since the PC has moving parts and the NS5 is solid state).

    I use an NS5 for traffic shaping, NAT (with four public IPs, ~10 private ones...another thing that Linksys routers can't do), and firewall duties and it works really well.

  121. Offtopic? by shepd · · Score: 2, Informative

    Now that's a really pathetic way to put down someone's opinion. especially when it is so ontopic.

    I will so metamoderate that person into non-moderation.

    Posting this at +2 again so people can read it:

    ou want it cheap on windows too?

    Your mission for this year:

    - Become proficient in C.
    - Port what you need from Linux to Windows (you suggested Linux already does everything you need).

    Problem solved.

    And sorry to say this, but 99% of the time this is the only way software becomes "cheap", is when someone who wants some software but can't/won't pay for it creates it (or ports it when available). Maybe you'll get lucky and it'll be ported for you, but I doubt it. Its just far easier to set up a linux router for this sort of thing.

    Unfortunately, windows is well known as one of the world's biggest "pay" OSes. You just won't find much free (of consequence) for it that didn't exist on another OS first -- even taping a simple phone conversation will cost far more in software than hardware (source: The Screen Savers).

    I wish you good luck on your search, though!

    --
    If you could be told what you can see or read, then it follows that you could be told what to say or think - BoC
  122. Cygwin Ports of *NIX software by Anonymous Coward · · Score: 0

    Cygwin is a nice application layer that runs on top of windows and you can run samba, sshd etc. on it. Try searching for some of the linux software mentioned with a cygwin port.

  123. that's not interesting by Anonymous Coward · · Score: 0

    That's some guy saying "I fucked up some settings on my computer and it started working right when I fixed them."

    Will the moderators spend the 2 seconds it takes to read AND understand the content of a post?

    1. Re:that's not interesting by Ilgaz · · Score: 0, Troll

      "That's some guy saying "I fucked up some settings on my computer and it started working right when I fixed them."

      If you will bitch, don't bitch under AC.

      NO, windows doesn'T come with those settings by default. Its a KNOWN fact... Fresh install windows, you will get stupid RWIN, MTU settings.

      Forgot what are NIC's for in fact? Corparate networks. Now, in DSL/Cable age, we use them at home.

      That post deserves credit since we are talking about Windows DSL performance problems here.

      Oh, btw, enable "passive ftp" for download on IE, DSL guys...

    2. Re:that's not interesting by Ilgaz · · Score: 1

      man... congragulations to the "genious" who market me as troll...

      and,yes...this is off-topic...

  124. first off by dmnic · · Score: 1

    well, I would uninstall "QOS" from your network properties list. not only is it not needed, but thats where a good portion of your bandwidth is going.

  125. Gripes from Nerds, Sarcasm that Matters. by Anonymous Coward · · Score: 0

    What a pompous ass.

  126. You should talk to your carrier now! by thype · · Score: 1

    It seems, as mentioned before, that you are having hardware issues, maybe configuration issues. Then again it may be your carrier. DSl, as well as all other WAN connectivity solutions are full duplex. At least I think it is. All other WAN technologies are full duplex, enless someone has some sort of funcky Ethernet connection, which can be changed to half/full within a configuration. Upstream should not effect downstream and vice versa. I would be calling your carrier and ask for a credit until they can get it to work correctly. Unless in your contract it explains that this connection is half duplex, then I geuss you are @#$%?!. I personally would not put up with something like that.

    1. Re:You should talk to your carrier now! by soccerisgod · · Score: 1

      It's ADSL he's talking about, and that means it all uses the same cable, as opposed to SDSL (synchronous) which uses seperate cables for up- and downstream. A decrease in download speed during uploads is normal for adsl.

      --
      If a train station is a place where a train stops, what's a workstation?
  127. Comment removed by account_deleted · · Score: 3, Informative

    Comment removed based on user account deletion

  128. TCP by KagatoLNX · · Score: 2, Informative

    > Full use of your upstream should NOT be
    > crippling your downstream so much, that's not
    > how it works. TCP should adjust accordingly.
    TCP normally piggybacks ACKs. That is, an ACK is sent out on traffic going the other way on the TCP connection. So, if there were only one connection, this isn't a problem.

    The problem lies in asymetric TCP sessions (huge data-filled going upstream with virtually data-less ACK packets coming back) over a congested upstream link. When your upstream gets saturated, the ACKs get caught between huge packets from whatever other traffic). Cable and DSL modems, for example, have a huge queue, so it is easy for your entire TCP window of ACKs to be waiting in queue. The lack of ACKs causes the remote end of the stream to pause sending and kills your throughput.

    In short, TCP is tuned for lower bandwidth links. Applications like satellite connections (where there is huge bandwidth but huge latency) could really benefit from insanely large windows, but you just don't see it much yet.

    --
    I think Mauve has the most RAM. --PHB (Dilbert Comic)
  129. Re:So... by Anonymous Coward · · Score: 0

    WinXP does lock up, like Win9x, for odd stuff on home XP Pro and work XP pro. Microsoft not there yet.

  130. Re: cbq sux, try htb by ldrolez · · Score: 1

    ...and the htb.init script.

    http://luxik.cdi.cz/~devik/qos/htb/
    http://sour ceforge.net/projects/htbinit

  131. ahhhh peace by winse · · Score: 1

    piece brother

    --
    this sig is deprecated
  132. Re:Doesnt this depend on xDSL type? ADSL .. SDSL.e by mini+me · · Score: 1

    I got adsl and have tried stuff like Direct Connect but its really useless since theres no way of limiting OUTGOING speed in DC or at least reserve a couple of kb/sec so it ends up with other ppl dlling from you and yourself getting a few bytes/sec because the more connx outgoing u got the bigger % of the speed it uses.

    There are some modified versions of DC++ floating around that have bandwidth limiting built in. Unfortunately they will not add the code into the main tree (at least that was the case last I heard) so you'll have to hunt for it, or add the code yourself.

  133. Re:Go cable, baby by H3XA · · Score: 1

    BBC and CNN are both still blocked but their are plenty of sources for the same news that are not blocked.

    I sometimes consider CNN and BBC to be the trolls (same crap over and over with hype and FUD whenever possible) of the news world...... I don't miss their websites.

    - HeXa

  134. simple upload limiting by juggy · · Score: 1

    In case you just want to limit your upload ratio, you should put something like this in a file in ip-up.d:

    echo waiting for traffic shaping...
    sleep 5
    tc qdisc add dev ppp0 root tbf burst brate rate ulrate latency 100ms
    echo traffic shaping set

    with brate: burst rate, something like 1500 and ulrate: upload rate (!), something like 114kbps

    1. Re:simple upload limiting by juggy · · Score: 1

      uhm sorry, of course this only works on Linux - but in case anyone is in search for a simple solution for that problem (i had it recently, now i have a firewall/routing-linux-box :), it is a simple solution ;)

  135. Where are you getting those 10-bit bytes? by Ashurnasipal · · Score: 1
    An 8KB TCP window is 80k bits
    B = Byte
    b = bit

    1B = 8b
    8B = 64b

    8KB = 64Kb

    8KB != 80Kb
    1. Re:Where are you getting those 10-bit bytes? by billstewart · · Score: 2

      Modems have start and stop bits on them. Yeah, OK, cable modems, so I should have done 8-bit bytes :-) DSL doesn't have them either, but DSL does have ATM headers which add about 12%. But even 64kbits takes half a second on your 128kbits upstream, which can easily annoy your downstream traffic.

      --

      Bill Stewart
      New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
  136. Coyote Linux by Acetylene5 · · Score: 1

    Try Coyote Linux I've been running a DSL firewall-router off a 486 motherboard, two ethernet cards, and a floppy disk for about a year now.... it's the best thing I've seen out there.

    --
    ---- "Physics is like sex: It has a practical use, but that's not why we do it." -- Richard Feyneman
  137. Do your own r&d by Anonymous Coward · · Score: 0

    Mr Gates....

  138. Just to make things clear :) by jackla · · Score: 1

    Actually I really can't use Linux for this(I don't have the H/W and there's little chance of this changing :/) And I did think of some kind of compact software that sits on the individual Windows machine and does the shaping, in fact there even exists an API dll that seems to be suited for this (LibNet for NT, I think) but no actuall application built on it. And about the whole Linux/Windows rant - I love Linux, I work with it every day but in this case I'm stuck with good(oh-oh), old windows.

    1. Re:Just to make things clear :) by DaKritter · · Score: 1


      Try vmware (noone else said it).

      This way you can run a virtual Linux machine inside Windows, and not buy any hardware (but vmware costs something).

      You would just set up the virtual Linux machine so it controls the ethernet card (DSL connection), and set up Windows to use the virtual Linux as router. It sounds weird that Windows would route all it's traffic to a program running inside it, but it works fine. Of course it costs some performance, but you run XP so your hardware should be capable enough.

      Maybe you can even set up Windows to use the DSL connection directly if the virtual Linux is not up, and when it comes up, let it reconfigure itself to use the Linux as gateway. A DHCP server on the Linux machine should do this.

  139. Re:So... by Anonymous Coward · · Score: 1

    the stabilty argument went out the window (pun intended) with WinME.

    WinME stable? Excuse my language, but, bullshit. I know that's false from experience. Had you stuck with Win2k and XP I wouldn't have responded.

    Besides until Windows is listed here, its stability is still relative.

  140. Sad excuse for a "broadband" connection by Bytenik · · Score: 1

    It is just so sad to see an ADSL connection with a 96kbps upstream rate. This is quite simply not a broadband connection. U.S. phone companies seem to be the worst offenders in this area.

    Our DSL in British Columbia is at least 640kbps upstream, which is much better, though still not great.

    ISPs will tell you that it's so people don't run servers. What about video conferencing ? One of the main reasons the "general public" gets broadband is to see the grandkids or other family and friends over the internet.

    I've even seen advertisements from the ISPs showing video conferencing as a feature. Sure, it will work if the other guy has a real broadband connection, but not if he's on a crappy ADSL connection too.

    It's just sick.

    --

    "Scientists prove we were never here."
    -- Devo

  141. not informative by Anonymous Coward · · Score: 0

    Pointing someone asking for XP help to a Linux How-To is not informative. He is already aware that this is possible under non-MS operating systems. He wants to know if you can do that with XP alone. Christ, don't the moderators actually read posts?

  142. Use a Better OS! Dump Windows Dude! by MrJerryNormandinSir · · Score: 1

    It's time for this guy to get a clue and dump
    windows!

    1. Re:Use a Better OS! Dump Windows Dude! by soccerisgod · · Score: 1

      Wow man.. that was so..helpfull!

      --
      If a train station is a place where a train stops, what's a workstation?
  143. Linux box by Anonymous Coward · · Score: 0

    Installing a simple FreeBSD or Linux box as a gateway you can prioritize upstream ACKs, using HFSC or CBQ, e.x., and you will increase HIGHLY your downstream performance.

  144. People have a choice on DSL providers? by aetherspoon · · Score: 1

    Wow, I wish it was like that here... in South Florida, my DSL ISP is Bellsouth.net. They are the only one. Before you tell me to go hunt for others and such, hear me out.

    If, lets say, I want to go to another ISP, bellsouth tells them that I can't receive DSL service in my area. (Bellsouth is also my local phone carrier. Gee, what surprise) ... no, I'm not kidding. Been through it. Ticks me off too. I'd take a speed cut if I didn't have to put up with my moronic ISP to say the least....

    --
    --- Ãther SPOON!
  145. How mature by Anonymous Coward · · Score: 0

    You do realise that if you continue your moderating rampage you're going to lose your rights when metamod bites you in the ass, huh?

    Because in metamod, you see, they only show the title of the story and the comment, which is 100% on topic.

    Please, mod this down too.

    Enjoy you little spat, moron.

  146. Re:So... by sethgecko · · Score: 1

    He probably meant that the stability argument went out the window when WinME went out the Window (i.e. they stopped selling it)

    --
    Be ot or bot ne ot, taht is the nestquoi.
  147. for maximum download while uploading try... by olla+podriga · · Score: 1

    The Wondershaper: http://lartc.org/wondershaper/
    Works nice for me

  148. Re:ECN and SACK...? Router, not OS by paugq · · Score: 1

    ECN and SACK are features that need to be supported by the routers.

    Even if WinXP supported them (maybe it does, I don't know), most of the routers that nowadays are running the Internet do not support ECN or SACK, so even if the XP's TCP/IP stack would tag your datagrams with ECN and SACK, you would notice nothing.

    In fact, if you look at the ECN help in the Linux kernel you'll see it recommends you not to use it, because it'll cause troubles.

  149. Hardware Router by Anonymous Coward · · Score: 0

    The Snapgear router has some outgoing traffic shaping capability, not sure if it would answer your needs directly. That is a linux-based router-firewall without hard-disk but you dont need to know linux to install it ....

  150. Good MSXP-traffic shapers / routers are available. by Ruudjah · · Score: 1

    Only pricey. Lightspeed systems has one. Everything can be monitored, shaped and routered. 10-day trial avalable. No bullshit or spam, http://www.lightspeedsystems.com/total_traffic_con trol.asp

  151. So in summery, the answer is..... by Anonymous Coward · · Score: 0

    So after reading page after page of bullshit about linux, the answer is NO it cannot be done in windows. Just summing it up if you had problems finding the answer. I also have the exact same problem and I still have no solution.

  152. If you ask specific questions, I'll try to answer by tlambert · · Score: 2

    If you ask specific questions, I'll try to answer in laymans terms, if you are genuinely interested.

    If you aren't interested in laymans terms, and it's just that I'm not a very good explainer (8-)), I can give you references to technical papers.

    -- Terry

  153. WinXP traffic shaping for dialup by ultrapenguin · · Score: 2, Informative

    Actually, I was quite surprised the other day. I actually used a built in modem on my laptop to dialup to check mail and download something I had to have at that moment. few megabytes download started, and I was actaully able to type over ssh link almost laglessly while the download was going at about 2.7k/second or so (this is a modem remember). And I remember on linux, if I didn't throttle the download using some download manager, things would get so laggy I had to wait like 20 seconds to see what I typed over ssh session. Surprising that the same kind of stuff doesn't work with DSL modems...

  154. What a mess by Archfeld · · Score: 2

    Damn Speakeasy won't even talk to me about what line speed I can get until I have severed my contract with Megapath ?!?!?! What a crock, what if they can't or won't support the set-up I have now, it will be to late then. I'd like to look into them but this kind of scratch the other ISP's back BS makes me ILL. It's my line but they can't talk until Megapath releases me ?!?! Fark em all

    --
    errr....umm...*whooosh* *whoosh* Is this thing on ?
  155. QoS on M$ by maximus21 · · Score: 1

    come on y0... guess you found out another thing that Windows is not suitable for.. HRM!!! You need a real firewall product on your os.. like checkpoint and the likes.. SORRY.. hehehehe subliminal message linux firewall/router