Upper/Lower Bounds On NAT Timer - TCP-IP

This is a discussion on Upper/Lower Bounds On NAT Timer - TCP-IP ; Hi All, I'm developing a UDP hole puncher. I would like to know the typical upper and lower bounds for NAT timer. Any information helps. TIA, -Le Chaud Lapin-...

+ Reply to Thread
Results 1 to 3 of 3

Thread: Upper/Lower Bounds On NAT Timer

  1. Upper/Lower Bounds On NAT Timer

    Hi All,

    I'm developing a UDP hole puncher. I would like to know the typical
    upper and lower bounds for NAT timer.

    Any information helps.

    TIA,

    -Le Chaud Lapin-

  2. Re: Upper/Lower Bounds On NAT Timer

    On Oct 7, 10:18*pm, Le Chaud Lapin wrote:
    > Hi All,
    >
    > I'm developing a UDP hole puncher. *I would like to know the typical
    > upper and lower bounds for NAT timer.
    >
    > Any information helps.
    >
    > TIA,
    >
    > -Le Chaud Lapin-


    For what it's worth, source code is available for TCP and UDP hole
    punching. Bryan Ford, who is the author of an article on hole-
    punching (the article can be found at http://www.brynosaurus.com/pub/net/p2pnat/
    ), also provided source code for his hole-punching technique. It's a
    program called "NAT Check" (or, natcheck.exe) and can be found here:

    http://midcom-p2p.sourceforge.net/

    Keep in mind that the hole-punching technique relies on the existence
    of an always-on server (sometimes called a "conspiring" server or a
    "rendezvous" server) that must be publicly accessible on the Internet
    and must be running a program that Mr. Ford calls "natserver.c"

    The code is probably written for Unix, but since it relies on Berkeley
    sockets, it probably can be adapted easily for Windows.

  3. Re: Upper/Lower Bounds On NAT Timer

    On Oct 9, 9:55*am, Malachy Moses wrote:
    > On Oct 7, 10:18*pm, Le Chaud Lapin wrote:
    > For what it's worth, source code is available for TCP and UDP hole
    > punching. *Bryan Ford, who is the author of an article on hole-
    > punching (the article can be found athttp://www.brynosaurus.com/pub/net/p2pnat/
    > ), also provided source code for his hole-punching technique. It's a
    > program called "NAT Check" (or, natcheck.exe) and can be found here:
    >
    > http://midcom-p2p.sourceforge.net/
    >
    > Keep in mind that the hole-punching technique relies on the existence
    > of an always-on server (sometimes called a "conspiring" server or a
    > "rendezvous" server) that must be publicly accessible on the Internet
    > and must be running a program that Mr. Ford calls "natserver.c"
    >
    > The code is probably written for Unix, but since it relies on Berkeley
    > sockets, it probably can be adapted easily for Windows.


    Thanks for the reply. My scheme is very similar (and pretty obvious I
    would say to anyone working with NAT), but there is one signficant
    benefit (I think): Current hole punching schemes work best when NAT
    does not keep changing the port #. My system, because it uses UDP as
    ethernet frames, will work even if the NAT changes the port number
    once every second. But of course, it helps to know what the lower/
    upper bounds for timing out the translation entry to optimize my
    scheme.

    -Le Chaud Lapin-

+ Reply to Thread