Internet wide IP broadcasting possible? - TCP-IP

This is a discussion on Internet wide IP broadcasting possible? - TCP-IP ; Hello, This question is purely out of my own personal curiosity. In the recent days I have been reading a lot about TCP/IP networking and I'm quite excited now I start to understand everything that happens under the hood of ...

+ Reply to Thread
Results 1 to 6 of 6

Thread: Internet wide IP broadcasting possible?

  1. Internet wide IP broadcasting possible?

    Hello,

    This question is purely out of my own personal curiosity. In the
    recent days I have been reading a lot about TCP/IP networking and I'm
    quite excited now I start to understand everything that happens under
    the hood of the internet after so many years of just being a clueless
    user .

    All this made me wonder about IP broadcasting though. I've read that
    broadcast packets can be sent within subnets or LANs (used by f.e.
    DHCP and ARP). A logical conclusion would be that it should be
    possible to send broadcast messages over the entire internet as well
    (packets directed to 255.255.255.255). There must be some convention
    that prevents this which I'm unaware of or something. Because if not,
    wouldn't all that noise be an extremely heavy burden for the entire
    net??

    Thanks in advance for helping a very interested but rather clueless
    networking newbie.

    Tim


  2. Re: Internet wide IP broadcasting possible?

    On Aug 14, 12:17 am, tawm wrote:
    > Hello,
    >
    > This question is purely out of my own personal curiosity. In the
    > recent days I have been reading a lot about TCP/IP networking and I'm
    > quite excited now I start to understand everything that happens under
    > the hood of the internet after so many years of just being a clueless
    > user .
    >
    > All this made me wonder about IP broadcasting though. I've read that
    > broadcast packets can be sent within subnets or LANs (used by f.e.
    > DHCP and ARP). A logical conclusion would be that it should be
    > possible to send broadcast messages over the entire internet as well
    > (packets directed to 255.255.255.255). There must be some convention
    > that prevents this which I'm unaware of or something. Because if not,
    > wouldn't all that noise be an extremely heavy burden for the entire
    > net??


    Well, you are both right and wrong :-)

    It /is/ possible to "send broadcast messages over the entire
    internet". But /not/ through the "broadcast" address. "Broadcast"
    reception is limited to the subnet you broadcast on.

    However, there are a group of IP addresses reserved (as "target"
    addresses) for something called "multicast". "Multicast" packets are
    the broadcast to the internet. You might want to Google for IP
    Multicast in order to get some background material. The Cisco website
    at http://www.cisco.com/univercd/cc/td/....htm#wp1020554
    has what looks to be a reasonable description.

    > Thanks in advance for helping a very interested but rather clueless
    > networking newbie.
    >
    > Tim




  3. Re: Internet wide IP broadcasting possible?

    In article <1187065040.066136.158410@22g2000hsm.googlegroups.c om>, tawm writes:
    > Hello,
    >
    > This question is purely out of my own personal curiosity. In the
    > recent days I have been reading a lot about TCP/IP networking and I'm
    > quite excited now I start to understand everything that happens under
    > the hood of the internet after so many years of just being a clueless
    > user .
    >
    > All this made me wonder about IP broadcasting though. I've read that
    > broadcast packets can be sent within subnets or LANs (used by f.e.
    > DHCP and ARP). A logical conclusion would be that it should be
    > possible to send broadcast messages over the entire internet as well
    > (packets directed to 255.255.255.255). There must be some convention
    > that prevents this which I'm unaware of or something. Because if not,
    > wouldn't all that noise be an extremely heavy burden for the entire
    > net??


    It would indeed be a burden if it worked.

    RFC 1812, section 5.3.5 prohibits the forwarding of 255.255.255.255
    broadcasts by routers. So your 255.255.255.255 broadcast is going to
    hit every host on the local subnet. And nothing else.

    Note also that ARP does not use IP broadcasts. ARP does not even use IP.
    ARP does use Ethernet link layer broadcasts.

  4. Re: Internet wide IP broadcasting possible?

    On Aug 14, 9:42 am, bri...@encompasserve.org wrote:
    > In article <1187065040.066136.158...@22g2000hsm.googlegroups.c om>, tawm writes:
    > > All this made me wonder about IP broadcasting though. I've read that
    > > broadcast packets can be sent within subnets or LANs (used by f.e.
    > > DHCP and ARP). A logical conclusion would be that it should be
    > > possible to send broadcast messages over the entire internet as well
    > > (packets directed to 255.255.255.255). There must be some convention
    > > that prevents this which I'm unaware of or something. Because if not,
    > > wouldn't all that noise be an extremely heavy burden for the entire
    > > net??

    >
    > It would indeed be a burden if it worked.
    >
    > RFC 1812, section 5.3.5 prohibits the forwarding of 255.255.255.255
    > broadcasts by routers. So your 255.255.255.255 broadcast is going to
    > hit every host on the local subnet. And nothing else.


    How ever, IMO, the way multicasting has been done in the Internet is
    not necessarily the way it "should" have been done.

    IOW, I think multicasting should pass the bright-20-year-old test. If
    a 20-year-old can write a multicast application in 1 week that has the
    potential to broacast to a billion users simultaneously from one
    camera-enabled PDA without becoming saturated (unless there is a
    deliberate DoS attack), then you know that you have true, Internet-
    wide multicasting.

    Defining this metric of virtue forces the assumption of certain
    architectural principles that are not at play right now, in either
    IPv4 or IPv6. That's why, IMO, we haven't seen the greatest potential
    for multicast is still ahead of us.

    -Le Chaud Lapin-



  5. Re: Internet wide IP broadcasting possible?

    In article <1187109487.120968.242250@q75g2000hsh.googlegroups. com>,
    jaibuduvin@gmail.com writes:
    > Defining this metric of virtue forces the assumption of certain
    > architectural principles that are not at play right now, in either
    > IPv4 or IPv6. That's why, IMO, we haven't seen the greatest potential
    > for multicast is still ahead of us.


    Metric of virtue? What is that supposed to mean?

  6. Re: Internet wide IP broadcasting possible?

    On Aug 15, 3:12 am, Caleb H wrote:
    > In article <1187109487.120968.242...@q75g2000hsh.googlegroups. com>,
    >
    > jaibudu...@gmail.com writes:
    > > Defining this metric of virtue forces the assumption of certain
    > > architectural principles that are not at play right now, in either
    > > IPv4 or IPv6. That's why, IMO, we haven't seen the greatest potential
    > > for multicast is still ahead of us.

    >
    > Metric of virtue? What is that supposed to mean?


    It is a more specific metric than "it works".

    Speficiation is sometimes needed when talking about whether something
    "works" or not. One could take essentially any tool and say, "there
    is no reason to improve this tool, because what we have works."

    The title this post asked, for example, "Internet wide IP broadcasting
    possible?" If you ask the person or persons who created the current
    "internet wide" IP broadcasting scheme, certainly the answer is "Of
    course." So that question should be qualified to receive a more useful
    answer:

    "To what extent is Internet-wide IP broadcasting possible?"

    I generally use my "Bright-20-Year-Old" test a rhetorical tool to
    remind us of the fact that the "To what extent..." part is very
    important.

    -Le Chaud Lapin-


+ Reply to Thread