lkml.org 
[lkml]   [2000]   [Sep]   [15]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    From
    SubjectRE: ISSUE: Linux bug/restriction prevents multicast participation
    Date
    I'd like to thank everyone who replied.  For the sake of brevity I'm
    replying with one message instead of several individual replies.

    > ak@suse.de asked "Sounds like a bug in the protocols. Why this
    restriction?"

    The complete port assignment specification is:

    from to application priority
    -------------------------------
    0 16383 unclassified lowest
    16384 32767 audio highest
    32768 49151 whiteboard medium
    49152 65535 video low

    The idea is to allow routers to be able to prioritize multicast traffic
    based on the port number that is used for the session (e.g., audio is given
    priority over video). This prioritization is actually implemented in some
    routers, although there is nothing to prevent a particular media type from
    using any particular port. SDR conforms to this port-assignment
    methodology, and thus affects the port selection choice for most global
    multicast traffic sessions.

    > The workaround is to not enable CONFIG_IP_MASQUERADE in the kernel
    configuration.

    That's what I was afraid of (Red Hat apparently enables CONFIG_IP_MASQUERADE
    by default), but I do like the code diff for the kernel. I look forward to
    it being incorporated in 2.2.18. Since I do development work, on occasion,
    for UC Berkeley's versions of vic and vat I was concerned about having to
    include a "recompile your kernel" step into the readme file for Linux--the
    typical user for these types of apps is used to multi-megabit (if not
    gigabit) network connections, but cowers at the thought of even performing a
    simple kernel update via an RPM.

    Meelis Roos reply goes a step further to question the manner in which the
    port assignments were hard-coded into the kernel with only a single
    purpose/application in mind. This question crossed my mind several times,
    but I didn't want to appear overly assertive in my first post.

    Thanks again,

    MD

    -
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    Please read the FAQ at http://www.tux.org/lkml/

    \
     
     \ /
      Last update: 2005-03-22 12:38    [W:0.049 / U:61.416 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site