lkml.org 
[lkml]   [2000]   [Feb]   [26]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    Date
    From
    SubjectRe: Socket migration
    On Sat, 26 Feb 2000, a9504915 wrote:

    > Hi, is there any information / projects that are about linux socket migration
    > (i.e. to migrate the socket to another machine as if the socket is created
    > there)? where can I find them? has it been tried before? Thanks in advance!

    I assume, you are talking about connected TCP sockets.

    I have never heard about any implementation, but it can be a very
    interesting thing to implement. In addition to obvious problems
    (serialization of state, reasonable protocol for transmitting it between
    machines) there is a problem related to the nature of TCP/IP -- it's easy
    to send a packet with address other than your own, but to receive one, you
    have to be able to receive _all_ packets in your subnet even if they are
    physically there in the first place (you aren't necessarily on
    non-switched Ethernet, in the same subnet with another box, and even if
    you are, you will have to set promiscuous mode and deal with all kinds of
    garbage, subnet is filled with). Some kind of proxying probably will work
    with remote syscalls like in MOSIX, or with some kind of proxy that can be
    enabled for particular IP addresses/ports set, however this is probably
    not what you are looking for -- performance of this thing will unlikely be
    high enough (I assume, migration of sockets is supposed to be used to
    solve performance problems, not to create additional ones).

    When I was thinking about TCP connections migration for HTTP I came to
    the conclusion that migration of half-connection (sending packets from
    migrated connection to remote end) with proxying another one (receiving
    packets from it) through IP in IP encapsulation or another network
    interface from "old" to "new" host will be sufficient for HTTP because of
    its asymmetric nature (most of data is passed from servers to clients),
    however it maybe much less beneficial for other protocols.

    --
    Alex

    ----------------------------------------------------------------------
    Excellent.. now give users the option to cut your hair you hippie!
    -- Anonymous Coward



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

    \
     
     \ /
      Last update: 2005-03-22 13:56    [W:0.019 / U:31.384 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site