Re: ip_local_port_range

From: David S. Miller (davem@redhat.com)
Date: Fri Mar 31 2000 - 23:36:37 EST

  • Next message: Boris Okun: "Re: problems with do_slow_gettimeoffset()"

       Date: Fri, 31 Mar 2000 23:15:22 -0500 (EST)
       From: "Albert D. Cahalan" <acahalan@cs.uml.edu>

       Why would we use fewer ports, especially when it leads to
       reduced performance? We aren't even in the legal range.

    Because it is a system resource and a protocol correctness
    issue, TIME_WAIT state has a specific purpose and machines
    with lower amounts of ram should not be using such a large
    portion of the available port space.

       IMHO we also need a way to force-expire old ports. Accepting
       new connections is more important than waiting to make sure
       the old ones are really dead.

    Look at the TIME_WAIT recycling mechanism present in the
    current 2.3.x, it is a dynamic scheme meant to early purge
    closed TCP sockets whenever feasible.

    Later,
    David S. Miller
    davem@redhat.com

    -
    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/



    This archive was generated by hypermail 2b29 : Fri Mar 31 2000 - 23:46:19 EST