lkml.org 
[lkml]   [1999]   [Oct]   [6]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
    /
    Date
    From
    SubjectRe: [2.2.{12,13p15}] PPP uploads fizzle out and die (tcpdump info incl.)
    On Wed, 6 Oct 1999, Clifford Kite wrote:

    > On Wed, 6 Oct 1999, Steve Dodd wrote:
    >
    > |I'm stuck. FTP downloads over my PPP link are fine, they pretty much saturate
    > |the connection when the remote host can send that fast. Uploads are a different
    > |matter: regardless of the remote host I'm uploading to, the xfer rate withers
    > |away to sod-all and a huge gob of data remains resolutely stuck in the
    >
    > Here is a third-hand suggestion that popped up in a reply to a similar
    > problem on comp.os.linux.networking:
    >
    > # RFC 1323 "TCP Extensions for High Performance"
    > echo -n 0 > /proc/sys/net/ipv4/tcp_window_scaling
    > echo -n 0 > /proc/sys/net/ipv4/tcp_timestamps
    > # RFC 2018
    > echo -n 0 > /proc/sys/net/ipv4/tcp_sack

    I snipped this off and executed it on both my machines.

    Here are the results:

    This sends with a mtu of 1500
    asyncmap 00000000
    pppd was started with no Van Jacobsen all cases with no quoted characters.

    ftp> bin
    200 Type set to I.
    ftp> put platinum.tar.gz
    local: platinum.tar.gz remote: platinum.tar.gz
    200 PORT command successful.
    150 Opening BINARY mode data connection for platinum.tar.gz.
    226 Transfer complete.
    109842 bytes sent in 36.3 secs (3 Kbytes/sec)

    This sends with a mtu of 576
    asyncmap 00000000

    ftp> bin
    200 Type set to I.
    ftp> put platinum.tar.gz
    local: platinum.tar.gz remote: platinum.tar.gz
    200 PORT command successful.
    150 Opening BINARY mode data connection for platinum.tar.gz.
    226 Transfer complete.
    109842 bytes sent in 31 secs (3.5 Kbytes/sec)

    This sends with a mtu of 256
    asyncmap 00000000

    ftp> bin
    200 Type set to I.
    ftp> put platinum.tar.gz
    local: platinum.tar.gz remote: platinum.tar.gz
    200 PORT command successful.
    150 Opening BINARY mode data connection for platinum.tar.gz.
    226 Transfer complete.
    109842 bytes sent in 31.2 secs (3.4 Kbytes/sec)

    This receives with a mtu of 1500
    asyncmap 00000000

    ftp> get platinum.tar.gz
    local: platinum.tar.gz remote: platinum.tar.gz
    200 PORT command successful.
    150 Opening BINARY mode data connection for platinum.tar.gz (109842 bytes).
    226 Transfer complete.
    109842 bytes received in 42.5 secs (2.5 Kbytes/sec)

    This receives with a mtu of 576
    asyncmap 00000000

    ftp> get platinum.tar.gz
    local: platinum.tar.gz remote: platinum.tar.gz
    200 PORT command successful.
    150 Opening BINARY mode data connection for platinum.tar.gz (109842 bytes).
    226 Transfer complete.
    109842 bytes received in 42.9 secs (2.5 Kbytes/sec)

    This receives with a mtu of 256
    asyncmap 00000000

    ftp> get platinum.tar.gz
    local: platinum.tar.gz remote: platinum.tar.gz
    200 PORT command successful.
    150 Opening BINARY mode data connection for platinum.tar.gz (109842 bytes).
    226 Transfer complete.
    109842 bytes received in 41.9 secs (2.6 Kbytes/sec)

    My tentative conclusion is that the mtu does nothing and the transfer
    rate is rotten.

    tcpdump shows that about every other packet is lost. Many more packets
    are lost if I get any other traffic on the link. For instance, if
    I ping a host (1-second pings), the data-flow disappears. Most all
    packets time out and are re-sent. The ping response is in 700 to
    900 ms and many are missing (18% packet loss). If there is no data
    transfer, I can ping with no packet loss.

    The ping transfer time on a quiet link is 190.0 ms with 8 data-bytes.
    This does not compute. Ping flood with 8 data-bytes shows 54% data
    loss. Ping flood with the standard 56 data bytes shows 100% data
    loss.

    Cheers,
    Dick Johnson
    **** FILE SYSTEM WAS MODIFIED ****
    Penguin : Linux version 2.3.13 on an i686 machine (400.59 BogoMips).
    Warning : It's hard to remain at the trailing edge of technology.


    -
    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 13:54    [W:0.227 / U:29.756 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site