freebsd-dev/share/man/man4/ttcp.4
Jordan K. Hubbard 1130b656e5 Make the long-awaited change from $Id$ to $FreeBSD$
This will make a number of things easier in the future, as well as (finally!)
avoiding the Id-smashing problem which has plagued developers for so long.

Boy, I'm glad we're not using sup anymore.  This update would have been
insane otherwise.
1997-01-14 07:20:47 +00:00

231 lines
6.6 KiB
Groff

.\" Copyright 1994, 1995 Massachusetts Institute of Technology
.\"
.\" Permission to use, copy, modify, and distribute this software and
.\" its documentation for any purpose and without fee is hereby
.\" granted, provided that both the above copyright notice and this
.\" permission notice appear in all copies, that both the above
.\" copyright notice and this permission notice appear in all
.\" supporting documentation, and that the name of M.I.T. not be used
.\" in advertising or publicity pertaining to distribution of the
.\" software without specific, written prior permission. M.I.T. makes
.\" no representations about the suitability of this software for any
.\" purpose. It is provided "as is" without express or implied
.\" warranty.
.\"
.\" THIS SOFTWARE IS PROVIDED BY M.I.T. ``AS IS''. M.I.T. DISCLAIMS
.\" ALL EXPRESS OR IMPLIED WARRANTIES WITH REGARD TO THIS SOFTWARE,
.\" INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF
.\" MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. IN NO EVENT
.\" SHALL M.I.T. BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
.\" SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
.\" LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF
.\" USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
.\" ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
.\" OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT
.\" OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
.\" $FreeBSD$
.\"
.Dd January 18, 1995
.Dt TTCP 4
.Os FreeBSD 2.1
.Sh NAME
.Nm ttcp
.Nd Transmission Control Protocol Extensions for Transactions
.Sh
.Fd #include <sys/socket.h>
.Fd #include <netinet/in.h>
.Fd #include <netinet/tcp.h>
.Ft int
.Fn setsockopt sock IPPROTO_TCP TCP_NOPUSH &One "sizeof One"
.br
.Ft ssize_t
.Fn sendto sock msg len MSG_EOF &sin "sizeof sin"
.br
.Ft ssize_t
.Fn sendto sock msg len MSG_EOF 0 0
.Sh DESCRIPTION
.Tn T/TCP
refers to a set of extensions to the
.Tn TCP
protocol (see
.Xr tcp 4 )
which permit hosts to reliably exchange a small amount of data in a
two-packet exchange, thus eliminating the extra round-trip delays
inherent in a standard
.Tn TCP
connection. The socket interface includes modifications to support
.Tn T/TCP ,
detailed here for the specific case, and in the
.Xr socket 2
and
.Xr send 2
manual pages for the protocol-independent support.
.Tn T/TCP
is defined in RFC 1644.
.Pp
The
.Tn T/TCP
extensions work by including certain options in all segments of a
particular connection, which enable the implementation to avoid the
three-way handshake for all but the first connection between a pair of
hosts. These same options also make it possible to more reliably
recognize old, duplicate packets, which in turn reduces the amount of
time the
.Tn TCP
protocol must maintain state after a connection closes. The
.Dq Li net.inet.tcp.rfc1644
MIB variable can be used to disable
.Tn T/TCP
negotiation at run time; however, the protocol has been designed to
ensure that attempts by non-T/TCP
systems to communicate with T/TCP-enhanced
ones automatically degenerate into standard
.Tn TCP .
.Sh TRANSACTION MODEL
The expected model of a
.Dq transaction
as used by
.Tn T/TCP
is a fairly simple one:
.Bl -enum
.It
A client program generates a request to be sent to the server, which
is small enough to fit in a single
.Tn TCP
segment, and sends a SYN PUSH FIN segment with options and data to the
server.
.It
The server program accepts the request in the same manner as for
regular
.Tn TCP
connections, interprets it, and generates a reply which may be small
enough to fit in a single segment. If it is, the reply is sent in a
single SYN PUSH FIN ACK segment with (different) options and data back
to the client. If not, then the connection degenerates into (almost)
the usual case for
.Tn TCP .
The server then closes its socket.
.It
The client reads the reply and closes its socket.
.El
.Sh CLIENT SUPPORT
Support on the client side is provided by extending the semantics of
the
.Xr sendto 2
and
.Xr sendmsg 2
system calls to understand the notion of
.Dq implied connect
and
.Dq send and shutdown.
To send the request in a transaction, the
.Xr sendto 2
system call is typically used, as in the following example:
.Bd -literal -offset indent
char request[REQ_LEN];
struct sockaddr_in sin;
int sock, req_len;
sock = socket(PF_INET, SOCK_STREAM, 0);
/* prepare request[] and sin */
err = sendto(sock, request, req_len, MSG_EOF,
(struct sockaddr *)&sin, sin.sin_len);
/* do something if error */
req_len = read(sock, request, sizeof request);
close(sock);
/* do something with the reply */
.Ed
.Pp
Note that, after the
call to
.Fn sendto ,
the socket is now in the same state as if the
.Xr connect 2
and
.Xr shutdown 2
system calls had been used. That is to say, the only reasonable
operations to perform on this socket are
.Xr read 2
and
.Xr close 2 .
(Because the client's
.Tn TCP
sender is already shut down, it is not possible to
.Xr connect 2
this socket to another destination.)
.Sh SERVER SUPPORT
There are two different options available for servers using
.Tn T/TCP :
.Bl -enum
.It
Set the
.Dv TCP_NOPUSH
socket option, and use normal
.Xr write 2
calls when formulating the response.
.It
Use
.Xr sendto 2
with the
.Dv MSG_EOF
flag, as in the client, but with the destination unspecified.
.El
.Pp
The first option is generally the appropriate choice when converting
existing servers to use
.Tn T/TCP
extensions; simply add a call to
.Fn setsockopt sock IPPROTO_TCP TCP_NOPUSH &One "sizeof One"
(where
.Va One
is an integer variable with a non-zero value). The server socket must
be closed before any data is sent (unless the socket buffers fill up).
.Pp
The second option is preferable for new servers, and is sometimes easy
enough to retrofit into older servers. In this case, where the reply
phase would ordinarily have included a call to
.Fn write ,
one substitutes:
.Pp
.Dl "sendto(sock, buf, len, MSG_EOF, (struct sockaddr *)0, 0)"
.Pp
In this case, the reply is sent immediately, but as in the client
case, the socket is no lnoger useful for anything and should be
immediately closed.
.Sh MIB VARIABLES
The
.Tn T/TCP
extensions require the
.Dq Li net.inet.tcp.rfc1644
MIB variable to be true in order for the appropriate
.Tn TCP
options to be sent. See
.Xr tcp 4
for more information.
.Sh SEE ALSO
.Xr send 2 ,
.Xr setsockopt 2 ,
.Xr inet 4 ,
.Xr tcp 4
.Rs
.%A R. Braden
.%T "T/TCP \- TCP Extensions for Transactions"
.%O RFC 1644
.Re
.Sh HISTORY
Support for
.Tn T/TCP
first appeared in
.Fx 2.1 ,
based on code written by Bob Braden and Liming Wei at the
University of Southern California, Information Sciences Institute, and
ported by Andras Olah at the University of Twente.