Question

is there any built in functionality (flag, parameter, whatever..) to send a TCP packet only then when its MTU is (nearly) full?

I hope I’m right with my assumption, that data is sent always and it doesn’t matter how much payload it contains (tried to find that out via wireshark).

best regards

Answer 1

I hope I’m right with my assumption, that data is sent always

No, a TCP stack typically bundles up data and send big segments (it doesn’t wait forever though, often just a little while) – usually Nagles algorithm or some variant is employed.

There might be other concerns too that impact how data are buffered up and sent, e.g. how congested the network is. Generally the TCP stack is very good at achiving max throughput, and normally you shouldn’t try to outsmart it.

If you need lower latency though, you can disable Nagle’s algoritm, by setting the TCP_NODELAY socket option

int ndelay = 1;
setsockopt(sock,IPPROTO_TCP,TCP_NODELAY,(char *)&ndelay ,sizeof(ndelay));

Linux provides the oposite as well, perhaps closer to what you’re asking , by the means of TCP_CORK.

TCP_CORK If set, don’t send out
partial frames. All queued partial
frames are sent when the option is
cleared again. This is useful for
prepending headers before calling
sendfile(2), or for throughput
optimization. As currently
implemented, there is a 200
millisecond ceiling on the time for
which output is corked by TCP_CORK. If
this ceiling is reached, then queued
data is automatically transmitted.
This option can be combined with
TCP_NODELAY only since Linux 2.5.71.
This option should not be used in code
intended to be portable.