Try restricting sendspace and recvspace to 256K. If you're adventurous, try
recompiling the kernel after bumping up the #define SB_MAX (in
/usr/src/sys/sys/socketvar.h) beyond the default value of 256*1024
-AG
-
I've been trying to up my TCP window size from the default 16K and it's caused
> > be possible to do interesting things things like state tracking
> > for individal connections, such that every time we move into the
> > next state instead of doing something like tp->t_state = TCPS_SYN_RCVD,
> > we can have something like tp->t_state |= TCPS_SYN_RCVD . The only
> > useful p
Luigi,
A slightly more concise way of making the same change would be as follows:
#define TCPS_HAVERCVDFIN(s) ( (s) >= TCPS_CLOSE_WAIT || ((s) >= TCPS_CLOSING &&
(s) != TCPS_FIN_WAIT_2) )
With the bit mask based scheme you have suggested, it might also be possible to do
interesting things
Would appreciate comments on this:
Currently in tcp_fsm.h its defined as :
#define TCPS_HAVERCVDFIN(s) ((s) >= TCPS_TIME_WAIT)
when IMHO it should be: (to consider all possible cases of having recd a FIN?)
#define TCPS_HAVERCVDFIN(s) ( ((s) >= TCPS_TIME_WAIT) || ((s) == TCPS_CLOSE_WAIT)
Thanks to all who responded. Luigi, I will also have a look at the code that you
posted. I did have a look at the code from the UCB Daedalus project, which was based
on the BSDI 2.0 code. I am currently basing most of my code changes on that work. The
changes broadly relate to the relatively n
Hi
I am currently working on implementing SACK on FreeBSD 4.3 (STABLE) . At some point in
the future I plan to contribute this patch to the FreeBSD source tree (4.3 or later).
I had a look at
http://www.freebsd.org/doc/en_US.ISO_8859-1/books/handbook/contrib.html .
With regards to the code con
