Jeff DeFouw <mrj@i2k.com> writes:
> They're already set to 65535, but I don't understand how increasing a
> window which seems to be showing 30k free affects the retransmit delays.
> Am I reading it wrong or am I just too uneducated on the inner-workings of
> TCP?
I think ppp over modem is a special case where a larger window is very
often bad depending on what hardware is in between --- the reason
(IMO) is that most of window is going to get queued up in your ISP at
the choke point --- their terminal controller/modem hardware --- and a
large window might be big enough to tickle problems on their end.
OpenBSD uses a window of 16384 and I find that that gives me a very
good transfer rate with linux-2.0.* without ever causing stalls which
will occur off and on when using the default Window.
BTW I have noticed one of my ISP's sending "source quench" ICMP
packets on occasion when using the default window (which Stevens says
is deprecated by RFC 1812):
[snip]
DISCUSSION
Research seems to suggest that Source Quench consumes network
bandwidth but is an ineffective (and unfair) antidote to
congestion. See, for example, [INTERNET:9] and [INTERNET:10].
Section [5.3.6] discusses the current thinking on how routers
ought to deal with overload and network congestion.
[snip]
How does Linux react to this?
> --
> Jeff DeFouw <mrj@i2k.com>
> I2K Staff - http://www.i2k.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/faq.html