On Thu, 4 May 2000, Alan Cox wrote:
> > > The Sun used a larger window (64K versus 32K).
> >
> > Check that Networking options ... IP: Allow large windows is enabled in
> > your kernel build. This could be related to case 2?
>
> We will only offer 32K windows if the box the other end offers no window
> scaling either way. This is because we care about interworking with old
> equipment with sign handling bugs.
These Sun boxes certainly shouldn't fall into that category, though,
surely!
> Turning on large window support gives a bigger window but while that
> reduces the number of zero window stalls it doesnt change total throughput
Unless the zero window stalls are causing the throughput bottleneck...
> > Using a larger receive window on the Linux end could help, anyway. Any
> > ideas WHY those packets were dropped, BTW?
>
> This is the important one I think
Yes... Could the two issues be related? (i.e. window problems causing
occasional dropped packets?)
James.
-
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/
This archive was generated by hypermail 2b29 : Sun May 07 2000 - 21:00:14 EST