> can not send larger ones. I did some bug hunting and I am believe that
> this because linux kernel slow down tcp connection speed. After each
> frame loss (and retransmitting) tcp_timer.c multiplies rlo by two, but
> never decreases it.
It decreases it when it gets frames that it knows are good and as the
cwnd opens.
> I have the following questions:
> 1. May be the problem is already resolved in linux 2.1.x?
2.1.x supports stuff like SACK/FACK/PAWS - that means it'll work better on
a bad link than almost anything else. If both sides support SACK (and everyone
is deploying it in latest OS releases) then bad links become a lot less of
a problem
> 2. Do my plan violates any RFC?
You should read RFC793 and the later RFCs. Paticularly the ones from 2000 up
about tcp stuff.
> 3. Can the sock structure be changed without recompiling all
> applications?
Yes
> 4. I am not familiar with the code, how I can know that the frame is
> transmitted without retransmission?
You mean if it was acked from the original or a retransmit - you dont
thats part of the fun
> 5. What precisely mean rlo field of sock structure. Where I can find
> additional documentation on sock structure?
The tcp parts of it are bet understood by reading RFC793
-
To unsubscribe from this list: send the line "unsubscribe linux-net" in
the body of a message to [EMAIL PROTECTED]
