[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [lwip-users] Recent tcp_rexmit() changes
From: |
K.J. Mansley |
Subject: |
Re: [lwip-users] Recent tcp_rexmit() changes |
Date: |
23 Jul 2004 13:51:04 +0100 |
On Fri, 2004-07-23 at 13:38, K.J. Mansley wrote:
> Perhaps we should just implement Selective ACKs - that would certainly
> help in this case!
Indeed, from the summary of the problem SACK is trying to solve:
"Multiple packet losses from a window of data can have a catastrophic
effect on TCP throughput. TCP uses a cumulative acknowledgment scheme in
which received segments that are not at the left edge of the receive
window are not acknowledged. This forces the sender to either wait a
roundtrip time to find out about each lost packet, or to unnecessarily
retransmit segments which have been correctly received."
(http://www.opalsoft.net/qos/TCP-90.htm)
Which seems to describe pretty much what you're seeing.
Would SACK be useful for you? i.e. Do the clients you're communicating
with support it?
Kieran
- [lwip-users] Recent tcp_rexmit() changes, Karl Jeacle, 2004/07/22
- Re: [lwip-users] Recent tcp_rexmit() changes, K.J. Mansley, 2004/07/23
- Re: [lwip-users] Recent tcp_rexmit() changes, Karl Jeacle, 2004/07/23
- Re: [lwip-users] Recent tcp_rexmit() changes, Karl Jeacle, 2004/07/25
- Re: [lwip-users] Recent tcp_rexmit() changes, K.J. Mansley, 2004/07/26
- Re: [lwip-users] Recent tcp_rexmit() changes, Karl Jeacle, 2004/07/26
- Re: [lwip-users] Recent tcp_rexmit() changes, Leon Woestenberg, 2004/07/26
- Re: [lwip-users] Recent tcp_rexmit() changes, K.J. Mansley, 2004/07/26
- Re: [lwip-users] Recent tcp_rexmit() changes, Leon Woestenberg, 2004/07/26
- Re: [lwip-users] Recent tcp_rexmit() changes, Sam Jansen, 2004/07/26
- Re: [lwip-users] Recent tcp_rexmit() changes, Kieran Mansley, 2004/07/27