>TCP's retransmission strategies, especially when SACK is available
>on the connection, should even make the most broken of networks
>perform decently.
The problem is that most of servers doesn' t support SACKS ;-). 2.2
releasing will fix this problem of course.
You can get pretty far even without SACKS, if only up to two or three
packets are lost in a single stream of data packets at a time. The
New Reno/Janey Hoe algorithms should resend them rather efficiently.
Actually, our TCP stacks should ack immediately when any out of order
data exists in the receive queue. This was broken in 2.1.x until
2.1.127, it is still broken in 2.0.36prexxx because Alan didn't want
to put that change in at this point.
Anyways, like I said lets wait until the next round of TCP fixes go
in.
Later,
David S. Miller
davem@dm.cobaltmicro.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/