yet again, I have found two possible bugs in the 2.3.x tcp
stack:
- outgoing data isn't being sent over a ppp link. The connection
seemingly stalls forever and eventually bleeds to death -- this
only occurs on _outgoing_ connections.
When I decrease /proc/sys/net/core/[rw]mem* to 8192, the data
is sent out slightly faster (at about one packet every 30 secs),
but still at an unacceptably slow pace.
Small transfers work great. tcp_timestamps is turned off.
- as soon as the program is terminated and the connection reaches
FIN_WAIT1 state, the data _is_ sent out about three times as fast
as in ESTABLISHED state, but it's still intolerable.
There is about 1% packet loss on my connection. This is a situation
out of my control so some software workaround will have to be found.
"Fix the link" is _not_ an option in my case -- I can't control the
occasional electric leakage of radio channels on the local phone
lines...
regards,
Rik -- Open Source: you deserve to be in control of your data.
+-------------------------------------------------------------------+
| Le Reseau netwerksystemen BV: http://www.reseau.nl/ |
| Linux Memory Management site: http://www.linux.eu.org/Linux-MM/ |
| Nederlandse Linux documentatie: http://www.nl.linux.org/ |
+-------------------------------------------------------------------+
-
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/