Another way to see the symptoms are:
hostname:/$> telnet localhost
Trying 127.0.0.1...
telnet: Unable to connect to remote host: Resource temporarily unavailable
Quickly after booting I see tcp_bind_bucket rise.
cat /proc/slabinfo | grep tcp_bind_bucket
But the number of connections etc.. do not reflect the same nubmers that I
see in the /proc/slabinfo.
I was forced to up the ip port range so that the 'pool' is large and my box
can stay up a bit longer.
echo "1024 60000" >/proc/sys/net/ipv4/ip_local_port_range
And quickly after that I started to graph the tcp_bind_bucket:
http://www.doxx.net/load/arsenic.slab..gif
The graph is updated every two minutes, and the machine is about to die,
but I think the graph shows the leak pretty well.
Any ideas?
-Barrett
-
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/