Hello, My linux 2.4.20-gentoo1.4 computer makes a pptp-connection to the Dutch ISP xs4all and also makes an ipv6-connection. The latter as follows: echo 1 > /proc/sys/net/ipv6/conf/default/forwarding echo 1 > /proc/sys/net/ipv6/conf/all/forwarding ip tunnel add unixc mode sit local 80.126.113.xxx remote 194.109.5.241 ttl 64 ip link set unixc up ip -6 route add ::/0 dev xs6all ip -6 route add 2000::/3 via 2001:888:10:c::1 ip -6 route add 2001:888:100c::/64 dev eth2 ip -6 addr add 2001:888:10:c::2/127 dev xs6all ip -6 addr add 2001:888:100c::1/64 dev eth2 ip -6 addr add 2001:888:100c::3 dev xs6all eth2 is my lan and eth1 is 10.0.0.150 for my speedtouch 510i adslmodemrouter (configged so it'll pptp itself in!). my radvd-conf is: interface eth2 { AdvSendAdvert on; AdvHomeAgentFlag off; prefix 2001:888:100c::/64 { AdvOnLink on; AdvAutonomous on; AdvRouterAddr off; }; }; There are too many timeouts happening for clients that get an ipv6-address by radvd. I think the cause lies within my lan! I could press ctrl+c and see what the output is of ping www.kame.net: --- www.kame.net ping statistics --- 5212 packets transmitted, 4403 received, +48 errors, 15% packet loss, time 5217762ms rtt min/avg/max/mdev = 357.390/383.197/1577.899/56.565 ms, pipe 2 tratz $ 15% is too much and so are 48 errors. I just don't know how to repair this, so does anyone else have the time to help me? And by the way, when I ping kame.net with my server, there isn't as much packetloss as is on my client. Mark