|
100% loss on a working tunnel
![[be]](/s/countries/be.gif) Shadow Hawkins on Friday, 06 February 2004 13:58:46
Hi,
I noticed that as of 18.00 last night i have a loss of 100% where my tunnel is working perfectly... I can ping my gateway from outside the network. I can ping to the internet and everything.
So my question is : what is wrong? Why doesn't my statistics work anymore, while i still have full ipv6 functionality on my network, and even to the internet?
Any advise?
100% loss on a working tunnel
1) Show your interface + routing tables
2) Show tcpdump's of you answering icmp requests to the server.
100% loss on a working tunnel
Shadow Hawkins on Friday, 06 February 2004 14:15:04
interfaces :
eth0 Link encap:Ethernet HWaddr 00:50:FC:73:24:5F
inet addr:192.168.0.254 Bcast:192.168.0.255 Mask:255.255.255.0
inet6 addr: fe80::250:fcff:fe73:245f/64 Scope:Link
inet6 addr: 2001:960:2:77::2/64 Scope:Global
//
lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
inet6 addr: ::1/128 Scope:Host
//
ppp0 Link encap:Point-to-Point Protocol
inet addr:217.136.48.xx P-t-P:217.136.48.1 Mask:255.255.255.255
//
sixxs Link encap:IPv6-in-IPv4
inet6 addr: 2001:960:2:77::2/64 Scope:Global
inet6 addr: fe80::c0a8:fe/64 Scope:Link
//
routing tables :
ipv6:
kes:~# ip -f inet6 route
2001:960:2:77::/64 dev eth0 metric 1024 mtu 1500 advmss 1220
2000::/3 via 2001:960:2:77::1 dev sixxs metric 1024 mtu 1280 advmss 1220
fe80::/64 dev eth0 proto kernel metric 256 mtu 1500 advmss 1220
fe80::/64 via :: dev sixxs proto kernel metric 256 mtu 1280 advmss 1220
ff00::/8 dev eth0 proto kernel metric 256 mtu 1500 advmss 1220
ff00::/8 dev sixxs proto kernel metric 256 mtu 1280 advmss 1220
unreachable default dev lo proto none metric -1 error -101 advmss 1220
kes:~# ip -f inet route
217.136.48.1 dev ppp0 proto kernel scope link src 217.136.48.x
192.168.0.0/24 dev eth0 proto kernel scope link src 192.168.0.254
default via 217.136.48.1 dev ppp0
100% loss on a working tunnel
eth0 Link encap:Ethernet HWaddr 00:50:FC:73:24:5F inet addr:192.168.0.254 Bcast:192.168.0.255 Mask:255.255.255.0 inet6 addr: fe80::250:fcff:fe73:245f/64 Scope:Link inet6 addr: 2001:960:2:77::2/64 Scope:Global <SNIP>
Ah, you have the tunnel prefix configured on your local network... that won't work. If you want to have IPv6 in your local network request a subnet and use that.
2001:960:2:77::/64 dev eth0 metric 1024 mtu 1500 advmss 1220
This entry makes all the echo replies go to your eth0 interface and they will never ever end up at the POP, too bad no cigar for you :)
100% loss on a working tunnel
Shadow Hawkins on Friday, 06 February 2004 15:40:00
IPV6 on my local network works ;)
but where does the ping request come from? i can route that specific ip thru the sixxs interface so that it will and up at the POP... is it 2001:bla77::1 ?
100% loss on a working tunnel
Shadow Hawkins on Friday, 06 February 2004 16:23:29
i have added a route for 2001:960:2:77::1 over the sixxs interface... now i get icmp6 request and reply's over that interface :
16:20:54.491213 2001:960:2:77::1 > 2001:960:2:77::2: icmp6: echo request
16:20:54.491449 2001:960:2:77::2 > 2001:960:2:77::1: icmp6: echo reply
16:21:02.051239 2001:960:2:77::1 > 2001:960:2:77::2: icmp6: echo request
16:21:02.051474 2001:960:2:77::2 > 2001:960:2:77::1: icmp6: echo reply
I noticed the stats were good for a few minutes... then went down again and that it started pinging again..... now i'm just waiting for the stats to update.
100% loss on a working tunnel
Notez bien that we will be restricting usage of the tunnel address space in such a manner that only the ::2 address of the /64 is routed to the other side of the tunnel. Thus if you make your setup like above it is bound to break.
100% loss on a working tunnel
Shadow Hawkins on Friday, 06 February 2004 17:30:44
Would you mind explaning me in dutch what you mean? (If you want it can be over e-mail)
PS statistics seem to work again :D
100% loss on a working tunnel
Shadow Hawkins on Friday, 06 February 2004 14:21:10
Here is tcpdump
14:20:15.010518 213.204.193.2 > 217.136.48.69: gw-120.ams-04.nl.sixxs.net > cl-120.ams-04.nl.sixxs.net: icmp6: neighbor sol: who has cl-120.ams-04.nl.sixxs.net (encap)
14:20:15.010735 217.136.48.69 > 213.204.193.2: cl-120.ams-04.nl.sixxs.net > gw-120.ams-04.nl.sixxs.net: icmp6: neighbor adv: tgt is cl-120.ams-04.nl.sixxs.net (encap)
14:20:17.171539 213.204.193.2 > 217.136.48.69: gw-120.ams-04.nl.sixxs.net > cl-120.ams-04.nl.sixxs.net: icmp6: echo request (encap)
14:20:19.009527 213.204.193.2 > 217.136.48.69: gw-120.ams-04.nl.sixxs.net > 2001:960:2:77::51: icmp6: neighbor sol: who has 2001:960:2:77::51 (encap)
14:20:20.009585 213.204.193.2 > 217.136.48.69: gw-120.ams-04.nl.sixxs.net > 2001:960:2:77::51: icmp6: neighbor sol: who has 2001:960:2:77::51 (encap)
14:20:21.009523 213.204.193.2 > 217.136.48.69: gw-120.ams-04.nl.sixxs.net > 2001:960:2:77::51: icmp6: neighbor sol: who has 2001:960:2:77::51 (encap)
14:20:24.880540 213.204.193.2 > 217.136.48.69: gw-120.ams-04.nl.sixxs.net > cl-120.ams-04.nl.sixxs.net: icmp6: echo request (encap)
14:20:28.009542 213.204.193.2 > 217.136.48.69: gw-120.ams-04.nl.sixxs.net > 2001:960:2:77::51: icmp6: neighbor sol: who has 2001:960:2:77::51 (encap)
14:20:29.009547 213.204.193.2 > 217.136.48.69: gw-120.ams-04.nl.sixxs.net > 2001:960:2:77::51: icmp6: neighbor sol: who has 2001:960:2:77::51 (encap)
14:20:30.009557 213.204.193.2 > 217.136.48.69: gw-120.ams-04.nl.sixxs.net > 2001:960:2:77::51: icmp6: neighbor sol: who has 2001:960:2:77::51 (encap)
14:20:32.539538 213.204.193.2 > 217.136.48.69: gw-120.ams-04.nl.sixxs.net > cl-120.ams-04.nl.sixxs.net: icmp6: echo request (encap)
100% loss on a working tunnel
I don't see your host replying to the ICMPv6 Echo Request, which makes it quite apparent that it doesn't ping and thus no statistics and thus your tunnel is marked as down.
|
![[ch]](/s/countries/ch.gif)
on Friday, 06 February 2004 14:02:38
Posting is only allowed when you are