|
HeartBeat WinXP client through Linux router problem
![[nl]](/s/countries/nl.gif) Carmen Sandiego on Thursday, 11 December 2003 09:10:17
Hi,
I am having a problem setting up my WinXP client to use IPv6 through a router performing NAT.
I installed the HB client, logged on and selected the tunnel (yes, so far no problem).
The HB tray icon shows a heart symbol every now and then.. no problems yet..
though, any ipv6 traffic won't come true...
"Show me some debug stuff" I hear... here goes:
WinXP: ipconfig (snipped irrelevant info):
Tunnel adapter SixXS:
Connection-specific DNS Suffix . : domain.local
Description . . . . . . . . . . . : Configured Tunnel Interface
Physical Address. . . . . . . . . : C0-A8-64-79
Dhcp Enabled. . . . . . . . . . . : No
IP Address. . . . . . . . . . . . : 3ffe:8114:1000::64b
IP Address. . . . . . . . . . . . : fe80::a:c0a8:6479%10
Default Gateway . . . . . . . . . : ::
DNS Servers . . . . . . . . . . . : fec0:0:0:ffff::1%2
fec0:0:0:ffff::2%2
fec0:0:0:ffff::3%2
NetBIOS over Tcpip. . . . . . . . : Disabled
I cannot see a problem here.. just for your information...
I put up an iptraf on the linux router which performs IPv4 NAT.
I see proto 41 traffic going out to the POP, and I see a ICMP dest Unreachable coming back. here's a dump of some lines shown in the iptraf session:
IP protocol 41 (100 bytes) from 80.60.183.201 to 212.19.192.219 on ppp0
ICMP dest unrch (proto) (128 bytes) from 212.19.192.219 to 192.168.100.121 on ppp0
IP protocol 41 (100 bytes) from 80.60.183.201 to 212.19.192.219 on ppp0
ICMP dest unrch (proto) (128 bytes) from 212.19.192.219 to 192.168.100.121 on ppp0
Yes, my private IP is 192.168.100.121, so I set 'Behind NAT' to 'on' in the SixXS HB client.
No proto 41 traffic is bounced into the firewall, as shown (or: not shown) in the firewall logs.
No further information was at my disposal just yet.. I might be able to find more information through netsh, though, not yet.. ;)
One more thing, in the 'Home' page on noc.sixxs.net I see the tunnel state to 'User Disabled' while the type is HeartBeat. Don't know if this is normal behaviour? Should it turn to User Enabled as soon as the HeartBeat is successful?
Thanks in advance for your time.
HeartBeat WinXP client through Linux router problem
a) Why don't you do it easy and let the Linux box handle the endpoint?
b) Did you forward proto-41 to the internal machine?
I also think that it is really odd that you will see ICMP's returning from the POP when your IP seems to be an RFC1918 address, unless iptraf gets its data after the NAT function has completed.
HeartBeat WinXP client through Linux router problem
Carmen Sandiego on Thursday, 11 December 2003 10:49:07
a) I want to connect my laptop to IPv6, not the particular router I am behind, because my laptop is what I carry around, not the router.
b) Does it have to be a static DNAT to the internal machine? I thought (and kinda hoped) that the conntrack of the router would see that proto-41 has to be forwarded, just like GRE traffic. That's what I wanted to check, but it did not come to that yet, as I did not see any proto-41 traffic returning yet which has to be routed.
My guess is that the information seen is already back-translated by the NAT table before iptraff sees it.. I'll add a tcpdump of the ppp0 (adsl) device in my next post, hope that tells more..
HeartBeat WinXP client through Linux router problem
Carmen Sandiego on Thursday, 11 December 2003 10:58:43
Okay.. weird..
just after I wrote the last post I pressed the send button and that caused my firewall announce an incoming connection of protocol 41 to my tcpkernel. so my connection is just yet ipv6.. (also, the tunnel shows as Enabled in the web interface of sixxs.)
connections to noc.sixxs.net are ipv6, pings look good, but a ping to my ipv6 router at home, mirage.ipv6.hoekstra.nu does not seem to be reachable..
4 pings in windows XP:
Pinging mirage.ipv6.hoekstra.nu [2001:888:118c::1]
from 3ffe:8114:1000::64b with 32 bytes of data:
Reply from 2001:7f8:2:1::1: Hop limit exceeded.
Request timed out.
Reply from 2001:7f8:2:1::1: Hop limit exceeded.
Reply from 2001:808:ffff:209::2: Hop limit exceeded.
Ping statistics for 2001:888:118c::1:
Packets: Sent = 4, Received = 0, Lost = 4 (100% loss),
I'm curious why this happens, any idea?
HeartBeat WinXP client through Linux router problem
XS4all seems to be having some problems:
jeroen@purgatory:~$ traceroute6 2001:888:118c::1
traceroute to 2001:888:118c::1 (2001:888:118c::1) from 3ffe:8114:2000:240:290:27ff:fe24:c19f, 30 hops max, 16 byte packets
1 gw-20.ams-02.nl.sixxs.net (3ffe:8114:1000::26) 19.44 ms 20.208 ms 19.675 ms
2 Amsterdam.core.ipv6.intouch.net (2001:6e0::2) 19.524 ms 18.938 ms 19.045 ms
3 ams-ix.tc2.xs4all.net (2001:7f8:1::a500:3265:2) 19.65 ms !H 20.585 ms !H 21.502 ms !H
Odd that your get your response from the UK6IX btw, check your routing...
inet6num: 2001:07F8:0002::/48
netname: UK6X-IX-20011009
HeartBeat WinXP client through Linux router problem
Carmen Sandiego on Thursday, 11 December 2003 11:47:21
doh:
http://www.xs4all.nl/nieuws/storingen/showreport.php3?id=726
currently no ipv6 available due to some movements..
what a coincidence..
thanks for your time... I'll have to wait until it gets online again ;)
|
![[ch]](/s/countries/ch.gif)
on Thursday, 11 December 2003 10:42:32
Posting is only allowed when you are