|
async transfer?!
![[pl]](/s/countries/pl.gif) Shadow Hawkins on Sunday, 11 January 2004 11:09:13
Hello
Can someone also look and tell me if have the same as me:
[root@fido:~]$ tracepath6 www.sixxs.net
1?: [LOCALHOST] pmtu 1280
1: gw-59.mun-02.de.sixxs.net 56.742ms
2: fe0-1-1-524.br1.enmuc.de.easynet.net asymm 3 89.552ms
3: so0-0-1-0.br0.ixfra.de.easynet.net asymm 4 98.924ms
4: 2001:6f8:0:210::1 asymm 5 67.890ms
5: 2001:6f8::51:20:1 asymm 6 78.403ms
6: so0-1-0-0.gr0.bbpar.fr.easynet.net asymm 7 77.651ms
7: 2001:6f8::65:120:1 asymm 8 84.421ms
8: no reply
9: no reply
10: no reply
11: no reply
12: no reply
13: no reply
14: no reply
15: se1.breda.ipv6.concepts-ict.net asymm 14 136.486ms
16: noc.sixxs.net asymm 15 130.430ms reached
Resume: pmtu 1280 hops 16 back 15
Output from tcpdump:
[root@fido:~]$ tcpdump -i sixxs -n | grep icmp
tcpdump: WARNING: sixxs: no IPv4 address assigned
tcpdump: listening on sixxs
11:06:02.459718 2001:6f8:900:3a::1 > 2001:6f8:93c::a: icmp6: time exceeded in-transit for 2001:838:1:1:210:dcff:fe20:7c7c
11:06:02.562561 2001:6f8:800:1003::209:55 > 2001:6f8:93c::a: icmp6: time exceeded in-transit for 2001:838:1:1:210:dcff:fe20:7c7c
11:06:02.708946 2001:6f8:800::51:1:1 > 2001:6f8:93c::a: icmp6: time exceeded in-transit for 2001:838:1:1:210:dcff:fe20:7c7c
11:06:02.795515 2001:6f8:0:210::1 > 2001:6f8:93c::a: icmp6: time exceeded in-transit for 2001:838:1:1:210:dcff:fe20:7c7c
11:06:02.900379 2001:6f8::51:20:1 > 2001:6f8:93c::a: icmp6: time exceeded in-transit for 2001:838:1:1:210:dcff:fe20:7c7c
11:06:03.024557 2001:6f8::51:11:2 > 2001:6f8:93c::a: icmp6: time exceeded in-transit for 2001:838:1:1:210:dcff:fe20:7c7c
11:06:03.139650 2001:6f8::65:120:1 > 2001:6f8:93c::a: icmp6: time exceeded in-transit for 2001:838:1:1:210:dcff:fe20:7c7c
11:06:24.300542 2001:838:0:10::2 > 2001:6f8:93c::a: icmp6: time exceeded in-transit for 2001:838:1:1:210:dcff:fe20:7c7c
11:06:24.439520 2001:838:1:1:210:dcff:fe20:7c7c > 2001:6f8:93c::a: icmp6: 2001:838:1:1:210:dcff:fe20:7c7c udp port 34194 unreachable
11:06:37.983586 2001:6f8:900:3a::1 > 2001:6f8:900:3a::2: icmp6: neighbor sol: who has 2001:6f8:900:3a::2
11:06:37.983674 2001:6f8:900:3a::2 > 2001:6f8:900:3a::1: icmp6: neighbor adv: tgt is 2001:6f8:900:3a::2
So, the easynet backbone is async? or maybe something at my side is wrong?
async transfer?!
You might want to check your routing tables. Next to that you can check GRH to see which traffic goes where, at least ASPath based.
Routing is usually assymetric btw. That is the nature of the internet and who buys from which transit provider.
async transfer?!
Shadow Hawkins on Sunday, 11 January 2004 18:45:59
This is my routing table:
root@fido:/etc/zebra]$ ip -6 r s
2001:6f8:900:3a::/64 via :: dev sixxs metric 256 mtu 1280 advmss 1220
2001:6f8:93c:a::/64 via :: dev bjaly metric 256 mtu 1280 advmss 1220
2001:6f8:93c:b::/64 via :: dev dsl metric 256 mtu 1280 advmss 1220
2001:6f8:93c:c::/64 via :: dev bash.pl metric 256 mtu 1280 advmss 1220
2001:6f8:93c:d::/64 via :: dev tmn metric 256 mtu 1280 advmss 1220
2001:6f8:993::/48 via 2001:6f8:93c:b::2 dev dsl proto zebra metric 1024 mtu 1280 advmss 1220
3ffe:4019::/32 via 2001:6f8:93c:c::2 dev bash.pl proto zebra metric 1024 mtu 1280 advmss 1220
3ffe:8320:16::/48 via 2001:6f8:93c:c::2 dev bash.pl proto zebra metric 1024 mtu 1280 advmss 1220
fe80::/64 dev eth0 metric 256 mtu 1500 advmss 1220
fe80::/64 dev eth1 metric 256 mtu 1500 advmss 1220
fe80::/64 via :: dev sixxs metric 256 mtu 1280 advmss 1220
fe80::/64 via :: dev bjaly metric 256 mtu 1280 advmss 1220
fe80::/64 via :: dev dsl metric 256 mtu 1280 advmss 1220
fe80::/64 via :: dev bash.pl metric 256 mtu 1280 advmss 1220
fe80::/64 via :: dev tmn metric 256 mtu 1280 advmss 1220
fe80::/64 dev vmnet8 metric 256 mtu 1500 advmss 1440
ff00::/8 dev eth0 metric 256 mtu 1500 advmss 1220
ff00::/8 dev eth1 metric 256 mtu 1500 advmss 1220
ff00::/8 dev sixxs metric 256 mtu 1280 advmss 1220
ff00::/8 dev bjaly metric 256 mtu 1280 advmss 1220
ff00::/8 dev dsl metric 256 mtu 1280 advmss 1220
ff00::/8 dev bash.pl metric 256 mtu 1280 advmss 1220
ff00::/8 dev tmn metric 256 mtu 1280 advmss 1220
ff00::/8 dev vmnet8 metric 256 mtu 1500 advmss 1440
default via 2001:6f8:900:3a::1 dev sixxs proto zebra metric 1024 mtu 1280 advmss 1220
I don't know how to check it on GRH.
I write it, because I remember your words - "async is the worse thing in ipv6" or something like that
async transfer?!
I don't know how to check it on GRH. I write it, because I remember your words - "async is the worse thing in ipv6" or something like that
No, sending traffic with a source address that doesn't belong to the upstream provider.
Thus, if you have the following setup:
ISP-A <--- 2001:db8:1::/48 ---> +--------+
| Client |
ISP-B <--- 2001:db8:2::/48 ---> +--------+
And Client sends traffic over ISP-A's connection using the prefix from ISP-B.
That causes a headache while debugging and also makes the originator totally untracable. Upstream ISP's should ofcourse be filtering that traffic and only allow traffic to be sent which originates from the prefixes that the customer was assigned or prefixes that where negotiated.
Otherwise Client in the above picture could start a DDOS attack using for example 2001:db8:2000::/48 and nobody will be able to trace where the traffic is coming from.
But asynchrouting in a backbone can't be avoided.
async transfer?!
Shadow Hawkins on Sunday, 11 January 2004 20:08:53
I understand, but, is there my fault?
I have sending using sixxs addresses over sixxs tunnel and getting async?
My /48 from sixxs is exported to other isp but at isp router is filtered and no exported to anywhere.
async transfer?!
The only thing that can be done about this is to clean up the 6bone mess and removing the crisscross of tunnels.
|
![[ch]](/s/countries/ch.gif)
on Sunday, 11 January 2004 18:18:56
Posting is only allowed when you are