On 5 mrt 2008, at 16:09, Leslie Daigle wrote:
As mentioned last week -- the wiki is now accessible:
http://wiki.tools.isoc.org/IETF71_IPv4_Outage
Right, it was accessible yesterday, but not anymore, at least, for me:
the pages don't load.
Looks like this is hosted on a 6to4 address. This isn't recommended,
because the quality of 6to4 reachability is highly variable. But
reachability doesn't seem to be the issue:
$ ping6 -c 4 wiki.tools.isoc.org
PING6(56=40+8+8 bytes) 2001:720:410:1001:21b:63ff:fe92:9fbb -->
2002:4e2f:6761::2
--- wiki.tools.isoc.org ping6 statistics ---
4 packets transmitted, 4 packets received, 0% packet loss
round-trip min/avg/max = 133.693/133.840/134.106 ms
So this looks like a path MTU discovery black hole. According to a
traceroute, the web server is terminating the 6to4 link itself and
this server is announcing a 1420 byte MSS:
10:54:54.464280 IP6 2002:4e2f:6761::2.http > 2001:720:410:1001:21b:
63ff:fe92:9fbb.49238: S 2106457901:2106457901(0) ack 809912022 win
5632 <mss 1420,sackOK,timestamp 1184434509 542018613,nop,wscale 7>
This is consistent with a 1500 byte IPv4 MTU - 20 bytes IPv4
encapsulation = a 1480 byte IPv6 MTU. However, the server is
unreachable (from where I'm sitting) for 1480 byte packets:
$ ping6 -c 4 -s 1432 wiki.tools.isoc.org
PING6(1480=40+8+1432 bytes) 2001:720:410:1001:21b:63ff:fe92:9fbb -->
2002:4e2f:6761::2
--- wiki.tools.isoc.org ping6 statistics ---
4 packets transmitted, 0 packets received, 100% packet loss
Can this please be fixed?
In general, people use a 1280 byte MTU for 6to4, which nicely avoids
the possibility of path MTU discovery black holes.
_______________________________________________
IETF mailing list
IETF(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf