11 Sep
2015
11 Sep
'15
10:44 a.m.
On 9/11/2015 11:58 AM, Noel Chiappa wrote:
Strictly true. However if Cogent were blocking them, they would most
likely block them at *ingress* rather than egress - more efficient. My
guess is that it is the next node - the first unidentified node - that
is doing the blocking.
From: Jon
Elson
so MANY others who could not access the
members.iinet page were finding
they got stopped at cogentco.
^^^^^^^^^^^^^^^^^^^
Well, to be precise 'Cogentco was the last node on the route which
responded'.
It's impossible to say whether i) that node tossed packets that tried to go
further; ii) it forwarded them to some other node (identity unknown) which
did toss them (and didn't allow/handle traceroute), etc.
One can't draw any conclusions about whether
it's i) from the fact that it's
_also_ still responding to traceroute packets sent to that address: one would
have to know whether it does the a) 'is this packet to a destination I'm
filtering' check before it does the b) 'I decremented the TTL and it's now
zero', or the other way around. If b) it could be the node that's dropping
the packets.
But given that other 'last hops' are also producing similar results, I'm
still thinking it's Ii.net which is tossing the packets, not the 'last hop'
one can see on traceroutes.
I expect that is correct.
JRJ