IPv6 and the outage 17/09 (82855)

[#1]

Anyone else notice this?

[#2]

Think I saw it as well. In Surrey (LSMOG), escaped the disconnect.

[#3]

Quote from: Bob

Anyone else notice this?

Yes

[#4]

Not sure if it's related but some of my traceroutes in IPv6 seem to convert to IPv4 after leaving PN's network:

2  2a02:16c8:0:1::1a  6.306 ms  6.594 ms  8.185 ms
3  2a02:16c8:1:2::11  7.777 ms  6.782 ms  9.561 ms
4  2a02:16c8::a  6.739 ms  5.956 ms  6.394 ms
5  40ge1-3.core1.lon2.he.net  16.673 ms  7.643 ms  19.163 ms
6  linx-gw1.thn.ncuk.net  6.230 ms  13.452 ms  8.153 ms
7  gi1-0-24-10-star1.core-rs2.thdo.ncuk.net  9.581 ms  9.385 ms  8.959 ms
Edit: Also seeing packet loss
--- 2a02:68:1::4 ping6 statistics ---
100 packets transmitted, 85 packets received, 15.0% packet loss
round-trip min/avg/max/std-dev = 6.540/20.169/129.238/31.687 ms

[#5]

They're just resolving to hostnames, not converting to IPv4 addresses. If you do the traceroute again with -n it should only show the IP addresses.

[#6]

On Windows pathping is useful as it will show the hop the packet loss occurs at.

[#7]

no
check both my graphs in sig I was affected on both ipv4 and ipv6.

[#8]

Quote from: avatastic

They're just resolving to hostnames, not converting to IPv4 addresses. If you do the traceroute again with -n it should only show the IP addresses.

Thanks - for some reason they didn't resolve to hostnames before. Ignore my comment anyway 
As for the packet loss - I assume the explanation is that the AG gateways were heavily (over)loaded with end users for a period last night due to the BNG outage.

[#9]

I suspect you're right Andy, just struck me as odd that it affected the IPv6 transport but not IPv4.

[#10]

Could it be the actual bandwidth capacity of the E320s though? I assume this doesn't differentiate with IPv4 and v6.

[#11]

Quote from: AndyH

I assume this doesn't differentiate with IPv4 and v6.

Thus my confusion. I doubt I'll be able to get any eyes on this one though. Far bigger fish to fry given the problems last night!