In reply to a post by Ixel:

I imagine TalkTalk Business isn't an option yet, mainly because I don't think they sell the 1000/115 package (yet?).

Highland76 is on 1000/115 with Talktalk business (directly from them).
I think it's just Talktalk retail who are yet to launch 1 gig packages.

It might well be area dependant.

Talktalk are unable to sell me any FTTP on my exchange.
They might be wholesaling backhaul for other ISP's on my exchange though.
No idea to be honest.

My connection and both gateways (the original .48.1 and the spun up .56.2) appear have been flat as pancakes for days which is really good.

I'm still on the .56.2 gateway so my baseline ping is around 9ms and previously it was 6ms. Presumably this is due to different routing or backhaul on the gateways. It's no great shakes, but obv. would be better to have the original lower baseline ping.

Hopefully that will be restored when we move back to the initial primary gateway.

There is still something very weird going on. My IPv6 BQM graph is now showing substantial amounts of blue (average latency increase) as well as tons of yellow (peak latency), but I don't really notice anything wrong in practice.

Testing with pings, to Google IPv6 it's as flat as a pancake at 5ms latency:

Text
1 23 45 6	root@nuc1:~# ping6 -q -c100 -i0.1 2001:4860:4860::8888 PING 2001:4860:4860::8888(2001:4860:4860::8888) 56 data bytes  --- 2001:4860:4860::8888 ping statistics ---100 packets transmitted, 100 received, 0% packet loss, time 9943ms rtt min/avg/max/mdev = 4.981/5.175/5.390/0.125 ms

But to the BQM server it's terrible: average latency 33ms and highly variable

Text
1 23 45 67	root@nuc1:~# ping6 -q -c100 -i0.1 -c500 2a02:68:1::164 PING 2a02:68:1::164(2a02:68:1::164) 56 data bytes  --- 2a02:68:1::164 ping statistics ---500 packets transmitted, 500 received, 0% packet loss, time 50184ms rtt min/avg/max/mdev = 4.357/33.327/115.437/19.528 ms, pipe 2root@nuc1:~#

Trying to another destination, one of bytemark's nameservers:

Text
1 23 45 6	root@nuc1:~# ping6 -q -c100 -i0.1 a.ns.bytemark.co.uk. PING a.ns.bytemark.co.uk.(a.ns.bytemark.co.uk (2001:41c8:2::3)) 56 data bytes  --- a.ns.bytemark.co.uk. ping statistics ---100 packets transmitted, 100 received, 0% packet loss, time 9977ms rtt min/avg/max/mdev = 9.607/9.827/10.132/0.164 ms

That's perfectly flat too (higher average latency, since their data centre is further up North, but that's just the laws of Physics)

I can't help thinking that:

1. This is nothing to do with Cerberus' BRASes - since a BRAS problem would affect all destinations, not just some
2. It must be a network problem do with transit to specific sites only (including TBB BQM)

Is this only apparent with IPv6?

I've been seeing the same thing on this gateway we were moved to. IPv4 is flat but IPv6 shows congestion somewhere, flat overnight then rising during the day.

IPv6 on gateway 46.37.56.2
https://www.thinkbroadband.com/broadband/monitoring/...

IPv6 on the original gateway immediately before the issues was flat.

Speed tests using Thinkbroadband and switching between an IPv6 and IPv4 test shows IPv6 often coming in slower on the single tests, nothing major but reflects what is seen on the BQM, so something somewhere on IPv6 is running pretty close or at capacity. This is hopefully temporary until we are swapped back to the original gateway, annoyingly delayed by a week.

I've pretty much decided I'm leaving Cerberus next month as my first year ends and I've paid my first year FTTPoD premium, I don't mind paying more if I feel like I'm getting more, but I haven't felt that has been the case these last couple weeks and we still haven't been reinstated back to as it was before.

@Pheasant yes only apparent on IPv6 on 46.37.56.2, IPv4 is flat as a pancake.

In reply to a post by Pheasant:

Is this only apparent with IPv6?

Yes. Testing to the two TBB endpoints (2a02:68:1::164 and 80.249.99.164), with "ping[6] -q -c100 -i0.1" - during the day v6 has high variability and peak RTT, but v4 is flat as a pancake.

I was moved back to the 46.37.48.1 gateway a few days ago and my IPv6 BQM is still showing bad latency between 8am and midnight. My IPv4 BQM is perfectly fine. I haven't noticed any real world latency issues but I still have an open ticket with Cerberus.

Thanks for the info, so it is not just this gateway some of us found ourselves switched to. Maybe the work delayed to the 14th Nov will add or switch back in some more capacity.

I got moved to the 46.37.48.1 gateway Yesterday and my IPv4 has been great ever since. I haven't reactivated IPv6 yet as due to migrate to Trunk Tomorrow.