FTTP Latency always change once pppoe reset

I'm currently having some issues with my isp (with Openreach). I feel bad because I don't know what the problem is (I'd love to know the reason behind the problem).

Speed is always as good as what the contract said. But the lowest latency I can get is 4ms (with bbc.co.uk or 8.8.8.8). My latency with Hyperoptic on the same test server at the same address is 1-2ms. If I reboot my router, the ping time sometimes will be 12 or 16 ms. It's not jitter, because the latency is very stable. But everytime I reboot the router or reset the pppoe connection, it has a chance (This sounds like a gamble to me) to go back to 4ms.

Hope anyone is interested in this matter, my isp told me that Openreach didn't have any meaningful reply. There's already been an engineer visit and the engineer said there was nothing wrong with their equipment.

Any guesses and suggested are welcome! I can also post tests here!

What’s the issue here exactly? 😎

Oh indeed, I will say I made it clear that I expected my latency to be stable at 4ms or less, not "random".

Or in other words, I want to know why this happened? Because according to my understanding, if it is the normal fluctuation of the Internet, it should return to the lowest value at a certain time.

Have you setup a BQM?

Which ISP? Openreach aren't an ISP.

If Zen for eg, they have Manchester and London exit points on the network, getting Manchester would result in higher latency.

Other ISPs may have the same setup?

I'm on BT Retail and in N.Ireland and my ping is always the same.

This is the case with BQM. I don't know what happened at 4 am. At 10 pm is one ISP maintenance (a core router reboot).

BQM: https://www.thinkbroadband.com/broadband/monitoring/...

I am currently on home broadband with Trunk Networks. I live in East London BTW, so I was really expecting better latency. I have friends in Hampshire and they have a steady 7ms to London on BT Business.

Thanks for your information!

In reply to a post by Otuuu:

This is the case with BQM. I don't know what happened at 4 am. At 10 pm is one ISP maintenance (a core router reboot).

BQM: https://www.thinkbroadband.com/broadband/monitoring/...

I don’t think that’s an issue with the Openreach access network. Those step changes look like a changes in gateway when the router drops and reinitiates the PPPoE session. That’s an issue squarely with Trunk Networks and their own backhaul / core network.

Trunk networks use various backhaul suppliers, one of those is Zen backhaul, and what you describe is typical of Zen, so I bet you are on Zen backhaul, you could always ask Trunk Networks to confirm.

Zen will randomly route you via Manchester or London when you connect, it is just random in order for Zen to split the traffic 50/50 to load balance.

As your gateway is in London, if Zen route you via Manchester then your latency is increased because packets now travel up to Manchester then back down to London to get to your gateway, rather than going to London directly.

One option is to just keep dropping PPP until you get back on a London route and keep doing this each time the connection flips over to Manchester when PPP drops for some reason. This is fairly safe and easy to do with a FTTP connection.

When I was on IDNet (using Zen backhaul) I found that Zen would drop PPP every few weeks overnight, and I think this is because over a few weeks enough customers are doing the route shuffle that their network becomes imbalanced (i.e. too few via Manchester and too many via London) so they unceremoniously drop loads of customers connections overnight to force new PPPs to come up via Manchester. So expect to keep doing this reshuffle every few weeks.

If you move ISP, check which backhaul supplier they use, as quite a few will use Zen, and you find exactly the same thing at your next ISP.