Zen latency huge drop.

So 3 nights ago at a few minutes past midnight i briefly lost internet but it came backup a few minutes later. After it reconnected i check my router graph on Thinkbroadband and my average latency went from 18ms down to 8ms

Did they change something recently as my connection has been about 18ms for the entire 19 months i've been with Zen or when it reconnected did i win some kind of lottery and now i never want to reboot my router ever again.


Thanks in advance.

Maybe you were moved over to the TalkTalk backhaul.

Obvs, I only half know what I'm on about and maybe someone with a full clue will be along shortly...

Several possibilities exist. On the modem side perhaps interleaving was on but is now off. Keeping an eye on modem stats can reveal if this happened.

For me though, it's the gateway, and it happens a lot. Take a look at this TTB graph from this morning.

My latency has doubled. I didn't even lose sync; the drop was with PPP.

Fast:

1 <1 ms <1 ms <1 ms fritz.box [192.168.0.1]
2 40 ms 12 ms 12 ms vt1.cor1.lond1.ptn.zen.net.uk [51.148.72.21]
3 12 ms 12 ms 12 ms ae-8.p1.thn-lon.zen.net.uk [51.148.73.158]

Slow:

1 <1 ms <1 ms <1 ms fritz.box [192.168.0.1]
2 23 ms 23 ms 24 ms vt1.cor2.lond2.ptn.zen.net.uk [51.148.72.24]
3 23 ms 23 ms 23 ms lag-8.p1.ixn-lon.zen.net.uk [51.148.73.188]

cor1.lond1 vs cor2.lond2.

Mines never done this so was a bit confused to see my latency drop my 10ms, not complaining but i'd like to know why.

Although i've just thought about something. Been arguing with Zen for a few months because BTT + openreach have both said my equipment supports their 900/110 FTTP but my order always gets auto cancelled by OpenReach for some reason.

Two months on i finally got Zen to speak to OpenReach to update their system so it will allow me to get the 900/110 and that was early this week so i'm now wondering if it's anything to do with that.

waiting for Zen to offer 900/900 with city fibre

i know talk talk and vodafone offer it

I see this regularly in both directions - as in gets worse or better. Worth checking your logs to see exactly which ip address port you are connecting too on Zen’s side. In my case connecting to an IP address ending .23 or .24 results in a >20ms latency figure, whereas if it ends .21 or ,22 it’s just 7ms. I can see a step change in the BQM.

No change in the Openreach DSL sync rate at all. This change in latency happens every week or so and I use the TTB BQM to check what the line looks like every day. I have raised this with Zen and they say the higher latency ports are still ‘in spec’ and that they can also see the same as I do.

If I notice a latency change on the BQM then I just get the Fritz! Modem to drop the connection without dropping the line and after a couple of tries I end up moving from a .23 or .24 address to a .21 or .22 and job done.

I can only assume it’s a dodgy line card (or whatever users connect too) in their racks.

https://imgur.com/1mxW5OP

That's basically how it always was and now how it's been for 3 days.

A Manchester vs London routing change can be worth about 10ms, depending on the details of where you are in the country and where you're measuring to.

I'm in York but honestly this is the first time It's ever jumped in any direction. Is it likely to stay this way or am i running on a timer?

In reply to a post by Steikums:

I'm in York but honestly this is the first time It's ever jumped in any direction. Is it likely to stay this way or am i running on a timer?

I'm honestly not sure, the latest information I have is that it was more likely to happen where Plexus (Zen's own backhaul network) had reached the handover exchange, but that it might still be load balanced to the further away one at a lower layer. My information is over a year old though, so it may no longer be accurate / relevant.