Mine is fixed now, so I can't comment!

In reply to a post by SteveBushell999:

So, probably less than surprised. I don't know what the common points between us are, but as Brandon Says:

We don’t believe many people are affected, though FakeJake from the forums for sure. There are very few speed complaints into our Tech Support .

Zen therefore believe our problems are at least similar. Could take a time ! I still believe that how the different OS's and TCP stacks handle the errors is interesting, NOT the problem, but pointing us somewhere. Perhaps if we knew WHY they behave so differently, it may guide us forwards?

I also wonder if we are entheusiasts, and maybe other people DO in fact have problems, but just assume it's 'them' perhaps?

Not saying anything more here, but monitoring it all. My connection is now solid, and fine.

Yeah there are going to be a lot of people, even on higher speed tiers, using mostly WiFi and not really noticing issues or writing them off as WiFi weirdness.

A timely reminder actually that most folks don't even think about broadband FTTC/P/whatever they just have 'WiFi'.

I'll give that a go in a bit.

Field engineer came to do a lot of tests (same bloke that visited Steve).

Was on the phone to Brandon for most of it. Loads of performance tests done (an assortment of generic internet speed tests plus iperf tests to a test server at Zen) both via router and also direct to ONT. Confirmed problem still present and more test results logged.

I guess the test CPE in the exchange is just to add to the evidence list. Very helpful to have lots of data when you're trying to get a 3rd party to look into things!

He even had the new 7530 AX Fritzbox with him. I believe they're still evaluating it. Perhaps I'll be able to get one at some point if I ask nicely 😉

I wouldn't get too excited about the 7530AX - see this 7530 compared to 7530AX They look pretty identical to me!
Good luck with the tests, sounds the same testing that was performed here, which just ended up inconclusive. Did he try Windows 10 straight into ONT on PPPOE.... What was the result?

In reply to a post by SteveBushell999:

I wouldn't get too excited about the 7530AX - see this 7530 compared to 7530AX They look pretty identical to me!

That page is complete pish.

The difference is that the 7530 AX has WiFi 6 whereas the 7530 does not. If you really want to know the difference (other than MU-MIMO that's it) best ask the money grinder direct.

And USB 3.0 .... Sorry, that page was the biggest hit I got looking for a comparison. Still, trying the AX here delivered zero difference, nor did 2 different 7530's even running release or beta code. As for the AX being faster on WiFi? I don't use WiFi for anything other than the phone or iPad TBH, and get 500Mbps, which is all I get on wired - both have mesh, but again my Mesh boxes aren't WiFi6 in any case

In reply to a post by SteveBushell999:

And USB 3.0 ....

Soz, nope, the 7530 already has USB 3.

Interesting that Zen brought other modems, will they flog me one? I'm asking about updating to the 7530 (which is why I know the specs!). Like you I don't care much about WiFi 6, but I'd take it. I'd also buy 7590 if available to replace my aged 7490. But you can't buy FritzBoxes 'officially' in the UK any more other than from Zen.

My pure guess is that there are subtle variations in the implementation of TCP congestion control algorithms at the OS level.

Machines running TCP based performance tests are detecting (perhaps variations in) packet loss and backing off as expected, but backing off to different degrees.

Question is where is the inconsistent pinch point along the route and why….

It would be good if Openreach could compare their configuration for a working GEA migrated Zen FTTP services with a non-working GEA migrated Zen service where both services have the same make of OLT/ONT combo, this would then at least prove its not down to a configuration mistake on their side.

Edited by deleted (Thu 12-May-22 16:33:07)

It could even come down to something like differences in firmware in a line card on otherwise identical kit…