What's my gateway

All this talk of moving gateways, what's mine - or where's mine?

nge001.edi-re1

Gonna take a guess at Edinburgh.

Not that I care much, just interested, not going to worry.

TalkTalk Business Next Generation Edge, Edinburgh. Juniper kit and you're on Routing Engine 2 of 2, the other being re0.

I think.

Makes sense. Or does it...

Our London router is connected to BNG1-IXN-LON-RE0. A previous reconnect was BNG4-THN-LON-RE0. But London either way as expected.

And our far north router is connected to nge001.gla-re0. So Glasgow, and that makes sense.

But the Glasgow router is connected to Edinburgh!

Later I'll see where the Edinburgh router is connecting to. Shall we take bets on Glasgow!

You're confusing two different things.

The NGE devices are TalkTalk Business equipment, the BNGs are on Zen's network. You'll connect to the Zen equipment through the TalkTalk Business equipment if you're provisioned via TTB.

Different routers report different things. The 'broadband gateway' Zyxel kit reports is the last link in the chain, the LNS. Mikrotik for instance reports the first link in the chain, the LAC.

Edited by XGS_Is_On (Mon 23-May-22 08:59:33)

In reply to a post by Woolwich:

Later I'll see where the Edinburgh router is connecting to. Shall we take bets on Glasgow!

Appears to be Manchester. Obviously...

🤷‍♂️

In reply to a post by Woolwich:

In reply to a post by Woolwich:

Later I'll see where the Edinburgh router is connecting to. Shall we take bets on Glasgow!

Appears to be Manchester. Obviously...

🤷‍♂️

Did you read my post mentioning the difference between the NGEs and the Zen endpoint / gateway / whatever?

In reply to a post by XGS_Is_On:

Did you read my post mentioning the difference between the NGEs and the Zen endpoint / gateway / whatever?

Aye, I'm just reporting what the router is reporting. Never let the facts get in the way of a good laugh.

In reply to a post by XGS_Is_On:

Different routers report different things. The 'broadband gateway' Zyxel kit reports is the last link in the chain, the LNS. Mikrotik for instance reports the first link in the chain, the LAC.

This is interesting - not something I had considered a possibility before.

In reply to a post by ft247:

In reply to a post by XGS_Is_On:

Different routers report different things. The 'broadband gateway' Zyxel kit reports is the last link in the chain, the LNS. Mikrotik for instance reports the first link in the chain, the LAC.

This is interesting - not something I had considered a possibility before.

Mikrotik reports them as Access Concentrators. A Zen GEA session might have something like:

BNG3.WH-MAN-RE0

While a Zen over TalkTalk Business session will have something like:

nge001.gla-re0

The BNG is Zen's kit and their naming convention. Broadband Network Gateway, and it's Huawei equipment. The NGE refers to a Juniper router owned by TalkTalk Business.

At least that's how it was last time I checked. Either way for obvious reasons you don't see the TTB device in traceroutes.

So I am now on zen's fttp 900 service.

I get "lo0" and "lag-3,2,1"
I am assuming this is all via london?

thanks

No idea: looks like part of a traceroute and doesn't have anything in it to give away where it is. lo0 just means loopback 0, which could be anywhere, every device with an Internet address also has a loopback and some just have a loopback, and the lag part just means a few different physical connections combined together in a Link Aggregation Group.

Edited by XGS_Is_On (Mon 30-May-22 19:02:04)

well I assume I am going through london as I can see

lo0-0.bng1.ixn-lon.zen.net.uk

Indeed you are. You're going through some equipment living in rented space in an Interxion data centre in London.

My first traceroute hop on Zen FTTC is (and has been for some time despite previous issues with disconnects) lo0-0.bng3.wh-man.zen.net.uk - which I presume to be Williams House, Manchester.

I'm curious what happens to my packets between my phone exchange and the first hop in a traceroute. It's not a Zen unbundled exchange (those are, I think, fairly rare) and it's not a TTB line, it's just Openreach. Who carries my data from the telephone exchange to, in this case, Manchester? I presume Zen don't operate a dedicated piece of fibre from every small-town exchange to their gateways. Do Openreach get the data there somehow? I can't get my head around it!

BT Wholesale if you know TTB aren't involved. There are about a thousand exchanges FTTC goes to and they connect to all of them. Zen use them, their own network, Plexus, and TalkTalk Business to get from Openreach to their gateways.

https://www.btwholesale.com/products-and-services/da...

Edited by XGS_Is_On (Thu 02-Jun-22 14:57:01)

In reply to a post by XGS_Is_On:

BT Wholesale if you know TTB aren't involved. There are about a thousand exchanges FTTC goes to and they connect to all of them. Zen use them, their own network, Plexus, and TalkTalk Business to get from Openreach to their gateways.

https://www.btwholesale.com/products-and-services/da...

From personal experience they have either re-configured their plexus or have another supplier to BTW BMC And TT-W I have had experience of all 3 lowest latency was TT W Best logical (the shortest routing /peering was BTW MBC using MCR peering (Geo location is N-WEST ) the highest throughput was either GEA supplier as no BRAS (IP profile rate limiting )

On Openreach Zen use in order of frequency TTB, BT Wholesale and their own Plexus.

TTB for cost, BTW for coverage, Plexus where they can as capacity and coverage permit.

Source: services on the same exchange provisioned over BTW then moved to TTB and service that was offered over Plexus as long as it was 330 or below.

No idea what BTW MBC is or what TT-W are, presume you mean BTW WBMC, BT Wholesale Broadband Managed Connect, and TTB, TalkTalk Business. Neither care about 'best logical' either way they use RADIUS from Zen to point them to gateways which as noted many times on here means the same customer on the same backhaul oscillating between Manchester and London. Nothing at all to do with 'best logical' it's basically round-Robin. If it were a best logical path a customer would never move unless the original site were unavailable and that clearly isn't the case.

In reply to a post by XGS_Is_On:

On Openreach Zen use in order of frequency TTB, BT Wholesale and their own Plexus.

TTB for cost, BTW for coverage, Plexus where they can as capacity and coverage permit.

Source: services on the same exchange provisioned over BTW then moved to TTB and service that was offered over Plexus as long as it was 330 or below.

No idea what BTW MBC is or what TT-W are, presume you mean BTW WBMC, BT Wholesale Broadband Managed Connect, and TTB, TalkTalk Business. Neither care about 'best logical' either way they use RADIUS from Zen to point them to gateways which as noted many times on here means the same customer on the same backhaul oscillating between Manchester and London. Nothing at all to do with 'best logical' it's basically round-Robin. If it were a best logical path a customer would never move unless the original site were unavailable and that clearly isn't the case.

BT Wholesale and TT Wholesale As it's only DATA transit Zen use from both but BTW routed me to the Manchester POP and all peering was out of Manchester NO via London unless there was an issue, TT wholesale , routed me to London no Manchester , Zen's GEA used to route directly to London but with high latency x2 almost to the 1st hop Zen gateway , but they had a single threat throughput issue and i was moved back to BTW , But this time get routed to Manchester but all the peering is done from London, and there is no single thread throughput issue, and base latency is more or less as with btw

In reply to a post by tommy45:

the highest throughput was either GEA supplier as no BRAS (IP profile rate limiting )

They are all GEA suppliers.
They all have some form of IP Profile to rate limit the downstream at the OLT.

The only difference is that BTW make their rate limiting more public/transparent by having an "IP Profile" that you can check.
That doesn't mean TTB or Zen don't rate limit though.

Openreach expect all CP's to rate limit your downstream before it reaches the L2S/OLT to prevent traffic being dropped.

In reply to a post by j0hn83:

In reply to a post by tommy45:

the highest throughput was either GEA supplier as no BRAS (IP profile rate limiting )

They are all GEA suppliers.
They all have some form of IP Profile to rate limit the downstream at the OLT.

The only difference is that BTW make their rate limiting more public/transparent by having an "IP Profile" that you can check.
That doesn't mean TTB or Zen don't rate limit though.

Openreach expect all CP's to rate limit your downstream before it reaches the L2S/OLT to prevent traffic being dropped.

The Downside to BTW IP profile is that it sets a lower throughput limit than others do,
By around 2mbps

This increases when DLM activates High re tx because end users effectively get hit twice , 1, by lower IP profile and 2 by increased headers due to the High RE TX on another layer ,

From my experience 9yrs now with VDSL Starting with Plusnet Even with a MAX BTW IP profile of 77.44 As it was back then, myt max through put was only around 73.5mbps

This didn't significantly increase after i migrated to Zen, again using BTW for data transit , It was

only once was i migrated to TTW GEA did it increase to 75.5mbps ( average ) and i have since been moved to Zen's GEA and throughput is similar but slightly lower So IMO the BTW Is further limiting throughput

Edited by tommy45 (Fri 03-Jun-22 13:24:28)

Reading this thread with interest and got me wondering.

Had an active FTTP connection over Openreach Fibre to Zen since 13 Jan 2018 (previously their LLU ADSL since March 2014), portal gives a plexus migration date of 18/1/208

Had a GEA migration in April 2021

Up until last week I would see the gateway reported as either -vt1.cor1.th-lon.ptn.zen.net.uk or vt1.cor2.th-lon.ptn.zen.net.uk

On checking the router logs (fritzbox 7530) on Wednesday I noticed there'd been a ppp drop about 4 in the morning (I'm aware of the router setting that allows you to specify a ppp session reset and mine is set to between 8-9) and the gateway was now showing as lo0-0.bng5.thn-lon.zen.net.uk
This has changed a couple of times when the router has reset ppp but it’s always one of the ones denoted as bng
I assumed the cor1-vt1 was Huawei kit since I’d seen it for a number of years since I was moved onto Plexus but after reading the comments here I’m now not so sure, I wondered if it was a movement onto Cisco or Juniper hardware as a result of the Government directive that Huawei kit has to be removed from networks by 2027.

Anyone know what I was connected to previously?

Just asking purely out of interest (I’m aware thn-lon indicates it’s in the Telehouse North DC in London)
Thanks

Edited by techguy (Thu 09-Jun-22 22:23:48)

I'm in TW2 and connected to vt1.cor1.lond1.ptn.zen.net.uk.

My Zen supplied Fritz!Box 3490 reports the gateway as being 51.148.72.21 and that's the reverse DNS for vt1.cor1...

Unfortunately I'm currently experiencing 5% packet loss to 8.8.8.8 and nearer 10% to 1.1.1.1 and by BQM is a world of red! Plus, my tested have all be over wired Ethernet, no Wi-Fi.

I've sent Zen support an email this evening.

In reply to a post by bsdnazz:

I'm in TW2 and connected to vt1.cor1.lond1.ptn.zen.net.uk.

My Zen supplied Fritz!Box 3490 reports the gateway as being 51.148.72.21 and that's the reverse DNS for vt1.cor1...

Unfortunately I'm currently experiencing 5% packet loss to 8.8.8.8 and nearer 10% to 1.1.1.1 and by BQM is a world of red! Plus, my tested have all be over wired Ethernet, no Wi-Fi.

I've sent Zen support an email this evening.

Hi

Unplug the router from power for two minutes to allow the session to drop.

Plug back in and wait 5 minutes or so, fixed issues on a number of ISPs by doing just that and Zen's own support message suggests restarting the router.

It could be a gateway issue but it could just be a session issue.

Edited by techguy (Thu 09-Jun-22 22:26:15)

Thanks. Could well have been a session issue.

I pulled the VDSL cable from the master socket, gave it 5 mins and then reinserted it.

Now a 100 ping test has zero packet loss.

Strange.

Glad it got you back up and running