My IP has changed (UBR has moved) and speeds much better now?

[Pontiac]

Hi fellas,

For as long as I can remember I've been on UBR4-enfi.blueyonder in all whois queries and that continued when I went onto 50 megs a few months back but I had disappointing speeds. Now, for some reason and without me contacting Virgin, for the first time I seem to have changed UBR?? cpc2-***cust**.hari.cable.virginmedia.com ?? Hari for Haringay? I was always on the enfield server?

Anyone else had this happen, and could this explain my recent speed increase? Speeds are now stable around 50 meg!!

[graf_von_anonym]

Virgin will move you from one UBR to another as load and requirements change. A UBR is only one of the two* elements that provides your broadband connection, the other is what's called the CMTS. The latter is where the "Cable Modem Terminat(es)", the former is where your bandwidth and services are provided through the actual physical cable infrastructure.

If memory serves, both "hari" (Haringey) and "enfi" (Enfield) are at the same "OSR", which is gobbledegook for "place where UBRs route through". In some circumstances they are physically colocated with their daughter UBRs, in others there's some distance. All that's happened here is you've been tuned from one bit of the network to another, so while it's the same cabinet and the street and all you're on a different channel (well, four, in the case of 50Mb).

So yes, loads of folk have it happen, few notice anything about it, and it's almost certainly why your speeds have improved/increased. Enjoy it while it lasts.**

* I know that there are more than two. You know that there are more than two. For the purposes of this explanation, however, we will assume that the atom is made of electrons and nuclei, rather than probability shells and congregated exotic matter.

** You do realise that every time you admit that you are happy with Virgin's services a mark is made beside your name in Branson's Book of Dooms, and when sufficient tallies have been made therein you will be plagued by the Nine Daemons Of Poor Service, yea, so it is written.

[Ignitionnet]

Hi Graf,

Ex-TW area customers are being moved to the ex-ntl naming convention, to go along with the core network. The ex-ntl naming conventions have two site names in them in some cases, one indicates the actual CMTS and the other the regional head end that used to contain the DHCP server.

So in my case my hubside where my CMTS is would be Mortlake and there I hit my local OSR then I hurtle off to the regional HE which is Croydon and the OSR there that connects to the VM core, giving me this:

cpc1-mort6-2-0-custxx.croy.cable.virginmedia.com

DHCP scope 1, Mortlake CMTS 6, Customer xx in the scope, Croydon RHE, on VM cable. Identifying detail censored to save a passing mod doing it just incase

In Enfield there are OSRs in the hubsite, same as Mortlake, and it's fed by the OSRs in Haringey. I think every site, or pretty much every site has its' own OSRs if this little hubsite which only had 2 CMTS before Virgin Media's upgrades put 4 more in had them.

[graf_von_anonym]

I forget that because while I'm on broo I'm also on renf, but I tend to only see broo. Flashes of RuneQuest there. Anyway, yes, you're almost certainly right that it's the naming convention switch, but 50Mb customers should be getting moved onto the newest equipment as a matter of course, which would also usually result in the UBR changing. Unless, of course, 4 was 50Mb uprated, but I thought it was standard practise to provide 50 on new hardware alone, since it was basically the front for the capacity upgrades.

[Ignitionnet]

I *think* the odd 10k that was being used for capacity relief pre-50Mbps had the external QAM and GigE SPA fitted to supply 50Mbit. Unsure if this ws one of them but I'd speculate it was a naming convention change then work was done to resolve issues and at that time the naming convention change was noticed

Yar Brookfield? hubsite, Renfrew PoP, cam, core and DHCP server in your case, the renf bit coming from the DHCP server your IP is/was served from.

[on in an hour!]

wooowwhhh to that conversation!!! i couldnt agree more!!

[graf_von_anonym]

Broomhill, but aye. It can be hard to remember which aspects of the various naming conventions are useful and which are now just historical artifact.

I am always forgetting that there were 10Ks deployed before 50Mb rollout, because I've tended to lump them together with the Motorola BSRs generationally. That, and what appears to have been the "rolling over" of any pre-50Mb capacity improvements into the 50Mb rollout plan meant a pretty big gap in the introduction of new hardware. Even before chucking in Virgin's usual commitment to consistency, of course.

Erm. Yes. That's enough spod talk I think.

[Ignitionnet]

The historical artifact now is just the numbers after the CMTS in the name. mort6-2-0 means Mortlake CMTS 6, the 2-0 isn't really relevant.

Yes it was quite clever of them, VXR is discontinued, needing to do capacity upgrades so having to run with next-gen kit, so why not do 50Mbit deployment at the same time, relatively little cost on top and oooohhhh those headlines It did entail a bit of a delay as you said. Putting in new kit where needed and doing a comprehensive 2 new CMTS per hubsite minimum overlay do take slightly different lengths of time so some people were a tad delayed. Hopefully with the 50Mbit done people are being resegmented onto the newer kit now for capacity relief.