cancel
Showing results for 
Search instead for 
Did you mean: 

Could this be the answer to the PPPoE anomaly?

nozzer
Hero
Posts: 3,298
Thanks: 676
Fixes: 3
Registered: ‎04-08-2009

Could this be the answer to the PPPoE anomaly?

I've just got to ask this question...  To Jameseh, BP, MT and anyone else who has been involved with the dropping PPP nuisance when using PPPoA, is what follows relevant? My exchange has been converted to 21CN, and although I know I am not actually using the 21CN network officially (I am on IPStream Max) I think this might be too much of a coincidence to ignore, especially as I have been solidly connected via PPPoE for a week now (because PPPoA was so unreliable, and won't now connect at all).
Take a look at this article (also supported by BTW SIN docs) http://www.samknows.com/broadband/21cn_overview.php. Scroll down to the part that deals with "High level 21CN network structure". Under there you will find this statement.
21CN involves the construction of an entirely new hierarchical network. Note that the network runs over Ethernet, rather than ATM (as in the current Colossus network). We'll start from the bottom level (the 5600 exchanges) and work our way up to the core of the new BT network.
Further down is the statement.. 
ATM is largely being replaced by Ethernet in 21CN. That said, there will still be traces of it since PPPoA (PPP over ATM) will still be used for handling DSL connections between end user premises and the MSAN in the exchange.
Might that explain why not everyone is being affected by the PPP drop problem? It could be that the apparent incompatibility of Broadcom chipsets with something BT have done on the backhaul is actually not what is seems to be, but rather the fact that PPPoE is now mandatory?
Just thought I would throw that one into the ring just in case!
17 REPLIES 17
James
Grafter
Posts: 21,036
Thanks: 5
Registered: ‎04-04-2007

Re: Could this be the answer to the PPPoE anomaly?

Nope on both counts!
There is still an element to ATM.  However, both encapsulation modes (as obviously you know!) work on ADSL2+.
It shouldn't have anything to do with the PPP drops being reported by some, but does explain why PPPoE works Smiley
nozzer
Hero
Posts: 3,298
Thanks: 676
Fixes: 3
Registered: ‎04-08-2009

Re: Could this be the answer to the PPPoE anomaly?

Yeah, spose so!
Has anything useful materialised from the testing you have been doing so far?
James
Grafter
Posts: 21,036
Thanks: 5
Registered: ‎04-04-2007

Re: Could this be the answer to the PPPoE anomaly?

No Sad
I have extremely limited examples, especially now Bill's (directFS) problem doesn't appear to be PPP related.
In other news though, BT are trying to fix the problems that have been reported with the speed being considerable below IP Profile.
nozzer
Hero
Posts: 3,298
Thanks: 676
Fixes: 3
Registered: ‎04-08-2009

Re: Could this be the answer to the PPPoE anomaly?

Is BT aware that there have been problems like I have had though? LS has indicated in my ticket (still ongoing) that you are carrying out testing and intending to liaise with BT. BP obviously has first hand experience of the problem (and MT is also aware). I assume you'll be able to decipher the initials.
Replies #14 and 15 in  http://community.plus.net/forum/index.php/topic,78341.0.html
jim:quote
James
Grafter
Posts: 21,036
Thanks: 5
Registered: ‎04-04-2007

Re: Could this be the answer to the PPPoE anomaly?

Yep - I'm fully aware, but the problem does not appear to be widespread.  We've had probably a handful of examples in here, and it does appear to be related to the chipset on the modem/router.
The more widescale problems are currently being address (exchange capacity on 21C and hopefully subsequently banded profiles).  for the number of examples we have on the PPP issues you've been reporting, we have to follow the normal faults process and escalate where required.
I know it's far from ideal!  But I'll do whatever I can to escalate your problem.
nozzer
Hero
Posts: 3,298
Thanks: 676
Fixes: 3
Registered: ‎04-08-2009

Re: Could this be the answer to the PPPoE anomaly?

OK thanks...    like I said before though, my connection on PPPoE is the best I've ever had with PN, so tread softly!   Wink
jim:quote
Oldjim
Resting Legend
Posts: 38,460
Thanks: 787
Fixes: 63
Registered: ‎15-06-2007

Re: Could this be the answer to the PPPoE anomaly?

James,
With respect to the router chipset link have you managed to match the problem to a specific chipset in the exchange
DMT used to be able to identify that with the Speedtouch Routers with the Broadcom chip but the information doesn't seem to be readily available with the Netgear and I don't know about the 585v7
James
Grafter
Posts: 21,036
Thanks: 5
Registered: ‎04-04-2007

Re: Could this be the answer to the PPPoE anomaly?

Do my best.
I've asked LS to get a copy of the engineers notes and to escalate it with the SFI Interventions Team (read as mroe helpful BT faults people).
I'll let you know how he gets on.  Be good to have it working well on PPPoATM too!
tobykim
Grafter
Posts: 105
Registered: ‎25-02-2009

Re: Could this be the answer to the PPPoE anomaly?

If it's any help to you my router was dropping I had a netgear DG834gV4 all the lights would be lit but I couldn't get on the web, funnily enough it only affected the PC connected via ethernet the wireless one was fine, I've swapped to the Speedtouch 585V7 and no drops at all, I believe they both have broadcom chipsets although not the same one I think the one in the speedtouch is 6338 and the netgear one is higher, my exchange hasn't been upgraded and I couldn't navigate to the router page
James
Grafter
Posts: 21,036
Thanks: 5
Registered: ‎04-04-2007

Re: Could this be the answer to the PPPoE anomaly?

Thanks.
for what's it's worth, I did look at all those who appeared to be affected by the PPP issue, but I wasn't able to see any similarities.  All on different exchanges, some upgraded, some not, etc.
nozzer
Hero
Posts: 3,298
Thanks: 676
Fixes: 3
Registered: ‎04-08-2009

Re: Could this be the answer to the PPPoE anomaly?

Quote from: Oldjim
James,
With respect to the router chipset link have you managed to match the problem to a specific chipset in the exchange
DMT used to be able to identify that with the Speedtouch Routers with the Broadcom chip but the information doesn't seem to be readily available with the Netgear and I don't know about the 585v7

I'm running DMT with a Voyager 2110 Jim, how do you get it to identify the DSLAM chipset? ie what's the telnet command (if there is one).
grahamt
Rising Star
Posts: 599
Thanks: 37
Registered: ‎05-04-2008

Re: Could this be the answer to the PPPoE anomaly?

I'm curious about how nozzer's connection manages to work over PPPoE. A brief bit of experimenting shows that if I switch my router to PPPoE the ADSL line itself stays up but the PPPoE status is given as 'N/A' (whereas with PPPoA it says 'Up' when all's well) and my WAN IP address is given as 0.0.0.0. Clearly, in this state I can't connect to anywhere.
I didn't expect PPPoE to work, given that PPPoA is used from the exchange to my router. So under what conditions does it work?
I'm using a ZyXEL router with a TI AR7 chipset.
Graham
nozzer
Hero
Posts: 3,298
Thanks: 676
Fixes: 3
Registered: ‎04-08-2009

Re: Could this be the answer to the PPPoE anomaly?

Don't ask!  Grin  It's a long story. Used to be on PPPoA on the Tiscali backhaul, now that won't work, only PPPoE on BT IPSMax.
grahamt
Rising Star
Posts: 599
Thanks: 37
Registered: ‎05-04-2008

Re: Could this be the answer to the PPPoE anomaly?

I take it all back! PPPoE *does* work for me.
When I first tried it, I hadn't also switched from VC to LLC encapsulation. After I saw nozzer mention LLC in another thread I tried again and it worked fine.
I don't expect it will change any of the other 'features' of my line but I'll leave it like this for a bit and see what happens.
Graham