Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Ragnarok]

I'd say it's more likely, the modem looks like it's behaving better. Some commented that having seen RETX enable=1 and

Considering the issues most vendors are having with lantiq hardware and making it fully compatible with G.inp, it's highly unlikely even in co-operation with lantiq BT got this update right first time, the update may be an old one too which would make it even more likely there are RETX incompatibilities.

I find this whole G.inp situation bizarre, it sounds like it's the norm for many other networks around the world. We know BT like to be slow and get it right, heck they even set stringent standards that modems/routers had to meet, but still ended up having big issues with lantiq hardware. I'm amazed the ECI modem issues never came up in testing, and it never crossed their minds that Lantic/ECI modems might be used on Huawei cabs and tested anyway. I would of thought they would of also thoroughly tested ECI cabinets before deploying them too.

There are some round here who seem to have a much better grasp of the issues than BT's PR department anyway, BT's also huge so it's probably a bureaucratic nightmare to get anything done or rolled out, I can imagine many of there technicians/engineers know exactly what's happening and are just as frustrated as some of us are, especially if we're all waiting on Lantiq for a fix.

[kitz]

For all intents and purposes it looks like their testing was silently roll it out to a few cabs in and around the Martlesham Heath area.   The testing was also done before any of the changes they made to the DLM late last year after the ASSIA case.  Its also possible if those cabs were the early enabled bunch (ie pre2013) that modems may have been matched to cabs.   I am surprised they didnt pick up some sort of inkling that something was wrong though.

>> it sounds like it's the norm for many other networks around the world.

In the other main thread I posted a link to another ISP, but it was in German and even with the aid of google translate I couldnt get a proper understanding what was going on, but it does appear they may have also had a similar issue with the lantiq chipsets earlier this year.

[Mark07]

Spotted an OR engineer in my ECI fibre cab (not the PCP, the actual fibre cab) on my way to work this morning, as I'm aware for customer connections etc they don't go into the fibre cabs themselves, so could it be they were installing some new hardware in there? 

[loonylion]

Spotted an OR engineer in my ECI fibre cab (not the PCP, the actual fibre cab) on my way to work this morning, as I'm aware for customer connections etc they don't go into the fibre cabs themselves, so could it be they were installing some new hardware in there? 

probably a heating pad

[Mark07]

Yeah that's what I was thinking, something totally unrelated and uninteresting  no change to my stats either

[burakkucat]

probably a heating pad

Or removing the Mossad installed "back-door"? 

[Mark07]

Or the cat6 I installed 

[themanstan]

https://community.bt.com/t5/BT-Infinity-Speed-Connection/G-INP-affecting-certain-eci-modems-resulting-in-a-drop-in-sync/m-p/1487001#M178832

3rd post from bottom on p37. appears a fix is coming out...

[kitz]

Thanks for that ^, but I think what SeanD actually meant is that its for those on the Huawei cabs.

Thank you for your continued patience.  As you know Openreach recently carried out a trial to address the reports on this thread with regards to G.INP.  Openreach are very happy with the results of the trial and after some consultation they have now decided to roll this fix out across the network.

The trials were only on Huawei cabs and theres been no definite news yet about the ECI cabs other than a review after the Huawei's have all been updated.  Im guessing here but I should think that will be late July before we hear anything more definite about the ECI's.

As regards to the 'fix' it may not be all wonderful for the ECI modems.  Interleaving appears to have been turned off, but from what Ive seen so far Error correction is still being applied which means no recovery of the loss of speed.    However we are waiting for the Huawei rollouts and then some kind soul such as BaldEagle will be able to test properly with stats to see whats happening.    Its not something I can do myself as Im on an ECI.

[Chrysalis]

kitz I post my 3 day montage for your viewing pleasure, cannot find our most recent posts, but this thread seems appropriate.

Since that resync event I have higher bitswapping and higher ES on my upstream, the ES rate on the DS is ok, but I have had 3 SES.

[kitz]

Im not sure what to make of it.   Your power is different, so that would account for the slightly less max attainable upstream,  and could cause more bitswap and errors, so I cant really draw anything conclusive to say a prep for g.inp (other than downtime).   It's not unusual for power to change upon a resync as the bitswap process can eventually cause power output to go up.. which then goes back down again at a fresh resync.
 
afaik the g.inp software config changes at the dslam should make any difference to your line conditions and its only when stage two is applied that you start seeing its effects.

[simoncraddock]

Your line is behaving in a similar fashion to how mine was Friday morning after a DLM resync around 3am. Afterwards I had steady stream of both CRC and ES errors from the moment it resynchronised which lasted around 30hrs before it lost connection twice in quick succession mid Saturday morning applying quite heavy interleaving.

At first I put it down to my newly installed Solar Panels but over the past few days I'm not so sure.

At the moment errors are really low but until interleaving is removed I can't really say if it's stabilised. Data as been uploading to MDWS for almost 48hrs so I can look for any pattern of errors that can possibly be explained.

[NewtronStar]

Well kitz if ECI cabs are to get G.INP the stage 1 should already be on the way to update cabinets and some of the Huawei users did notice strange going on's before the start of stage 2.

[Semmy]

Spotted an OR engineer in my ECI fibre cab (not the PCP, the actual fibre cab) on my way to work this morning, as I'm aware for customer connections etc they don't go into the fibre cabs themselves, so could it be they were installing some new hardware in there? 

Could always be an incremental tie job.

[Mark07]

Could have been anything I guess, but nothing has changed on my line