cancel
Showing results for 
Search instead for 
Did you mean: 

FAO: CRT -- ECI Modem Latency Problem with G.INP

AndyH
Grafter
Posts: 6,824
Thanks: 1
Registered: ‎27-10-2012

Re: FAO: CRT -- ECI Modem Latency Problem with G.INP

No ISP has commented or reported what has been published by Openreach on retransmission since the 'fix' was applied, but this is not unexpected. Even the likes of A&A and Zen have been quiet, despite there being a couple of brief updates as to the fix and the number of lines that have been updated so far.
There are many technical details from briefings that ISPs could share, but they chose not to. A small number of people on here will be unhappy about that, but ISPs act in the interest of the majority of their customers.
I don't think PN can extract any information, other than running the GEA line test. If reported the line profile setting conflicts with what the modem is saying, then by all means post away so that PN can report this to their suppliers. This forum will be the best way of reporting this information as there are some technical staff on here who deal direct with Openreach and BTw on things like this.
alext05
Grafter
Posts: 162
Registered: ‎16-12-2013

Re: FAO: CRT -- ECI Modem Latency Problem with G.INP

Thanks a lot for, Andy, your clarification. Just one a question about your recommendation for approaching your ISP in the first instance.
What would you suggest that the users who have lost G.INP on upstream with compatible equipment would need to say to their ISP in order to get it fixed, apart from the fact that they know it should be working and perhaps linking to this thread.
Also, do you personally know of anyone who has retained G.INP after the fix or had it fixed? Thanks.
AndyH
Grafter
Posts: 6,824
Thanks: 1
Registered: ‎27-10-2012

Re: FAO: CRT -- ECI Modem Latency Problem with G.INP

I am pretty sure there are examples on here of people who have retained the upstream retransmission.
Those users who believe or know they have lost upstream retransmission (when they think it should not have been removed) need to post on here so Plusnet can forward it to their suppliers for investigation. If there is a conflict in what the GEA test shows and what the modem is saying, then Openreach need to be made aware so they can investigate.
Not every line will have retransmission applied - it's only those that DLM are applying it to. What we don't know is what the thresholds/criteria for retransmission are (something which I think is due to be discussed in due course though).
The other thing to point out is that not all lines need retransmission. Those which are stable and of high quality, should not need it. If you're a gamer, it can be counter productive in terms of latency when packets are retransmitted.
alext05
Grafter
Posts: 162
Registered: ‎16-12-2013

Re: FAO: CRT -- ECI Modem Latency Problem with G.INP

Quote from: AndyH
If you're a gamer, it can be counter productive in terms of latency when packets are retransmitted.

Thanks. Wow. That is interesting. If it indeed applied to, let's say my own line, would I be able to ask my ISP to get it disabled or they would have no control of that?
Quote from: AndyH
Contrary to what you, and some other people might think, the technical engineers/developers within Openreach aren't a bunch of clueless idiots. Many have extensive experience and knowledge of telecommunications/broadband engineering and know much more than any forum enthusiasts will ever know.

I am not doubting that since they are producing software of such importance, but to be fair, they made a right hash for users of incompatible equipment with initial release of G.INP.  Wink
deathtrap
Grafter
Posts: 1,064
Thanks: 4
Registered: ‎23-04-2013

Re: FAO: CRT -- ECI Modem Latency Problem with G.INP

Quote from: AndyH
I am pretty sure there are examples on here of people who have retained the upstream retransmission.
Those users who believe or know they have lost upstream retransmission (when they think it should not have been removed) need to post on here so Plusnet can forward it to their suppliers for investigation. If there is a conflict in what the GEA test shows and what the modem is saying, then Openreach need to be made aware so they can investigate.
Not every line will have retransmission applied - it's only those that DLM are applying it to.
The other thing to point out is that not all lines need retransmission. Those which are stable and of high quality, should not need it. If you're a gamer, it can be counter productive in terms of latency when packets are retransmitted.

Since they started rolling out what they are calling this fix , Quite a few of those who have made their stats publicly available the majority have G.inp applied regardless of how well their circuit was performing prior to that rollout back in March, even line with attainable rates is excess of 100,mbps and lots of spare SNR and had never had DLM intervene since their circuits where activated   still G.inp was enabled regardless And just as i cannot prove that g.inp was applied to all circuits on huawei cabs , you cannot prove that it was selective
So that's one myth that you have peddled on here  discredited with some actual fact not BT make believe
Secondly a lot of those circuits that had g.inp applied in both directions  have since the fix  was announced had g.inp removed from their upstream  which has been replaced with a fast path profile, which has a higher level of FEC applied so some are seeing a reduction in upstream sync or attainable rates, latency was unaffected
As for not everyone who had been g.inp'ed has had it removed from the upstream, well that is quite possibly due to the fact that they as yet haven't finished rolling it out to every Huawei cab So time will tell, As for customers reporting this to their ISP I wish them luck with that one,
As for G.inp being an issue for gamers I didn't notice anything different about it, as with most others who had G.inp bidirectionally applied it didn't increase latency, Interleave and its delay are a games enemy  as are poor peering links and network congestion (regardless of who's network it resides in )
chrcoluk
Grafter
Posts: 1,990
Thanks: 5
Registered: ‎11-12-2013

Re: FAO: CRT -- ECI Modem Latency Problem with G.INP

Quote from: AndyH
Contrary to what you, and some other people might think, the technical engineers/developers within Openreach aren't a bunch of clueless idiots. Many have extensive experience and knowledge of telecommunications/broadband engineering and know much more than any forum enthusiasts will ever know.
If you have specific examples of users that are not seeing DLM applying retransmission correctly, then you need to tell them to raise it through their ISP. There are appropriate channels that need to be followed; you cannot just approach Openreach and tell them they are giving out wrong or misleading information.

What a stupid response.
So you refusing to fast track feedback simply because you pulling the corporate line.
Do a bit of searching its not hard to find what is happening.
Here is the facts, despite what your beloved openreach tell you.
Users with g.inp compatible broadcom devices are having g.inp disabled on the upstream.
Yes openreach are incompetent even if you do not wish to acknowledge, what is amusing is a month ago you was telling me that g.inp is needed on the US because of unstable lines, yet as soon as openreach changed their stance and decided to remove it now you telling users they dont need it, so your view changes to match what openreach say, its as if you are their puppet.
Also I am still waiting to hear what your affiliation is with openreach, I dont ask ignition the same question as he has openly let people know he works in the isp industry, whilst you claim to not work in that industry yet have access to their NGA discussions and the like.
Strat
Community Veteran
Posts: 31,320
Thanks: 1,609
Fixes: 565
Registered: ‎14-04-2007

Re: FAO: CRT -- ECI Modem Latency Problem with G.INP

Mod Note
To no-one specifically...please ease back on the language as it's getting a bit personal.
Windows 10 Firefox 109.0 (64-bit)
To argue with someone who has renounced the use of reason is like administering medicine to the dead - Thomas Paine
AndyH
Grafter
Posts: 6,824
Thanks: 1
Registered: ‎27-10-2012

Re: FAO: CRT -- ECI Modem Latency Problem with G.INP

Why should I discuss my business dealings on here? I fail to understand the relevance of this and why you are so concerned about who I do business with and why. I am a Plusnet customer and that's the end of it.
You need to post specific examples to backup your claims. You cannot go to Openreach making wild claims with no evidence. Not only would it be unprofessional, you're going to end up with zero credibility and they won't invite you to any working groups.
alext05
Grafter
Posts: 162
Registered: ‎16-12-2013

Re: FAO: CRT -- ECI Modem Latency Problem with G.INP

I don't get why people need to be abusive to a user that, I personally believe, tries to help, provides information in good faith and often backs it up with a relevant explanation or even a screenshot. Now, what he says might be contrary to what others know or think and people would like to disagree -  and that's fine. But this can be done in manner that stimulates finding a common solution rather than actually be counter-productive.
What everyone really should be angry at is their own ISPs, who would have plenty of relevant information from Openreach but choose to withhold it for the reasons best known to themselves.
If Plusnet do not have a firm confirmation from Openreach on what happens to G.INP on upstream after the latest fix, they should chase it from them - as users are widely reporting that it's being removed -  and update us appropriately. If they do know - perhaps they should be brave enough to let us know the truth.
It is also possible that was the ultimate intention of Openreach and they still work to make it happen for compatible equipment, it's just wasn't the immediate priority as it was important to improve connections for millions of users affected with the initial release as soon as possible, otherwise it would have taken too long for the fix to come.
It's Plusnet that your have a contract with. Not Openreach or AndyH. The ball is in firmly in Plusnet's court on this one. They have (or should have) all the data. Jeez.
InterZoom
Rising Star
Posts: 226
Thanks: 25
Fixes: 1
Registered: ‎15-08-2014

Re: FAO: CRT -- ECI Modem Latency Problem with G.INP

My latest exciting profile is:
0.128M-80M Downstream, Retransmission Low - 0.128M-20M Upstream, Error Protection Off
As of yesterday midday, with the ECI modem.
Will be interesting to see whether it has changed with the Huawei.
No further response re. getting the line fault fixed, though.
(And upstream sync speed remains extremely low.)
---
Troubleshooting:
The Limitations of Traceroute & Ping
Latency: Connection "fast" but internet sluggish? Bufferbloat FAQ
Black Holes: Worth noting that the Plusnet Hub One router has an MTU of 1488 bytes.
deathtrap
Grafter
Posts: 1,064
Thanks: 4
Registered: ‎23-04-2013

Re: FAO: CRT -- ECI Modem Latency Problem with G.INP

@InterZoom If that is as a result of this so called fix being rolled out, then it will work in the same way on the Huawei HG612  and all other G.inp compatible devices ,
InterZoom
Rising Star
Posts: 226
Thanks: 25
Fixes: 1
Registered: ‎15-08-2014

Re: FAO: CRT -- ECI Modem Latency Problem with G.INP

The point is that the Huawei modem itself is reporting G.INP off on the upstream.
We'll see what the new DLM profile says.
The wider point is that I would really rather like my line fixed. Please. Openreach. (Who tell me by email that PlusNet need to deal with fixing Openreach balls-ups because that's who I pay.)
---
Troubleshooting:
The Limitations of Traceroute & Ping
Latency: Connection "fast" but internet sluggish? Bufferbloat FAQ
Black Holes: Worth noting that the Plusnet Hub One router has an MTU of 1488 bytes.
deathtrap
Grafter
Posts: 1,064
Thanks: 4
Registered: ‎23-04-2013

Re: FAO: CRT -- ECI Modem Latency Problem with G.INP

@  AndyH & alext05 Proof of the changes is available on line, and from a source independent from BT ect , and I have proof from the stats collected from even before G.inp was initially rolled out, and i'm reporting  from my own experience and from the evidence from that independent source  i have seen the changes so know 100% that BT have removed G.inp on the upstream and that is intended as that is what their fix is, so that ECI and other devices that are incompatible with G.inp work without adding latency due to DLM reverting to a fall back profile (interleaving on the upstream with associated error correction coupled with a loss of  upstream sync speed ) that is how the buffoons have fixed it, not
When we all know what they should of done was to recall all those G.inp incompatible devices that they BTOR and BT consumer supplied to EU's and replaced them with compatible devices
God only knows what mess will occur should they ever decide to roll out vectoring  and there is still the question regarding those ECI cabs are they G.inp compatible ? we know that they aren't vector compatible as they are , And Andy H still thinks they are intelligent ,
alext05
Grafter
Posts: 162
Registered: ‎16-12-2013

Re: FAO: CRT -- ECI Modem Latency Problem with G.INP

I am not disputing the proof you are referring to - I have no knowledge or qualifications to do that. I am just a curious user trying to find solutions to the issues and researching what is available to me, much more like yourself perhaps. I have also noticed the discrepancy between what Andy was saying and many other users, and took it into account.
What I am saying is that certain things often are not as clear cut as we would like them to be, and that's why the difference of opinion might occur. If you have any further probing questions to Andy that might help you to establish where he comes from, by all means keep asking. And if this is done in a constructive manner, I am sure that he would be more than willing to help us all to clear any misunderstandings we all might have.
I also do hope that someone from Plusnet will take an effort to tell us whether both sides are right to an extend or not, since clearly this issue is very important to some users.
Looking forward to my fix, by the way.  Smiley
30FTTC06
Pro
Posts: 2,286
Thanks: 108
Fixes: 4
Registered: ‎18-02-2013

Re: FAO: CRT -- ECI Modem Latency Problem with G.INP

@InterZoom Mine didn't show Retransmission enabled when I first had g.inp enabled either.
This snapshot is taken from my Pre trial g.inp enabled modem on the 27/3/15
# xdslcmd profile --show
Modulations:
G.Dmt Enabled
G.lite Disabled
T1.413 Disabled
ADSL2 Enabled
AnnexL Disabled
ADSL2+ Enabled
AnnexM Disabled
VDSL2 Enabled
VDSL2 profiles:
8a Enabled
8b Enabled
8c Enabled
8d Enabled
12a Enabled
12b Enabled
17a Enabled
30a Enabled
US0 Enabled
Phone line pair:
Inner pair
Capability:
bitswap On
sra On
trellis On
sesdrop Off
CoMinMgn Off
24k On
phyReXmt(Us/Ds) Off/On
TpsTc AvPvAa
monitorTone: On
dynamicD: On
dynamicF: Off
SOS: On
Training Margin(Q4 in dB): -1(DEFAULT)
#

Then on the 30/4/15  0.128M-44M Downstream, Retransmission Low - 0.128M-7.2M Upstream, Retransmission Low
I'm curious if that is still the same if anybody is able to have a look for me please, 'cus' the people behind the button are giving me some spurious random answers  Huh