ECI Cabinets

[#1]

Since my set up changed over 17 days ago, it has been rock steady and has not dropped the connection at all according to my router stats or that I have seen.

I used to get 4~5 ES errors every hour prior to G.INP being applied, but now average 2 ES per day - Download speed up to 64.2 Mbps from 61.4 Mbps and consistent.

Well impressed with the results, and hope the trial is a success.  😀

VMT to @bobpullen for his assistance, and work, with regard to this trial for us involved.

[#2]

@ryant704 looks like interleaving has been re-applied and your line stability assessment isn't great, suggesting your line is occasionally dropping out.

[#3]

@bobpullen

I posted a few weeks ago and from the responses I got my line appeared to be banded, I tried a reboot last week but it just dropped my sync rate lower (see attached).

I did ask CS if they would ask to see if the DLM could be reset but again it seems nothing has happen.

I am getting better ping times so at least latency has improved but currently I am 2Mbs below where I was.

If anything can be done I would be grateful.

[#4]

Update, Three Weeks in and still holding no change from when G.INP was added, will continue to monitor and update the thread as and when needed.

[#5]

Rock solid for me still. 

Does anyone remember the old trial and the first initial results so we can compare it to how this trial is going? 

I assume the old trial was dismissed as participants had a lot of instability. So far I can’t fault it. Low pings, speed gains, solid SNR, no fluctuations in anything I can monitor. 

I’d say good news for a wider roll out unless the engineers find anything in the data. 

[#6]

Connection has been steady since having g.inp applied although I have done a few reboots since. One thing I'd like to mention though is I'm getting UDP packet loss spikes every few seconds but I had this before g.inp was applied too. Running a UDP test using StarTrinitty Continuous Speed Test I get intermittent loss on the download and after a couple of minutes frequent loss on the upload as seen here.

EDIT: Can't seem to zoom into the first image so here's a direct link https://i.postimg.cc/38Ww6Xnd/packetloss.png

With SQM QOS enabled in OpenWRT to combat bufferbloat (capped at 2.5/10) There is a lot less packet loss but still frequent small spikes as seen here

I can see this loss also in in-game networking stats such as in cs:go and pubg and can also here small robotic breaks in voip in games and other applications. Can also see the loss visually with laggy players/disappearing actions and events when i get the spikes.

[#7]

16 days into the trials, I think.
Slight fluctuations in the maximum rate and the SNR. Looks like auto readjustments.
No other changes. Line still rock solid.

[#8]

All good on my end.
Using it for video conference calls and WiFi calling for work all day with the kids playing online and streaming on the TVs and FaceTiming there friends and hasn’t missed a beat and the line speed is holding up with no resync’s.

[#9]

Had a reset last night about 2am back to where I was before G.INP not sure why as everything was stable and looked good Has anyone else had a reset ?, @bobpullen  Has G.INP been disabled on my end for some reason ? 

[#10]

Same results as last week :

”Slight fluctuations in the maximum rate and the SNR. Looks like auto readjustments.
No other changes. Line still rock solid.”

No resets, G.INP still enabled. Current stats can be seen here 

[#11]

Still no change from the reset, If any staff member is around and free can they please check to see if G.INP has been disabled I'm guessing it has but the Hub-One doesn't give any stats to truly determine that.

[#12]

---

@Terranova667 wrote:

Had a reset last night about 2am back to where I was before G.INP not sure why as everything was stable and looked good Has anyone else had a reset ?, @bobpullen  Has G.INP been disabled on my end for some reason ? 

---

I'll have somebody take a look. Bit early for me to glean anything due to the delay in the service layer data being fed through.

I can see a bunch of downstream errors in the days leading up to it though which might explain things.

[#13]

Thanks for that Bob, errors would explain it, would be nice to know just how bad things are, but as the Hub-One is so limited in not providing anything for errors and other stats it's very hard to see what is happening.

A case of wait and see if DLM re-applies G.INP.. that's if it is removed and not something else, which at this stage is a guess based on the sync being back to where it was Prior to it being applied.

[#14]

No problems here, noise margins remain stable, speeds are consistent, no dlm resets…

Data rate:9.995 Mbps / 39.998 Mbps

Maximum data rate:15.920 Mbps / 61.985 Mbps

Noise margin:10.1 / 12.8

Line attenuation:14.5 / 13.5

Signal attenuation:14.4 / 13.5

[#15]

I’ve had a resync this morning and lost about 1 meg.
My noise margin has been sitting at 5db for about 5 days and now back to 6db.
Just think of all that crosstalk in the cabs at the moment where everyone is at home working and playing all day.