Kitz Forum
Broadband Related => ADSL Issues => Topic started by: Weaver on April 23, 2019, 03:46:34 AM
-
I have noticed that my line #2 is also showing cyclic upstream SNRM variation, but a different waveform and a much smaller amplitude, only ~1dB high. Times are UTC, day is 2019-04-22
(https://i.ibb.co/H7RKfPy/849-EB5-E9-E87-F-41-BB-9-EB0-942-BD20092-E5.jpg)
Line #2, despite the number, is the most recently installed one, installed last summer.
* What on earth is the short duration high state period - ~5:05pm - 7:25pm UTC - all about ?
-
To my eyes the time period you mention is when the line is operating normally, at the 6dB target.
It's the 1dB drops between 11am - midnight that are when some form of interference is present.
What causes the 1dB drops in upstream is anyone's guess.
-
John - you make a very good point. I was thinking about the shape, it’s not just a simple high/low waveform as for line 3. So maybe some device got turned off for that period. Will see what happens on other days.
Right now line 2 is ‘low’ @ upstream SNRM of 5 dB, and attenuation is 41.2 dB, power is 12.4 dBm. So the question is: is this a variation of noise or attenuation or output power? I’ll check when the graph goes high and compare.
With line 3 I think the conclusion was that it has to be a line fault [?] (was that agreed?) because it is attenuation that goes up and down, not noise, so the question is whether or not this is similar.
-
With line 3 I think the conclusion was that it has to be a line fault [?] (was that agreed?) because it is attenuation that goes up and down, not noise, so the question is whether or not this is similar.
Without looking back through the earlier posts, I believe that was my conclusion . . . the attenuation of the physical metallic pathway should not change in such a fashion.
-
Even weirder waveform. This is not about a 24 hour period at all.
(https://i.ibb.co/44K45hz/D04315-A9-2-FB9-4793-9012-69-EA5-BCE38-A9.jpg)
-
This is my cordless phone charging it's battery left. Right not. DS SNR.
-
What on earth is the short duration high state period - ~5:05pm - 7:25pm UTC - all about
Both patterns look like two shifts going to work and turning something on.
At 7km your line could pass a lot of noisy things on way to exchange.
-
At 7km your line could pass a lot of noisy things on way to exchange.
Except for the majority of that length it is just moorland, overlying granite bedrock. ;)
-
Do the Belgians care?
-
This has possibly returned - it is a bit soon to tell. But there is a step 1.6 dB high showing in the most recent graph:
(https://i.ibb.co/7SB18JZ/4-A973-D78-DD41-4777-A2-CC-4-ED98-B8-B9-AF5.jpg)
See https://forum.kitz.co.uk/index.php/topic,23742.msg401882.html#msg401882
-
Line 2 - Tue 2019-09-10 to Wed 11. It seems that line 2 has now become the new line 3. Compared with how it was a month ago, now the waveform is regular, with a 24 hour period and 1.5dB height
(https://i.ibb.co/bJqcSNp/DE76-EB7-A-741-B-4-BC7-82-B4-7-A226-C8-BE27-A.jpg)
-
I don't think I can offer any insight into the step changes in SNRM you have, other than saying I have experienced similar in my downstream and have no explanation. Lights switching on? Some sort of construction tools being used?
But I do enjoy seeing graphs from my firmware! :D
V2 is coming, but I need everyone to be able to build it.
-
Your firmware has been absolutely invaluable during the month of hell recently. It has shown exactly how bad and how widespread the problems have been.
As for cutting out some candidate explanations: it only applies to one line and not to others; it’s upstream-only.
I need to look at the numerator and denominator of a ratio here, I think, and I have been lazy in not doing this earlier in this case. I need to look at the attenuation, power output and the noise level in the high and low parts of the square wave once more. If we assume for the moment that it is not the case that the output power of a transmitter is going up/down, then either the attenuation could be going up/down or the absolute noise level could be going up down. For the former, I would assume a change in resistance due to a ‘semi-contact’ make/break - I don’t have the words I need here. For the latter, interference would seem more likely, but then we are stuck with explaining how it can only be one line.
Since line 2 seems to have taken on the upstream-waveform of line 3, while line 3’s upstream has been cured, I do wonder if someone possibly swapped E-sides when they fixed the recent month of hell. Would that be a possible candidate explanation?
-
Since line 2 seems to have taken on the upstream-waveform of line 3, while line 3’s upstream has been cured, I do wonder if someone possibly swapped E-sides when they fixed the recent month of hell. Would that be a possible candidate explanation?
We don't know if you now have four lines, divided into D-sides & E-sides, or just one line with a D-side & an E-side plus three EO lines. However we do have a strong hint that the fourth, latest, line you had installed has "taken" an E-side from one of the two cabinets (on the A87, at Harrapul (I hope I have spelt the name correctly)) for its deployment.
However pair swapping does seem to be indicated by your latest observations.
-
That’s a good point. Line 2 is the latest, and has an E-side and indeed we don’t know about the others, I shouldn’t have made that mistake.
Yes, it could be taken as evidence for pair-swapping; line 2 became like the new line 3, but also more significantly, line 3 changed; its troublesome behaviour went away. So that suggests line 3 also has an E-side now, no?
-
So that suggests line 3 also has an E-side now, no?
I wouldn't go as far as making that assumption. The disruption could be coupled into an E-side, D-side or an EO line. Once we reach the T-junction with the A87, there are various premises that could be the disrupter's source.
At the moment, without sight of the up-to-date network records, we are just guessing as to how the infrastructure merges into the background. For example, you have mentioned where it is possible to see one (or more) cables lying beside the road on the way to Heasta. I would be interested to know where it/they go(es) underground, before continuing westbound along the A87.
-
Indeed, I have probably assumed too much. It was a convenient way of explaining the disappearance of the nuisance line 3 upstream square wave.
The copper must be in ducts [?] somewhere between the Claymore restaurant and the Broadford exchange along the main road. Then further east there are plenty of FTTC cabs now in Sculamus and Breacais according to that codelook website [?] where there are a large number of houses stretching half a mile or so further east and in the direction of the Skye bridge.
On the high moor to Heasta there is still only the one bundle visible by the roadside, according to Janet. Near the main road, in the suburb of Harapul, where the Heasta road leaves the main road at the Claymore restaurant (and the local vet’s) there are about eight or nine houses right by the Heasta road, between the main road and a cattle grid where the road goes into to open pastures. All visible in Google Street View of course. I have no idea where exactly the copper is that services those houses and continues on towards Heasta. I suspect the copper is buried in a very shallow inadequate way on that stretch from the northern cattle grid where the houses peter out down southwards to the sharp little kink in the road where the bundle crosses the stream and starts to climb towards Heasta. There has been cable damage in the past in those fields, where animals are fed and vehicles pull off the road at the Harapul fank, where the ground is soft. It’s only at that bridge where you can start to see the copper, so in terms of topology, I’m wondering if the only significant node is that PCP discussed earlier, by the Claymore restaurant. And that spot, near the south roadside green cab, you’ll recall, is near where Janet saw workmen active around the time when things first went to hell last month. They were not BT, but builders with a mini digger, working on a driveway.
-
. . . in terms of topology, I’m wondering if the only significant node is that PCP discussed earlier, by the Claymore restaurant.
That aligns with my suspicion.
And that spot, near the south roadside green cab, you’ll recall, is near where Janet saw workmen active around the time when things first went to hell last month. They were not BT, but builders with a mini digger, working on a driveway.
Ah, yes. That's the location I can't pin-point (by searching for the corresponding background features in the current, old, Google Maps Street View image).
-
And now it seems things have gone back to the way they were.
* Line 3 has reassumed its old square wave behaviour. :-( Large (~4dB) height. I have set the line 3 upstream target to 9dB again in order to guard against a drop from 6dB all the way down to 2dB, which actually happened earlier today.
* Here’s line 2 for 2019-09-17 … 18 :
(https://i.ibb.co/86XWp43/8-D90-A2-ED-2-DF2-4-AA3-996-C-215-ADD3468-E3.jpg)
Combined upstream throughput reported by speedtesters has maybe improved slightly, again this is odd because the sync rates of line 3 has worsened slightly following a retrain this morning.
Live sync rates:
#1: down 2913 kbps, up 532 kbps
#2: down 2762 kbps, up 512 kbps
#3: down 2927 kbps, up 272 kbps
#4: down 2850 kbps, up 496 kbps
Firebrick current upstream rate limiters' IP PDU tx rates (egress speeds), in-force right now ::
#1: 454050 bps
#2: 436981 bps
#3: 232146 bps
#4: 423325 bps
Total combined rate: 1.546502 Mbps
Fractional upstream speed contributions:
#1: 29.360% [███████████████ ‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒]
#2: 28.256% [██████████████ ‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒]
#3: 15.011% [████████ ‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒]
#4: 27.373% [██████████████ ‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒‒]
-
I have to admit that I am totalled baffled by these current happenings. ???
-
I am baffled too. That would now seem to have killed the theory that an E-side swap was carried out; cannot have had two E-side swaps, with a more recent one in the last couple of days; surely no way?
One thing that I need to recheck: is the SNRM variation due to attenuation variation, which is itself due to variation of resistance not variation in tx power output? Or is the SNRM due to changes in the noise level, be it crosstalk or other unrelated random interference - sources of interference going on / off?
Can check that by looking at the detailed stats for attenuation and power output.
-
I think all modems only update the attenuation figure when the line trains. Therefore the only way to tell if the attenuation has changed would be to trigger a retrain.
-
I had not thought about that. Thanks for letting me know! ;D That explains a few things in the past. :-[ I’ll make sure to take account of that then.
I’ve just blown it for today, forgot and now I’ve missed my chance so next opportunity for a partner snapshot will have to be tomorrow now.
-
Line 2 continues its bizarre upstream waggling, same 1.5dB height and it has not taken on the role of "the new line 3" at the moment
(https://i.ibb.co/7gL6Y2j/826-D3-BB3-99-E5-4-BC2-9-C73-0-E19-FCF4-C842.jpg)