Kitz Forum
Broadband Related => FTTC and FTTP Issues => Topic started by: skyeci on April 27, 2017, 11:49:20 AM
-
Hi,
Any ideas what leads to large drops in snr... I have had 3 recent resyncs within the last 24 hours. At one point the DS snr dropped to 0.8. I have asked my ISP (sky) to do some tests and not surprisingly they said all was normal...
Thoughts about what can cause the large snr drop appreciated.
thanks
-
This is rather like what I was experiencing for some 4 months last summer with TT and an ECI cabinet, you can see some wild swings on MDWS between May and September 2016. TT always said they cold find nothing wrong when they tested. It was never fixed - it just disappeared despite 3 BTOR visits.
Stuart
-
Thanks - I recall all the trouble you had. I have have had a lengthy open case with sky which eventually resulted in rein engineers visting . They have advised there is rein present but unable to resolve even though its affecting their network. The reason for my question was even though the line is affected by rein supposedly it normally doesnt suffer with all these resyncs which seems to be a new thing now..
-
is mdws down right now? I get a blank white page.
Since I cannot load it can you confirm your US margin is going down to close to 2db snrm?
-
From the posted graph, I assumed the vertical axis was 0, 4, 8, 12, 16, but the "1" digit had been cropped off.
-
sorry - this might be a bit better. Couple of snapshots around the resyncs, Since the 4 resyncs with overall loss of 4mb on the ds sync but attainable still below sync with snr under the target I am not seeing any further resyncs but I am seeing regular SES. At the time of the resyncs the 24 hour period was 535 ses or something along that lines.
It just seems to be getting worse but OR seem unable to help any further. Sky never seem to detect any faults either. They have said they won't pursue things until the sync goes below 60.5 - currently around 67 sync with attainable lower than sync which has always been the problem.
Thanks
-
looks a mess, sadly tho the way this service is supported is very simplistic, they dont really care about snrm that fluctuates, it is why there is a snr margin so fluctuations can occur.
All they really care about is 2 factors, line stability and line performance. By stability I mean outages or packet loss not a flat lined SNRM.
So unless your line drops out excessively or the sync speed falls below the figure given to you by sky there is not much you can do.
-
here's hoping it will just give up eventually then something might be done perhaps...
thanks for taking a look
ses 550
-
I am wondering if there is a developing high resistance joint in the circuit.
Have you recently tried a Quiet Line Test and a Faultsman's Ring-Back, via 17070, whilst watching what DSLstats reports?
-
Hi bc
Quiet line test says it can hear echo but not noise. I notice the ds snr falls a few points (e.g 5.8 to 5.5)when making a call out- is this normal?
I told the rein engineer about the echo on the test but was not concerned with the result.
qln around the worst of it and hlog around that time of the 3 resyncs I think possibly after, not sure tbh
thanks
-
I notice the ds snr falls a few points (e.g 5.8 to 5.5)when making a call out- is this normal?
There will be extra loading applied to the circuit when a telephone has "looped the line" to originate or answer a call, so I would not be too concerned with an observed drop in SNRM of, say, 0.2 - 0.3 dB.
qln around the worst of it and hlog around that time of the 3 resyncs I think possibly after, not sure tbh
Both the QLN and the Hlog plots that you have shown do not give me any cause for concern (although you have only shown a section of each). A G.993.2 (VDSL2), Profile 17a, circuit has 4096 sub-carriers available (numbered from 0 to 4095).
If the circuit was being troubled by a HR (or semiconducting) joint, it would most likely become visible in the SNRM plot -- coinciding with usage of the telephone.
-
Ok cheers.
Will see how things progress in time..
-
I am wondering if there is a developing high resistance joint in the circuit.
Have you recently tried a Quiet Line Test and a Faultsman's Ring-Back, via 17070, whilst watching what DSLstats reports?
More trouble this evening.. DS SNR just hit 0.0! - bands look a mess. 140+ SES..
wish I could get this sorted.. - any further thoughts?
cheers
[attachment deleted by admin]
-
More trouble this evening.. DS SNR just hit 0.0! - bands look a mess. 140+ SES..
wish I could get this sorted.. - any further thoughts?
There is definitely a fault in the metallic pathway but as to the nature of the fault, I am not too sure. Initially I thought there might be a HR or semi-conducting joint. However the plots you have shown do not "fit the pattern".
I wonder if WWWombat has noticed anything? :-\
-
resync at -2.0 on the DS... so gradual loss from 73mb sync only a few months ago.
seems to have a mind of its own..
all over the place now sync wise..
-
are you manually resyncing it when its low to recover the speed? if yes dont do that, let it stay low so they see it on their screens.
-
I haven't touched it mate - all the resyncs are happening on its own. I am not messing with it in anyway..
4 automatic re-syncs now
-
I agree at a potential HR fault; some signs from the variation in SNR.
But it's not quite right yet. Perhaps not often enough, yet.
MDWS takes snapshots of Hlog, QLN, SNR/tone ... but only at hourly intervals, on the hour. And the SNR here just happens to be "normal" at those snapshots.
Keep monitoring?
-
Sky rang diagnostics.. no fault found as per usual. They did however accept my graphs from dsl stats and have told me they will open it again with openreach. Need to wait now for further feedback...how long who knows..
-
There is definitely a fault in the metallic pathway but as to the nature of the fault, I am not too sure. Initially I thought there might be a HR or semi-conducting joint. However the plots you have shown do not "fit the pattern".
Hi bc
Line looks a mess again. Regular ses and loss of ds snr. Around 5pm yesterday continual levels of crc's have appeared from practically nothing.. Would this have any significance on the potential cause of the problem?
Many thanks
Average error rates for 25 May 2017
CRC erors per hour: 9178 Down, 0.14 Up
FEC erors per hour: 0 Down, 2.00 Up
HEC erors per hour: 4617 Down, 0 Up
ES per hour: 40.4 Down, 0.14 Up
SES per hour: 15.6 Down, 0 Up
-
I think you got a new crosstalker as snrm and attainable has dropped :( the lower snr margin may be the cause of the increased errors.
-
to be fair Chrys my attainable has always been under sync right back from day one... Sky have run tests today and confirmed they can see all the errors etc.. - they have no explanation and suggested I wait till it falls below if it ever go below the 60.4mb guaranteed sync..They did offer an engineer but I see no point as I have had several plus REIN engineers that resulted in the issue not being fixed..
I just wish it would break once and for all and perhaps something might get sorted even if it meant no service for sometime whilst investigated again. :(
-
Check it, you will see a sudden step down recently. 5.57 pm may 24th
you was between 5.9 and 6.1db, after that time its now between 5.1 and 5.3 (excluding spikes).
Also you wont get this fixed without an engineer.
to clarify the new crosstalker is not the cause of your original fault but now the snrm has weakened it has made your symptons worse, thats what I meant.
-
I'll agree that the evidence points towards a new crosstalker . . . However the circuit defect, giving rise to SES, ES, HEC and CRC, still remains an unknown. :(
-
Yeah SES is rising at a steady even rate, very bizarre, it be interesting to see if this stops if he resyncs again at 6db (with lower sync speed).
-
Wont be messing with it tbh..Just going to leave it and see what prevails....
-
I'll agree that the evidence points towards a new crosstalker . . . However the circuit defect, giving rise to SES, ES, HEC and CRC, still remains an unknown. :(
Like magic.... the crc's and ses have appeared to clear again as stopped at 9pm ish.. Snr back up for now. Either cross talker disappeared or a or the source of noise/fault?? has relented again till the next time.... ???
-
I can only assume the storms last night may be the cause of this. Hopefully dlm will leave me alone :fingers: -ouch!
-
That's nasty. :(
Just hoping that the DLM will recognise it as a wide area event and decline to interfere. :fingers:
-
Lets hope. Might just get away with it this time..
-
Well another bunch of continued ses and us loss of snr. 5mb lost on DS sync as DLM stepped in..
Sky say I am now under the guaranteed speed and as this is another speed drop in the last 15 days they will now approve an engineer.....
Lets see what prevails this time...
-
well could do with some advice please. After a this period of persistent SES and CRC's they have both stopped again. This seems to be a frequent event lately. I have an engineer for tuesday but is there any point proceeding with it if the fault/issue has temporarily cleared again? - I don't want the usual no fault found which has happened a number of times in the past...
given the past issues since early april just interested what you do in the same position? Stick with appointment and show the evidence gathered or cancel the appointment until the next time hoping it will be worse. Because the snr has risen again its given me the overhead back but I am tempted to leave it and see what happens rather than force it back up....
many thanks - help greatly appreciated...
-
I really wish I could suggest something, but in a lot of respects your line shows several similarities to mine especially if you compare say over a 60 or 90 day period. Its almost like there's an emerging HR fault that never fully materialises, but at other times it can at times look like REIN.
Like you I often get the SNRm oscillations. Like you I can suddenly lose quite a lot of sync speed and it will fall below the BTw recommended speed.
Like you I get the bursts of CRC/ErrSecs/SES that can last for several hours and will cause DLM to kick in. Its been doing this on and off for over 9 months.
On the other hand its now in a period of stability and performing quite nicely despite being below the Target SNRM.
All I can say is from observation sometimes it seems to correlate to a disturber line doing a resync, although it can happen after a remote resync.. or another trigger can be that everyday without fail I get a spike in CRCs which will usually happen between 10-11am.
Performing a resync will quite often stop the errors dead in their tracks, but by the time I notice its usually too late and DLM takes action.
-
Thanks Kitz
How much sync have you lost do you think? I am 10mb plus from last november on the DS...
-
One other item to add to the list of similarities is that both skyeci and kitz have G.993.2 metallic pathways that terminate on ECI Hi-FOCuS M41 Mini-Shelf MSANs.
-
Thanks Kitz
How much sync have you lost do you think? I am 10mb plus from last november on the DS...
Not quite sure.... from memory prior to this happening (Oct 2016) my attainable was circa 84Mb. The lowest Ive been is 67Mbps.
My BTw estimate was 80 - 70.6 Mbps, so Ive fallen below it on several occasions. I get the burst errors which look very similar to yours and then DLM takes action. Then its fine for a while. Last week there was some sort of remote sync and I came back up before my disturber which is why I'm now sat at 78Mbps, but usually its around 72Mbps.
One other item to add to the list of similarities is that both skyeci and kitz have G.993.2 metallic pathways that terminate on ECI Hi-FOCuS M41 Mini-Shelf MSANs.
Yup indeed and it is something Ive said before that there's a couple of us on MDWS that could be getting similar symptoms.. all on ECI's :(
Whatever it is, it started for me after what I presume was some downtime at the cab/exchange in about Oct last year. Unlikely to be a crosstalker coming onboard at 4am causing an outage of about 20mins. :(
-
Yup I lost complete service the other week for 2 hours. Phone ok but no bb. Checked neighbour but seemed ok.
Last speed drop was 15 days or so. Considering the erratic behaviour of snr and errors it seens unlikely to be a cross talker but I am not sure I will ever find out..
-
engineer arrived yesterday 13/06/2107 and carried out tests - no fault found as a number of visits in the past have also found. I showed him the graphs from past events in april/may & june. He agreed something is obviously not right but said he can do nothing about it.
Sadly today DLM lowering speed again by another 4mb on the DS, removing my fastpath which I have had for many months and applying interleaving & delay to both up and down. I now hold the highest level of interleaving on mydslwebstats :no:
Sky have issued deadlock and said I can leave penalty free if I wish - not that it will help as we know its the same bit of copper. They say due to the factors outside of their control they are unable to help any further.
looks like upstream was the cause of the issue today, dropping to 0db...
I wonder if G.INP should it finally come back to ECI will help this situation at all?
-
My interleaving was over 1200 as well, I thought I had my DLM manually reset this morning following a conversation with an openreach bod, but turns out it was automated, so you could if line is stable be back to fast path in 2 days.
Of course with an ongoing fault it doesnt mean a lot. As you could just as easily be back to interleaving 2 days after.
You should have reported right when your sync was really low, a clear no brainer fault in such a situation.
not too happy with sky at the moment tho, as well as the weird congestion behaviour in the mornings, I now have these pings on fast path, it used to be 7-8ms.
C:\Windows\system32>ping -t -4 bbc.co.uk
Pinging bbc.co.uk [212.58.246.78] with 32 bytes of data:
Reply from 212.58.246.78: bytes=32 time=14ms TTL=57
Reply from 212.58.246.78: bytes=32 time=13ms TTL=57
Reply from 212.58.246.78: bytes=32 time=14ms TTL=57
ipv6 is a bit lower but only 12ms.
tbb is a bit dodgy on single threaded ipv4 tests but ipv6 is flatlined max. I am wondering with so much of sky's traffic now going over ipv6 (think its more than half) they may have dialled back their ipv4 peering/transit capacity.
a single threaded ipv4 test of one of my servers is fine tho.
-
My interleaving was over 1200 as well, I thought I had my DLM manually reset this morning following a conversation with an openreach bod, but turns out it was automated, so you could if line is stable be back to fast path in 2 days.
Of course with an ongoing fault it doesnt mean a lot. As you could just as easily be back to interleaving 2 days after.
You should have reported right when your sync was really low, a clear no brainer fault in such a situation.
not too happy with sky at the moment tho, as well as the weird congestion behaviour in the mornings, I now have these pings on fast path, it used to be 7-8ms.
ipv6 is a bit lower but only 12ms.
tbb is a bit dodgy on single threaded ipv4 tests but ipv6 is flatlined max.
a single threaded ipv4 test of one of my servers is fine tho.
I did report it when the sync was at 57mb but due to changes again in the snr the tests and engineer was all to late..
Seems I am stuck with it..
-
lol... after putting up with my eci cab causing all sorts of grief over the last couple of years and one neighbour in particular who knocks off 6mb on my attainable and sync everytime the modem is left on I took Sky's suggestion of moving seriously ! ;D .. providing all goes well I hope it will feel like winning the broadband lottery- new place has FTTP pre installed and ready to go.. :clap:
hopefully in around 6 weeks I can stop moaning about my eci cab forever more. I sure won't miss pcp31 and it's associated
:shoot: eci cab one bit!....