Print

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

[Chrysalis]

yeah I just posted my observations, I never said it was a fault (in another thread, no post here)

More so it was a possible factor that could affect stability and in turn DLM behaviour.

As far as his engineer goes I dont really have much to say, he has ordered one with the sole intention of getting a DLM reset which the engineer may not take kindly to, I think the good news tho is that its not a SFI but a boost engineer instead, so the boost engineer is there with the aim of boosting speed, and a DLM reset will likely be part of that task.

[William Grimsley]

Yes, I likely won't let him leave without a DLM reset, that's his sole purpose as you say.

[licquorice]

It will be very interesting to see if a straightforward engineer instigated DLM reset is any different to the the DLM reset performed when William upgraded and does indeed remove the banding, or whether a 'special' reset is required to remove banding.

[William Grimsley]

We shall have to see, I'll go with the latter seen as kitz has mentioned about this "capped C profile" issue I may be experiencing.

kitz, would you like me to create one thread for everything from now on so everything isn't getting confused?

[kitz]

Oh, god now spiking on Downstream SNR Margin, not a crosstalker fault... :/ I have someone next to me on FTTC who has 50% of the speed I have so they could have a fault now...

Breathe.   Ive just had a look.  It looks like a crosstalker spike.  Dont panic.

Im trying to go over some points on your bullet list and Im having to look through some of your early posts.
Unfortunately I have a class at 10am, and Im not even dressed yet.. so Im going to have to leave this until later

[William Grimsley]

I'm perfectly calm, sorry it's not coming over like that, I come over a bit angry when it's not intended, just loving the banter here!

Ok, now BT are saying the engineer is coming in the afternoon, thankfully I don't have much on but I've messaged back saying that's incorrect.

Ah, off we go. They say I was "wrongly advised" it was tomorrow morning. High level complaints are going to call me back shortly.

[kitz]

OK its taken me a while because Ive had to go searching through your old posts for some info that I needed.

I suggest you change your bullet point list slightly to

• 30/01/2016 - BT Home Hub 5 started rebooting in excess of 20 times per day,
• 01/02/2016 - DLM Intervened and banded the line at 35000 kbps
• 02/02/2016 - Bought a new Billion BiPAC 8800NL router and since that date the line has been completely stable running with surplus SNR Margin (9.7dB)
• 2 months have passed and DLM has still banded my line and has kept the Downstream Rate at around 35 Mbps even though the Downstream Attainable Rate has sat around 43 – 46 Mbps continually.
• The 55/10 package came available so thought upgrading to that would cause a DLM reset.
• 18/04/2016 - 55/10 product upgrade complete. However the line is still banded and interleaving and FEC overheads has followed causing the Downstream Attainable Rate to drop further.
• The line is stable, experiences very low error rates,  line banding still remains in place.

[William Grimsley]

Thank you, kitz.

Do you mind if I use your precise words for tomorrow?

[kitz]

I also suggest that you print out the following on a separate piece of paper to show the engineer just so that he has a record of your stats

30/01/16 - Line stats during the period of 20+ reboots per day

   Data Rate:   8099 / 40000
   Maximum Data Rate:   8099 / 43854
   Noise Margin:   6.3 / 7.3


01/02/16 - Line stats after DLM Intervened

     Data Rate:   8447 / 35000
     Maximum Data Rate:   9038 / 44340
     Noise Margin:   6.3 / 9.6


02/02/16 - Line stats with new router

              Downstream   Upstream
   Line attenuation (dB):     25.8      0.0
   Signal attenuation (dB):   Not monitored     
   Connection speed (kbps):   35000      8493
   SNR margin (dB):           9.7      6.7
   Power (dBm):               12.0      5.8
   Interleave depth:          8      2
   INP:                       54.00      53.00
   G.INP:                     Enabled
   Attainable Rate (Kbps)     44479   8897     

   RSCorr/RS (%):             0.0001      0.0010
   RSUnCorr/RS (%):           0.0000      0.0000
   ES/hour:                   0      0



18/04/2016 - Full Line stats immediately after product upgrade. 
   Note: Cap still in place. G.INP removed and Interleaving and Forward Error Correction applied.  Overheads from FEC redundancy reduce line rate.

xDSL
Mode   VDSL2
Traffic Type   PTM
Status   Up
Link Power State   L0
Downstream   Upstream
Line Coding (Trellis)   On   On
SNR Margin (dB)   6.6   6.3
Attenuation (dB)   25.8   0.0
Output Power (dBm)   12.0   5.8
Attainable Rate (Kbps)   45458   8015
Rate (Kbps)   34999   8015
B (# of bytes in Mux Data Frame)   47   239
M (# of Mux Data Frames in an RS codeword)   1   1
T (# of Mux Data Frames in an OH sub-frame)   64   42
R (# of redundancy bytes in the RS codeword)   14   0
S (# of data symbols over which the RS code word spans)   0.0436   0.9514
L (# of bits transmitted in each data symbol)   11368   2018
D (interleaver depth)   743   1
I (interleaver block size in bytes)   62   120
N (RS codeword size)   62   240
Delay (msec)   8   0
INP (DMT symbol)   3.50   0.00
OH Frames   1157914   321353
OH Frame Errors   2   1
RS Words   296335661   767079
RS Correctable Errors   5147   0
RS Uncorrectable Errors   21   0
HEC Errors   6   0
OCD Errors   0   0
LCD Errors   0   0
Total Cells   218471761   0
Data Cells   4946678   0
Bit Errors   0   0
Total ES   1   1
Total SES   0   0
Total UAS   30   30

[William Grimsley]

I'm simply overwhelmed with how much work you've put in. I'll send you a PM.

[kitz]

No I dont mind at all, hopefully it should get all the info across you need to the engineer.

Good luck for tomorrow 

[kitz]

Finally to finish off.

I already mentioned yesterday that I was aware that Williams upstream was showing some slight instability.   I said at the time that I suspected there was an amount of what could be normalish SNRm variance of about 1.5dB, plus something similar to REIN that happened during the hours of 1pm and 4:30 pm.

Having looked at Williams stats today, I am still of the opinion that Williams line looks like it has a daily swing of ~ 1.5dB.
It does now indeed look like yesterdays 1 - 4:30 incident could have been REIN.   Yesterday was a beautiful afternoon and it could have been something like a neighbour with power tools in the garden  (A couple of years ago I saw similar when my neighbour was jet washing his drive).

Whilst this still leaves a daily swing, but to be perfectly honest this is going to be exceedingly difficult for Openreach to track down and it will just be written off as performing within normal parameters.   There could be a very early HR fault, but again until it gets MUCH worse, Openreach wont really do anything until it actually impacts voice.    I know from personal experience if they cant find a 10dB HR spiking until it impacted voice, I doubt they will do anything for 1.5dB when the line is not experiencing excessive errors.

I feel that to give William the best chance of getting the DLM reset, then for now we should concentrate on the downstream issue.
If the upstream variance gets a lot worse or if it starts affecting voice, then the situation should be reviewed at that time.

[William Grimsley]

Brilliant post, +1.