Frequent bouts of extremely high error rates - noise margin steady

[#1]

Any ideas on what could be causing this?
The error rates have been relatively high for a few weeks on my line, and for periods of time will increase significantly (to the point of almost 100% packet loss). Yet, noise margin remains steady.
Bitswapping is not occurring as frequently as I'd seen in the past (even though it is enabled and still bitswapping very infrequently).

[#2]

Sounds like some issues ive been having, ive ordered a new router just incase... i hope it is the router otherwise ill be out of pocket

[#3]

I seem to be having the same problem, very high error rates yet noise margins seem fairly steady.
I'm finding i have to reboot the router to get web pages to load faster, it slows down to a crawl.
ukste20,  I'm using the same router as you  TP-Link8960N, let me know how you get on with the new router
Fifer

[#4]

I'm using a Netgear DG834G. I tried my Draytek Vigor 2850Vn's internal modem and it wouldn't even sync at all (the Netgear eventually connects (slowly) after taking several attempts at negotiating -> connecting -> negotiating -> connecting).

[#5]

CX, can you post the ADSL stats from the modem/router.
Fifer, sounds very similar to ukste20's problem.

[#6]

(I think that the "Fast" can be ignored - it's Interleaved)

Status: ShowtimeRetrain Reason:	20
Channel: FAST, Upstream rate = 444 Kbps, Downstream rate = 1895 Kbps
Link Power State: L0
Mode:			ADSL2
Channel:		Fast
Trellis:		U:ON /D:ON
Line Status:		No Defect
Training Status:	Showtime
		Down		Up
SNR (dB):	7.3		14.6
Attn(dB):	67.0		40.0
Pwr(dBm):	16.4		12.0
Max(Kbps):	2464		17547
Rate (Kbps):	1895		444
			G.dmt framing
K:		60(0)		19
R:		10		10
S:		2		2
D:		16		8
			ADSL2 framing
MSGc:		59		24
B:		59		18
M:		2		2
T:		1		2
R:		10		10
S:		1.9923		2.6666
L:		522		144
D:		16		8
			Counters
SF:		2484118		2010653
SFErr(CRCErr):	3804		19404
RS:		80733858		190070
RSCorr:		27816678		133982
RSUnCorr:	5220		0
HEC:		1670		7436
OCD:		4		0
LCD:		0		0
Total Cells:	179747311		3434690937
Data Cells:	152736205		398036970
Drop Cells:	0
Bit Errors:	0		342851
ES:		82506		906
SES:		14		3427
UAS:		308		858
AS(Uptime):	40213
INP:		1.22		2.22
PER:		16.18		20.00
delay:		7.96		5.33
OR:		32.12		12.00
Bitswap:	567		41
Total uptime:	220897

Total time = 1 days 13 hours 28 min 30 sec
SF  = 13710832
CRC = 180837
LOS = 0
LOF = 0
ES  = 82506
Latest 1 day time = 13 hours 28 min 30 sec
SF  = 2988964
CRC = 35573
LOS = 0
LOF = 0
ES  = 10785
Latest 15 minutes time = 13 min 30 sec
SF  = 50002
CRC = 79
LOS = 0
LOF = 0
ES  = 71
Previous 15 minutes time = 15 min 0 sec
SF  = 55578
CRC = 131
LOS = 0
LOF = 0
ES  = 118
Previous 1 day time = 24 hours 0 sec
SF  = 5360978
CRC = 122147
LOS = 0
LOF = 0
ES  = 56981
15 minutes interval [-30 min to -15 min] time = 15 min 0 sec
SF  = 55641
CRC = 105
LOS = 0
LOF = 0
ES  = 96
15 minutes interval [-45 min to -30 min] time = 15 min 0 sec
SF  = 55578
CRC = 120
LOS = 0
LOF = 0
ES  = 107
15 minutes interval [-60 min to -45 min] time = 15 min 0 sec
SF  = 55579
CRC = 158
LOS = 0
LOF = 0
ES  = 143
Showtime Drop Reason:	20
Last Retrain Reason:	20

[#7]

Hi CX,
Just wanted to post one of those connection graphs, there have been one or two disconnections showing recently:
<img src="http://ccgi.psmith12.plus.com/visradius/generated/image13512488849802.png" />
I suspect that's more a symptom of whatever's causing the interference though, if you've already confirmed it's not the internal wiring or anything plugged into it then it'd probably be worth raising a fault if only to see the results of the diagnostics.
Don't suppose you're hearing any noise on the phone line at all?

[#8]

Hi,
Thanks for the connection graph. I think there may be more actual loss of sync events than that, since PPP usually survives a resync.
We don't use the landline much these days, and the quiet line test doesn't highlight anything (although the fact it hangs up after a few minutes, as well as the "quiet line test" voice limits long-term testing). On a couple of occasions there has been a loud click during a call, but it could well have been coming from the "other end" of the call.
We're currently using a Pressac SSFP (FTTC style) at the master socket. Problems were occurring with the previous ADSLNation filter too. From the SSFP the DSL runs along a second pair in the cable up to a dedicated socket. The problems persist when plugging directly in the front of the SSFP and also in the test socket behind.
I will raise a fault.
Fault raised.

[#9]

Thanks for that, hope we manage to get it sorted for you - our faults team will update you with the test information as soon as is possible.

[#10]

The errors seem to be increasing again.  Checked this morning and there was 118 TOTAL SES now almost 8000.
The router was rebooted at approx  2.00 pm yesterday so its been up for 26 hrs,  seems like an awful lot of errors.  I don't think there has been any loss of connection but i notice the SNR Margin has dropped from 3.9 to 2.4.
Just recently been moved over to ADSL2+ and since then the error rates have shot up. I used to get occasional bursts of noise causing errors on ADSL1 but not like this.
Here are some router stats, they don't look good to me
Mode: ADSL_2plus
Traffic Type: ATM
Status: Up
Link Power State: L0

Downstream Upstream
Line Coding(Trellis): On On
SNR Margin (0.1 dB): 24 57
Attenuation (0.1 dB): 255 102
Output Power (0.1 dBm): 0 128
Attainable Rate (Kbps): 18748 1204

Path 0 Path 1
Downstream Upstream Downstream Upstream
Rate (Kbps): 16044 1207 0 0

MSGc (# of bytes in overhead channel message): 63 10 0 0
B (# of bytes in Mux Data Frame): 156 27 0 0
M (# of Mux Data Frames in FEC Data Frame): 1 8 0 0
T (Mux Data Frames over sync bytes): 3 6 0 0
R (# of check bytes in FEC Data Frame): 12 4 0 0
S (ratio of FEC over PMD Data Frame length): 0.3124 5.9029 0.0 0.0
L (# of bits in PMD Data Frame): 4327 309 0 0
D (interleaver depth): 96 4 0 0
Delay (msec): 7.49 5.90 0.0 0.0
INP (DMT symbol): 1.6 0.20 0.0 0.0

Super Frames: 5742041 505727 0 0
Super Frame Errors: 0 0 0 0
RS Words: 1188602568 2723818 0 0
RS Correctable Errors: 88466838 4294967237 0 0
RS Uncorrectable Errors: 11663287 0 0 0

HEC Errors: 4294967286 418876 0 0
OCD Errors: 37850 0 0 0
LCD Errors: 36 0 0 0
Total Cells: 3494751875 2407238 0 0
Data Cells: 2909302 13192 0 0
Bit Errors: 0 0 0 0

Total ES: 7804 0
Total SES: 7374 0
Total UAS: 32 4294967268
Fifer
P.S  there are now over 11000 Total SES  1 Hour later

[#11]

Things appear to have settled at around 1950kbps. Obviously a huge step down from the 2400kbps it was working at before but at least the error rates are remaining low, although it's still dropping upon incoming calls.
I'm running a Python script to grab the sync rate, overall SNR, per-tone SNR, errored seconds and CRC via telnet and write it to a database so I can track any trends with time. All figures appear steady right now but I wish I'd started doing this sooner so I could compare to before.

[#12]

If you paste in the results of the telnet command 'adsl info --QLN', I can generate a graph for you to show where the interference may be coming from, e.g. radio stations, etc.

[#13]

@CX,
if your sync is dropping on incoming calls, you probably have a line fault, but you should try another filter if you haven't already.
Can you hear any crackling or other noises when using the phone or do you have any other problems with incoming or outgoing calls?
@Fifer,
I' d try another modem/router if you have a spare modem/router or can borrow one. Not only is that error rate somewhat high, but the Interleave depth is showing as 96!!

[#14]

Quote from: Anotherone

if your sync is dropping on incoming calls, you probably have a line fault, but you should try another filter if you haven't already.

It drops even when the router is plugged directly into the test socket, no filter.

[#15]

Hi CX,
Your fault is being investigated and our faults team will update your ticket later today with their findings. If there's anything else in the meantime then please let me knoe.
Chris