Kitz Forum
Broadband Related => FTTC and FTTP Issues => Topic started by: Jon21 on November 19, 2019, 01:18:44 PM
-
In the early hours of Friday morning, my connection dropped and came back up with interleaving applied. I think it was because there was a large burst of ES on Thursday. Not sure what caused it, I've gone back to using a Draytek Vigor 130. I've also replaced the DSL cable from the master socket to modem. The RJ11 end was looking a bit suspect, sheathing had broken at the crimp on the plug. Ordered another one from Run-IT-Direct and the build quality seems to be improved compared to the old one. Was just wondering how long it takes to go back to Fastpath?
Stats from Draytek:
---------------------- ATU-R Info (hw: annex A, f/w: annex A/B/C) -----------
Running Mode : 17A State : SHOWTIME
DS Actual Rate : 59824000 bps US Actual Rate : 14116000 bps
DS Attainable Rate : 69388504 bps US Attainable Rate : 14121272 bps
DS Path Mode : Interleave US Path Mode : Fast
DS Interleave Depth : 872 US Interleave Depth : 1
NE Current Attenuation : 20 dB Cur SNR Margin : 6 dB
DS actual PSD : 6. 4 dB US actual PSD : 13. 0 dB
NE CRC Count : 0 FE CRC Count : 5073
NE ES Count : 0 FE ES Count : 4040
Xdsl Reset Times : 0 Xdsl Link Times : 1
ITU Version[0] : fe004452 ITU Version[1] : 41590000
VDSL Firmware Version : 05-07-06-0D-01-07 [with Vectoring support]
Power Management Mode : DSL_G997_PMS_L0
Test Mode : DISABLE
-------------------------------- ATU-C Info ---------------------------------
Far Current Attenuation : 26 dB Far SNR Margin : 6 dB
CO ITU Version[0] : b5004946 CO ITU Version[1] : 544ed086
DSLAM CHIPSET VENDOR : < IFTN >
-
It all depends.
My connection has had interleaving applied for the past 14d after a burst of ES (400 or so) in a couple of hours period.
Way below the 2440 max allowed in the day but must have hit some unknown limit perhaps per hour.
I have had a maximum of 5 ES/hr since normally 0/1 with 10 CRC/hr.
No idea why it is being reluctant to remove this time in the past it used to take 3 or 5 days maximum but the last couple of times it has happened (storms) it was 10d and 12d iirc.
I'm also on an ECI cab which isn't great.
I may try switching off for 30m or so and see if a resync will wake up the DLM, 14d is getting ridiculous.
-
Not wishing to upset anyone but I've been interleaved (and capped at 60mbps) now for nearly 4 years on an ECI cabinet, even BTOR have no idea why!
Stuart
-
All you can now do is make sure your VDSL2 connection stays stable for next 2 weeks with out any forced resyncs and very few errored seconds per day like under 10 good luck.
-
Ideally we need to see ES rates over a period of time.
To ktz392837 I think 5 ES/hour "whilst" interleaved is not particularly low. Bear in mind the green threshold is probably very low. Green on Interleaving will be way way lower than green on fast path.
Also the recovery period increases if you have more incidents, I dont know if it ever reverts back to a 2 day recovery e.g. if you have no incident for a year, but the recovery time will increase if it has repeated incidents.
-
My 15d since interleaving applied is averaging
Down ES 0.33/hr
Up ES 1.22/hr
There has been a down high of 9 ES and upload high of 25 ES in a hour period iirc.
It is very annoying it is not getting removed I wish we had some hard numbers on the limits.
Hopefully one morning I will wake up and DLM has reset to normal overnight :)
-
Some more stats. Not quite a days worth. Seems to be about 1 Down ES a day and 9 Up ES a day, at the moment.
[ Counters: Showtime ]
Near End Far End Note
ElapsedTime : 67888 67888 (seconds)
CRC : 5 5084
FEC : 40243 66160
HEC : 0 0
CRC_P : 1 0
CRCP_P : 0 0
CVP_P : 0 0
NCD : 0 0
LCD : 0 0
ES : 1 4049
SES : 0 3
LOSS : 0 0
UAS : 0 0
LOFS : 0 0
[ Counters: 15Min ]
Near End Far End Note
ElapsedTime : 449 449 (seconds)
CRC : 0 0
FEC : 21 2
HEC : 0 0
CRC_P : 0 0
CRCP_P : 0 0
CVP_P : 0 0
NCD : 0 0
LCD : 0 0
ES : 0 0
SES : 0 0
LOSS : 0 0
UAS : 0 0
LOFS : 0 0
[ Counters: 1DAY ]
Near End Far End Note
ElapsedTime : 77849 77849 (seconds)
CRC : 5 5084
FEC : 40243 66160
HEC : 0 0
CRC_P : 1 0
CRCP_P : 0 0
CVP_P : 0 0
NCD : 0 0
LCD : 0 0
ES : 1 4049
SES : 0 3
LOSS : 0 0
UAS : 0 0
LOFS : 0 0
-
Yay back to fast path, sync speed returned, interleaving removed.
What solved it for me:
1. Posting to Kitz forums
2. Unplugging from line for 6+ hours
3. Absolutely pure coincidence the DLM does as it likes you have no control over it
Number 3 it is ;)
Let's hope the nearly 2000 upload ES reported by DSLstats in the hour it resynced is just a stats glitch and not actually ES otherwise DLM will no doubt intervene tonight!
-
To give you hope when i first flashed to newer firmware, I had the same ES spike reported, the ES were not showing in the telnet stats (just dslstats). and my DLM was not affected. IF it shows in your live telnet stats I would perhaps be a bit concerned, but fingers crossed.
-
Don't want start new thread so will ask here as I recently back on fastpatch (again) also on ECI cabinet.
What is acceptable amount of errors per day, CR, HEC, ES, only these errors appear on my line. I'm on ee broadband so not sure about profile, where even ee seems to don't know that.
-
I think its 2880 for speed profile, and 1440 for standard profile, these I think are right from memory, the numbers are on the kitz wiki as well. This is for ES, I dont think they bother with CRC or HEC.
-
In that case I may remove cap completely over the weekend, at the moment I capped my line to 65Mb/s from available 76Mb/s and getting just over 100 ES per day.
-
Yay back to fast path, sync speed returned, interleaving removed.
What solved it for me:
1. Posting to Kitz forums
2. Unplugging from line for 6+ hours
3. Absolutely pure coincidence the DLM does as it likes you have no control over it
Number 3 it is ;)
Let's hope the nearly 2000 upload ES reported by DSLstats in the hour it resynced is just a stats glitch and not actually ES otherwise DLM will no doubt intervene tonight!
So roughly ~15/16days to go back to Fastpath. DLM hasn't relented on my connection yet, just a waiting game I guess.
-
I think is all depend from conditions, I was waiting over two weeks with both upstream and downstream capped and where on upstream fastpatch kicked after three days on downstream didn't for over 15 days. I decided to remove cap from upstream as higher snr due to cap cause strange high snr on u0 band (if I good remember). After 3 days with cap from 69 to 58 on downstream only, fastpatch kicked in on downstream. I personally think that time doesn't matter, I have like 300k fec errors in over 15 days (no other errirs at all) and dlm didn't give me fastpatch, so there must be other conditions like snr on all bands which tell dlm if something else may be wrong with line, just basically on my own experience.
-
I always used to go off the 2880/day rule but from my experience over the past three months or so I feel that interleaving is taking longer to be removed.
It used to take 2 or 3 days but the last few times for me it has been approximately 12d, 10d and 16d.
If you get a massive burst of ES (eg 500) the DLM can still kick in overnight even if you are no where near 2880 for the rest of the day.
-
Since I reworked my extension with proper cat5e cable I don't get errors in burst of hundreds or thousands , 1-2/15min sometimes 0 and then in other 15min 1-2 ES.
-
I always used to go off the 2880/day rule but from my experience over the past three months or so I feel that interleaving is taking longer to be removed.
It used to take 2 or 3 days but the last few times for me it has been approximately 12d, 10d and 16d.
If you get a massive burst of ES (eg 500) the DLM can still kick in overnight even if you are no where near 2880 for the rest of the day.
Those are the red thresholds for increased DLM stability, not thresholds for it to be removed (green).
-
I'm not sure what DLM is playing at right now.
This spike got me interleaved on the 16th of Oct. Total Err secs that day was just 450. I had no disconnects and the modem had been up for over 2 months.
Its still not been removed over 5 wks later, despite an average of just 10 Err secs per day.
-
you think its possible kitz the data we have has now been obsoleted?
That seems crazy for that day, you basically just had an extra 20 ES or there about's.
-
I'm glad it is not just me that is/was having problems as at least it maybe a global change and not specific to me.
Kitz I wasn't kidding when I said I left mine disconnected for 6hrs (afternoon) I wonder if that actually did something and kicked it into action. It was weird that it was removed that night.
-
you think its possible kitz the data we have has now been obsoleted?
Even if it has changed, that is a ludicrously low figure - Way too low to be workable for most lines.
A spike of 25 Err Secs should not trigger DLM - Thats if it was attributable to the spike at all. My total Err Secs for that day AND whole proceeding week only came to 2096
That seems crazy for that day, you basically just had an extra 20 ES or there about's.
Yup. 25 to be precise.
Nor should it be still stuck nrly 6 wks later. More like something is a bit broken.
I'm glad it is not just me that is/was having problems as at least it maybe a global change and not specific to me.
I don't know if its related or not but I've had 2 small outages within the past few weeks both at around 1am where there was no backhaul connectivity.
An outage at that time of night is usually Openreach messing with something.
-
I don't know if its related or not but I've had 2 small outages within the past few weeks both at around 1am where there was no backhaul connectivity.
Correction 4 of them. I hadn't noticed one yesterday morning at 1:30 plus another. The two I had noticed were only because I was watching Netflix and the stream came to a halt.
Anyhow I caved in and thought I'd try my luck contacting PN just before they closed to ask for a DLM reset.
Call was answered on first ring and CS rep seemed entirely competent and knowledgeable - we were chatting whilst he ran the required fault tests before a reset can be requested. He agreed that the line was stable with low error rate for past few months and he could see no visible reason why it had been applied and that in theory it should have been removed a few weeks ago.
-----
I know I could have capped and am aware that a reset will probably leave me with default interleave for a short while... but I am hoping that a reset will clear any accumulating data where the DLM gets stricter for each subsequent event.
-
Cut the long story I decided to rework my extension cable as I have aluminium installation on extensions in my house (house 2 years old) , after rework was done modem connected running constantly for over 14 days no any errors at all apart of 100k fec per day on downstream upstream maybe 3k. After 14 days decided to give a breath for dlm, modem disconnected for 9h from morning to late afternoon, turned back on for over 10 days, same figures, no any errors apart of FEC. Decided to cap line once capped fast patch kicks in on upstream after 3 days, for another 15 das with cap on booth upstream and downstream no luck with fastpatch for download. However I noticed that snr on U0 band go up like hell comparing to other bands so I decided to remove cap from upstream and keep it only on downstream to drop snr on U0 band, once I did that after 3 dsys fastpatch kicked in on downstream as well. Yesterday I removed cap completely and amount of errors looks sensible, just over 130 CRC, over 320 HEC, just over 100 ES. Hopefully DLM will leave me alone for now once I improved my line.
-
The recovery following a reset is still about 2 days at least so that's something, I was back on fast path on 3rd day after I changed isp.
I wonder now as well with the default interleaving, if the first move to fast path counts as the "first" recovery meaning the first time you get DLM'd it counts as a second event?
-
The reset occurred this morning.
Straight to INP 0
>> first time you get DLM'd it counts as a second event?
Would have thought it would count the events from the ILQ records as per usual.
A trigger event would be ILQ red or ILQ crimson.
----
and just as I hit send, my phone advised me that the daily morning spike is still there. ~2k CRCs in one minute but Err Secs is 107 so should hopefully be ok.
-
Interleaving was removed just before 4am this morning. (Ignore any red spikes, that's me messing with the router). About 9-10 days to go from Interleaved back to Fastpath.
(https://www.thinkbroadband.com/broadband/monitoring/quality/share/thumb/3346ede3d2b0098f33705216a00dcf8bf47c3ffa-25-11-2019.png) (https://www.thinkbroadband.com/broadband/monitoring/quality/share/3346ede3d2b0098f33705216a00dcf8bf47c3ffa-25-11-2019)
-
Just noticed i'm unfortunately back on interleaved. Had a burst of only 300 or so ES a few days ago and that night it immediately reapplied. The max 2880 es/day rule is definitely not applying anymore for me. Something else must be being taken into account.
-
I got daily 300 ES and still on fastpatch, so must be something else in your case...
Maybe too much CRC or HEC errors ?
-
Kitz was DLM'd with an extremely low ES count so yes the most plausible explanation is now some extra parameters are been monitored and used for decision making.
-
I wonder if these possible new rules are for ECI cabinets only. Wouldn't surprise me.
ISPs should be given the ability to adjust the interleave setting and db values.
Decent ISPs could then add features to their control panel so end users can adjust with appropriate disclaimers.
I wonder why it takes so little to activate interleaving but then so much to get rid? Very annoying.
-
That's true, I was fighting for fastpatch over two months and so far so good two weeks on fast patch. I had few modem reboots and is still there. :cool:
-
I am the next victim of this new DLM :(
had some ES way below the old limit, question is will I escape its graft without intervention?
-
You need wait and see, is your line on G.inp?
-
You need wait and see, is your line on G.inp?
i wish, its heaviest interleaving ive ever had on vdsl, circa 14ms delay added for 500 one off ES
-
You can always cap your line to make it more stable.
-
You can always cap your line to make it more stable.
Not sure the new DLM pays much attention to that either. I tried relaxing my Plusnet line and it wouldn't get rid of interleaving until I got into the g.INP trial.
-
I have had 2 ES in the past day, its about as stable as it will get.
The line to be honest is typically stable from day to day, it was only a moderate amount of noise that came from nowhere, and the old DLM wouldnt have taken action.
I agree with kitz that this new DLM is too sensitive.
The delay value is 1401, to those in the know I dont know how heavy that is, but the latency is pretty big from 8ms to 24ms. The interleaved I had on the default was much lighter only an addition 8ms instead of 16ms.
-
>> The delay value is 1401,
That is the interleaving depth not delay.
Openreach uses INP [on the ECIs] to control Error Protection. When using INP, the D value is the number of bits in the data block and this can change depending upon your bit load. Thus that figure could change just by performing a resync.
See Impulse Noise Protection (INP) (https://kitz.co.uk/adsl/interleaving.htm#INP)
The total number of bits in a DMT symbol can vary depending upon how many bits are contained in each sub-carrier bin. For example tone 40 may have sufficient SNR to load 14 bits, whilst tone 400 may only contain 5 bits. The number of bits carried in each bin is called a symbol.
Using our above example tone with 14 bits and an INP value of 3, then the system must ensure that sufficient error protection is applied to ensure recovery of 42 bits of contiguous data for that particular tone.
The figures you need to be looking at are these in your full stats.
Bearer 0
INP: 3.00 0.00
INPRein: 0.00 0.00
delay: 8 0
INP = 3 is Interleave low, giving an additional delay of 8ms. The other setting they use is INP=4 Interleave High
Actually I notice from your stats that the D value is now 1409, so your bit load must have changed slightly
-
ok thanks kitz, maybe there is other reasons as well for the larger jump in latency.
-
Since the changes it has been taking 10-14 days to remove interleaving from my ECI line. That is even averaging very low daily ES <10 and usually <5 after the event.
Have never tried capping line as concerned if I had another event I could end up with a perma cap based on my user cap settings that I can't get rid of.
It is very sensitive to adding interleaving and complete opposite to removing it.
Touch wood I am on the longest run of fastpath for quite some months even though I had a ES scare a couple of weeks ago.
No idea why it didn't add interleaving the new DLM criteria is just madness in my opinion. I am praying Ginp does eventually come to ECI cabinets.
-
If the ES spike happened on a bunch of users on the cabinet at the same time, then it may have been detected as a wide area event.
-
Since the changes it has been taking 10-14 days to remove interleaving from my ECI line. That is even averaging very low daily ES <10 and usually <5 after the event.
Have never tried capping line as concerned if I had another event I could end up with a perma cap based on my user cap settings that I can't get rid of.
It is very sensitive to adding interleaving and complete opposite to removing it.
Touch wood I am on the longest run of fastpath for quite some months even though I had a ES scare a couple of weeks ago.
No idea why it didn't add interleaving the new DLM criteria is just madness in my opinion. I am praying Ginp does eventually come to ECI cabinets.
In past I was waiting for over 15 days and nothing didn't changed on my line, cap line was only solution to go back on fast path even that line was very stable with low amount of errors. Few weeks back I changed modems from eci to vigor 2760, next day totally interleaved, so cap line on zyxel, after 3 days fast path back on upstream, next 3 days fast path back on upstream, swapped back to vigor and still on fast patch. Every line perform different, this is what I learned from this forum ;)
-
I have had 2 ES in the past day, its about as stable as it will get.
The line to be honest is typically stable from day to day, it was only a moderate amount of noise that came from nowhere, and the old DLM wouldnt have taken action.
I agree with kitz that this new DLM is too sensitive.
The delay value is 1401, to those in the know I dont know how heavy that is, but the latency is pretty big from 8ms to 24ms. The interleaved I had on the default was much lighter only an addition 8ms instead of 16ms.
Can you post your stats?
xdslcmd info --stats if possible
I've only every seen 8ms or 16ms delay added with Interleaving.
-
@J0hn the link to Chrys's line stats are in his sig, which is how I got these (https://team-rebellion.net/Chris/fullstats.htm) from in my post above and also noticed the D value is now 1409.
Bearer 0
INP: 3.00 0.00
INPRein: 0.00 0.00
delay: 8 0
-
@J0hn the link to Chrys's line stats are in his sig, which is how I got these (https://team-rebellion.net/Chris/fullstats.htm) from in my post above and also noticed the D value is now 1409.
Bearer 0
INP: 3.00 0.00
INPRein: 0.00 0.00
delay: 8 0
Thanks I'm going blind :D
-
Yeah not sure why I thought D was delay, sorry guys, I havent looked at stats for ages so, kitz corrected me on that one.
If I get a 3 day recovery I will expect fast path either tomorrow or more likely friday, the error rate is currently extremely low, 0 errors for the current 17 hours today and 15 in total for the 2 days 17 hours uptime.
But seems more likely on this new DLM, it will be end of next week. :(
j0hn if you interested in what triggered the DLM, my historical stats are available in the sig, check for date 2020-04-06.
After recovery I am tempted to cap the line to try and get at least to 9db DS SNRM.
-
Starting to think the original issue and another issue today is down to someone doing work on cabinet or loop.
Reason been today I had 2 spikes of errors, but also a extended period of downtime with no sync at all, for something to cause complete loss of sync for over an hour suggests work in the cabinet or something.
This type of situation shows how broken DLM can be, a 24/7 interleaved state for error spikes a week apart, and powerless against a "real" problem such as the outage.
-
Graphing has stopped again for my line dslstats silently closed, yesterday I waited till past midnight and manually uploaded the previous day to history, I will do the same today, the reason is if I restart dslstats it wipes the existing graph and then lose the existing days data. So will resume after midnight, and with todays graph intact in history (up to 5pm).
Also after yesterdays outage my latency went down 9ms, I am back at the levels that I had on the default ECI profile. Sync speed I Capped slightly above the 67mbit banding level, aware I am risking a 74mbit banding, but this should prevent a 67mbit banding.
-
Interleaving was removed yesterday. or I should say earlier today.
-
Hi,
I am having High FEC errors and my line is interleaved on downstream, stats below. I am connected to an ECI cabinet and am using A Draytek Vigor 130 modem, on BT infinity FTTC.
---------------------- ATU-R Info (hw: annex A, f/w: annex A/B/C) -----------
Running Mode : 17A State : SHOWTIME
DS Actual Rate : 35000000 bps US Actual Rate : 10666000 bps
DS Attainable Rate : 44606048 bps US Attainable Rate : 10707098 bps
DS Path Mode : Interleave US Path Mode : Fast
DS Interleave Depth : 1514 US Interleave Depth : 1
NE Current Attenuation : 20 dB Cur SNR Margin : 5 dB
DS actual PSD : 6. 4 dB US actual PSD : 13. 0 dB
NE CRC Count : 1 FE CRC Count : 4046
NE ES Count : 1 FE ES Count : 2646
Xdsl Reset Times : 0 Xdsl Link Times : 4
ITU Version[0] : fe004452 ITU Version[1] : 41590000
VDSL Firmware Version : 05-07-06-0D-01-07 [with Vectoring support]
Power Management Mode : DSL_G997_PMS_L0
Test Mode : DISABLE
-------------------------------- ATU-C Info ---------------------------------
Far Current Attenuation : 27 dB Far SNR Margin : 6 dB
CO ITU Version[0] : b5004946 CO ITU Version[1] : 544ed086
DSLAM CHIPSET VENDOR : < IFTN >
[ Counters: Showtime ]
Near End Far End Note
ElapsedTime : 385676 385676 (seconds)
CRC : -1 4046
FEC : 265415461 53814
HEC : 0 0
CRC_P : -1 0
CRCP_P : 0 0
CVP_P : 0 0
NCD : 0 0
LCD : 0 0
ES : 0 2646
SES : 3 31
LOSS : 0 1
UAS : 0 0
LOFS : 0 0
[ Counters: 15Min ]
Near End Far End Note
ElapsedTime : 150 150 (seconds)
CRC : 0 0
FEC : 374657 0
HEC : 0 0
CRC_P : 0 0
CRCP_P : 0 0
CVP_P : 0 0
NCD : 0 0
LCD : 0 0
ES : 0 0
SES : 0 0
LOSS : 0 0
UAS : 0 0
LOFS : 0 0
[ Counters: 1DAY ]
Near End Far End Note
ElapsedTime : 10050 10050 (seconds)
CRC : 0 0
FEC : 24097803 0
HEC : 0 0
CRC_P : 0 0
CRCP_P : 0 0
CVP_P : 0 0
NCD : 0 0
LCD : 0 0
ES : 0 0
SES : 0 0
LOSS : 0 0
UAS : 0 0
LOFS : 0 0
-
Bit of a thread ressurection but, connection is back to interleaving. I'm guessing it's probably because of the 3 peaks in error seconds? Didn't have much more than 500 ES in the day so should be below the thresholds. Getting very tempted to give Virgin Media another go. Just a little unsure as they were terrible around here a few years ago, hopefully they have improved the network with it now being capable of 1Gb in this area.
adsl: ADSL driver and PHY status
Status: Showtime
Last Retrain Reason: 1
Last initialization procedure status: 0
Max: Upstream rate = 13712 Kbps, Downstream rate = 78845 Kbps
Bearer: 0, Upstream rate = 13680 Kbps, Downstream rate = 66078 Kbps
Link Power State: L0
Mode: VDSL2 Annex B
VDSL2 Profile: Profile 17a
TPS-TC: PTM Mode(0x0)
Trellis: U:ON /D:ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 6.2 6.0
Attn(dB): 19.6 0.0
Pwr(dBm): 13.3 6.8
VDSL2 framing
Bearer 0
MSGc: 18 19
B: 51 238
M: 1 1
T: 64 13
R: 12 16
S: 0.0250 0.5560
L: 20448 3669
D: 1293 1
I: 64 255
N: 64 255
Counters
Bearer 0
OHF: 32181228 160228
OHFErr: 10 21
RS: 3943269355 3133355
RSCorr: 42106 354
RSUnCorr: 705 0
Bearer 0
HEC: 156 0
OCD: 0 0
LCD: 0 0
Total Cells: 2283865645 0
Data Cells: 24030966 0
Drop Cells: 0
Bit Errors: 0 0
ES: 7041 6842
SES: 61 4
UAS: 66 53126
AS: 51781
Bearer 0
INP: 3.00 0.00
INPRein: 0.00 0.00
delay: 8 0
PER: 1.60 9.07
OR: 119.34 22.04
AgR: 66197.41 13701.63
Bitswap: 5987/6015 2/2
Total time = 34 days 4 hours 10 min 47 sec
FEC: 42106 354
CRC: 10 21
ES: 7041 6842
SES: 61 4
UAS: 66 53126
LOS: 1 0
LOF: 7 0
LOM: 0 0
Latest 15 minutes time = 10 min 47 sec
FEC: 1055 3
CRC: 0 0
ES: 0 0
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Previous 15 minutes time = 15 min 0 sec
FEC: 1197 1
CRC: 0 0
ES: 0 0
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Latest 1 day time = 4 hours 10 min 47 sec
FEC: 16361 45
CRC: 0 3
ES: 0 3
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Previous 1 day time = 24 hours 0 sec
FEC: 25745 309
CRC: 10 18
ES: 311 3429
SES: 11 2
UAS: 38 26580
LOS: 1 0
LOF: 7 0
LOM: 0 0
Since Link time = 14 hours 22 min 59 sec
FEC: 42106 354
CRC: 10 21
ES: 3 3429
SES: 0 2
UAS: 0 26552
LOS: 0 0
LOF: 0 0
LOM: 0 0
NTR: mipsCntAtNtr=0 ncoCntAtNtr=0
>
-
Kitz has mentioned in a few recent posts that the DLM has changed a bit and it checks over a smaller period than 24 hours now. So some people have triggered DLM without exhausting the daily limit.
-
Interleaving was removed just after 2am this morning. 14 days this time.
-
Gone straight back to interleaved. I can't see anything obviously wrong with the line, although DSLStats is showing a fairly constant 50-60 ES/hour for yesterday. I had a feeling that DLM would stick interleaving back on in the early hours, so pre-emptively placed an order with Virgin. Hopefully the network has improved since I was last with them, although have 14 days to find out. Can always use this one as a backup connection until the contract runs out.
adsl info --stats
adsl: ADSL driver and PHY status
Status: Showtime
Last Retrain Reason: 1
Last initialization procedure status: 0
Max: Upstream rate = 13774 Kbps, Downstream rate = 78772 Kbps
Bearer: 0, Upstream rate = 13803 Kbps, Downstream rate = 66026 Kbps
Link Power State: L0
Mode: VDSL2 Annex B
VDSL2 Profile: Profile 17a
TPS-TC: PTM Mode(0x0)
Trellis: U:ON /D:ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 6.2 6.0
Attn(dB): 19.6 0.0
Pwr(dBm): 13.3 6.6
VDSL2 framing
Bearer 0
MSGc: 18 18
B: 51 238
M: 1 1
T: 64 21
R: 12 16
S: 0.0251 0.5511
L: 20432 3702
D: 1293 1
I: 64 255
N: 64 255
Counters
Bearer 0
OHF: 34940478 683670
OHFErr: 22 35
RS: 354670280 335164
RSCorr: 97574 237
RSUnCorr: 1035 0
Bearer 0
HEC: 188 0
OCD: 0 0
LCD: 0 0
Total Cells: 2847941229 0
Data Cells: 12538367 0
Drop Cells: 0
Bit Errors: 0 0
ES: 8051 14692
SES: 83 8
UAS: 142 106390
AS: 56265
Bearer 0
INP: 3.00 0.00
INPRein: 0.00 0.00
delay: 8 0
PER: 1.61 8.71
OR: 119.25 22.03
AgR: 66145.61 13824.86
Bitswap: 10834/10859 8/8
Total time = 50 days 4 hours 48 min 39 sec
FEC: 1693805 3815
CRC: 6707 351
ES: 8051 14692
SES: 83 8
UAS: 142 106390
LOS: 3 0
LOF: 22 0
LOM: 0 0
Latest 15 minutes time = 3 min 39 sec
FEC: 1722 2
CRC: 2 0
ES: 1 0
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Previous 15 minutes time = 15 min 0 sec
FEC: 2407 2
CRC: 0 0
ES: 0 0
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Latest 1 day time = 4 hours 48 min 39 sec
FEC: 44530 139
CRC: 12 17
ES: 3 15
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Previous 1 day time = 24 hours 0 sec
FEC: 53044 98
CRC: 10 18
ES: 325 3801
SES: 12 2
UAS: 38 26649
LOS: 1 0
LOF: 7 0
LOM: 0 0
Since Link time = 15 hours 37 min 43 sec
FEC: 97574 237
CRC: 22 35
ES: 5 3800
SES: 0 2
UAS: 0 26622
LOS: 0 0
LOF: 0 0
LOM: 0 0
NTR: mipsCntAtNtr=0 ncoCntAtNtr=0
>
-
Can’t be sure the Virgin line won’t get interleaved assuming it’s delivered over OR copper in the same bundle as the existing line. Does the interleave stop you using the line in some way perhaps?
-
Can’t be sure the Virgin line won’t get interleaved assuming it’s delivered over OR copper in the same bundle as the existing line. Does the interleave stop you using the line in some way perhaps?
Virgin have their own network. To be perfectly honest, not really. The slightly higher upload speed (gone for M200) will be useful for my mum when she is working from home. Tbh, if it doesn't meet my expectations, I'll just cancel Virgin within the 14 days but I'm willing to give them a go.
-
Ah so your able to get a cable connection then, that should be much better than FTTC
-
Ah so your able to get a cable connection then, that should be much better than FTTC
That's a whole different can of worms, when cable gets contention it tends be much much worse than DSL. But when it works well, its pretty awesome.
-
That's a whole different can of worms, when cable gets contention it tends be much much worse than DSL. But when it works well, its pretty awesome.
This, pretty much. When it works, it’s great. Last time I was with Virgin, there was upstream congestion, which was painful at peak times.
Interestingly, there’s a few Openreach vans/lorries that have turned up today. Seem to be checking the ducts in the pavement. Not sure what for, I know there’s FTTP planned for here but seemingly anytime up till 2025. Unless it’s for the planned 5G mast at the top of the road? Do Openreach do work for mobile providers?
-
I think if I had a cable service I wouldn’t have left it for FTTC !
-
That's a whole different can of worms, when the cable gets contention it tends to be much much worse than DSL. But when it works well, its pretty awesome.
this is correct, I had Virgin for 18 months in a shared house of 10! circa 2015 and got 106Mbps 24,7 over LAN, so was VERY LUCKY, but I've seen many customers suffer from Virgin Congestion and it's not only bad, but it won't get fixed until almost the whole street of Virgin customers have complained!!! as it's a "shared bandwidth" type of setup where each postcode has its own node which may be 1000Mbps or 10,000Mbps shared between a street, if all started to download at the same time you ain't gonna get your speed. they rely on people downloading at different times so that overall you get good speed - even though my experience was pleasant, I WOULD NOT take out a contract with them from experiences in the past with the sheer amount of customers i worked for complaining of Virgin speeds.
as for Interleaving and Fastpath - I used to get FASTPATH for 9 months getting 80/20 until i had that "MISHAP" from midnight till 4am when i lost 10Mb+ speed and been the same since 3 years ago... I'm getting less than 1 errored second per hour and zero CRC's - yet, FP won't be enabled on my Plusnet connection by DLM....it also had a problem recently where my IP was static when it should of been dynamic, going downhill fast in my book PlusNet
-
Gone straight back to interleaved. I can't see anything obviously wrong with the line, although DSLStats is showing a fairly constant 50-60 ES/hour for yesterday. I had a feeling that DLM would stick interleaving back on in the early hours, so pre-emptively placed an order with Virgin. Hopefully the network has improved since I was last with them, although have 14 days to find out. Can always use this one as a backup connection until the contract runs out.
adsl info --stats
adsl: ADSL driver and PHY status
Status: Showtime
Last Retrain Reason: 1
Last initialization procedure status: 0
Max: Upstream rate = 13774 Kbps, Downstream rate = 78772 Kbps
Bearer: 0, Upstream rate = 13803 Kbps, Downstream rate = 66026 Kbps
Link Power State: L0
Mode: VDSL2 Annex B
VDSL2 Profile: Profile 17a
TPS-TC: PTM Mode(0x0)
Trellis: U:ON /D:ON
Line Status: No Defect
Training Status: Showtime
Down Up
SNR (dB): 6.2 6.0
Attn(dB): 19.6 0.0
Pwr(dBm): 13.3 6.6
VDSL2 framing
Bearer 0
MSGc: 18 18
B: 51 238
M: 1 1
T: 64 21
R: 12 16
S: 0.0251 0.5511
L: 20432 3702
D: 1293 1
I: 64 255
N: 64 255
Counters
Bearer 0
OHF: 34940478 683670
OHFErr: 22 35
RS: 354670280 335164
RSCorr: 97574 237
RSUnCorr: 1035 0
Bearer 0
HEC: 188 0
OCD: 0 0
LCD: 0 0
Total Cells: 2847941229 0
Data Cells: 12538367 0
Drop Cells: 0
Bit Errors: 0 0
ES: 8051 14692
SES: 83 8
UAS: 142 106390
AS: 56265
Bearer 0
INP: 3.00 0.00
INPRein: 0.00 0.00
delay: 8 0
PER: 1.61 8.71
OR: 119.25 22.03
AgR: 66145.61 13824.86
Bitswap: 10834/10859 8/8
Total time = 50 days 4 hours 48 min 39 sec
FEC: 1693805 3815
CRC: 6707 351
ES: 8051 14692
SES: 83 8
UAS: 142 106390
LOS: 3 0
LOF: 22 0
LOM: 0 0
Latest 15 minutes time = 3 min 39 sec
FEC: 1722 2
CRC: 2 0
ES: 1 0
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Previous 15 minutes time = 15 min 0 sec
FEC: 2407 2
CRC: 0 0
ES: 0 0
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Latest 1 day time = 4 hours 48 min 39 sec
FEC: 44530 139
CRC: 12 17
ES: 3 15
SES: 0 0
UAS: 0 0
LOS: 0 0
LOF: 0 0
LOM: 0 0
Previous 1 day time = 24 hours 0 sec
FEC: 53044 98
CRC: 10 18
ES: 325 3801
SES: 12 2
UAS: 38 26649
LOS: 1 0
LOF: 7 0
LOM: 0 0
Since Link time = 15 hours 37 min 43 sec
FEC: 97574 237
CRC: 22 35
ES: 5 3800
SES: 0 2
UAS: 0 26622
LOS: 0 0
LOF: 0 0
LOM: 0 0
NTR: mipsCntAtNtr=0 ncoCntAtNtr=0
>
50-60ES/hour is DLM territory, am surprised you were moved back to fast path in the first place. Ideally a line shouldnt be looping between the two.
Usually to get back to fast path the FEC rate has to be very low which would equate to a much lower ES than 50 an hour.
-
this is correct, I had Virgin for 18 months in a shared house of 10! circa 2015 and got 106Mbps 24,7 over LAN, so was VERY LUCKY, but I've seen many customers suffer from Virgin Congestion and it's not only bad, but it won't get fixed until almost the whole street of Virgin customers have complained!!! as it's a "shared bandwidth" type of setup where each postcode has its own node which may be 1000Mbps or 10,000Mbps shared between a street, if all started to download at the same time you ain't gonna get your speed. they rely on people downloading at different times so that overall you get good speed - even though my experience was pleasant, I WOULD NOT take out a contract with them from experiences in the past with the sheer amount of customers i worked for complaining of Virgin speeds.
as for Interleaving and Fastpath - I used to get FASTPATH for 9 months getting 80/20 until i had that "MISHAP" from midnight till 4am when i lost 10Mb+ speed and been the same since 3 years ago... I'm getting less than 1 errored second per hour and zero CRC's - yet, FP won't be enabled on my Plusnet connection by DLM....it also had a problem recently where my IP was static when it should of been dynamic, going downhill fast in my book PlusNet
I think when I was last on cable the bandwidth ratios were much worse then that. Carl has explained in the past node sizes are not consistent from area to area, due to different companies doing the original cable rollouts. In my city the node sizes are very large, and docsis bandwidth pools were not that large when I had cable, I think was more like a few hundred mbit shared between a dozen or so streets.
I do agree when cable isnt congested it works very well. There is areas which are low on subscribers, and as a result the capacity is over provisioned (for VM standards) and they get great performance, then areas like mine where almost everyone has VM and it was originally built with massive node sizes.
The people I know still on cable in my city do still have congestion, but because the rated speeds are so high now along side the docsis shared pools of bandwidth, it no longer drops as low as a couple of mbit, like when I was on there, friend of mine on the 350mbit service usually gets around 20-100mbit in the evenings depending on how bad it is that given day.
I expect in cityfibre areas it has probably improved and it wouldnt surprise me if VM are telling those customers they completed a network upgrade. :D
-
Even with interleave it looks a pretty tidy line to me. Can’t see how much improvement you would notice on FP unless your a gamer of course.
-
50-60ES/hour is DLM territory, am surprised you were moved back to fast path in the first place. Ideally a line shouldnt be looping between the two.
Usually to get back to fast path the FEC rate has to be very low which would equate to a much lower ES than 50 an hour.
It's only seemed to have changed in the last few weeks. Could normally go months being on fast path. Whether it's down to local noise, I'm not sure. Included a screenshot of the average error rates from DSLStats (unfortunately only got 4 days worth).
-
Sadly it happens, I am now interleaved probably for the foreseeable future when previously I was fast path for several years (except very short spells of interleaving once every 2 years or so).
The DSL tech is vulnerable to noise, as well as crosstalk (when no vectoring), so errors appearing out of nowhere isnt unheard off.
-
Usually to get back to fast path the FEC rate has to be very low which would equate to a much lower ES than 50 an hour.
There's no direct correlation between the number of ES a line receives and the number of FEC the same line may receive.
I had similar ES numbers to Jon21 but over 100 times more FEC.
Some types of interference cause no ES at all but cause 100,000+ FEC per min (Homeplugs do this for example).
-
It could be that you get a lot of FECs but they are all corrected, either by Reed-Solomon CRCs, poss correction helped by interleave, or corrected by L2 retransmission (ie G.INP). If they’re corrected, then they don’t turn into "CRC" count events or ES.
FECs are only potential trouble. I think Kitz has written something about this.
-
Virgin Media kit has been delivered. Interesting that the connections on the cable are now push fit. Cable is also quite a bit thinner than the old coax. Won’t have time to activate it until later tonight.
-
Well the BQM doesn't exactly fill me with confidence, looks horrendous (ignoring the 2 red blocks, which was me). Yet weirdly getting the full speeds that I've gone for. Going off topic though.
(https://www.thinkbroadband.com/broadband/monitoring/quality/share/thumb/c86b5e13b611906f2910fa5b0d6221f434d80c34.png) (https://www.thinkbroadband.com/broadband/monitoring/quality/share/c86b5e13b611906f2910fa5b0d6221f434d80c34)
Edit: Have cancelled under the cooling off period. Looks almost exactly like when I was last with Virgin, congested. Surprisingly easy to do though, no fuss.
-
There's no direct correlation between the number of ES a line receives and the number of FEC the same line may receive.
I had similar ES numbers to Jon21 but over 100 times more FEC.
Some types of interference cause no ES at all but cause 100,000+ FEC per min (Homeplugs do this for example).
You made other posts, that said otherwise?
There would naturally be a correlation, FEC are fixed CRC, the more CRC, the more ES, which means more FEC on an interleaved line.
Now of course you can have situations where you have batches of CRC on low numbers of ES, typical in noise bursts, that would give a high CRC to ES ratio and also high FEC to ES ratio, maybe thats what you mean, but regardless, there seems to be no counter for fixed errored seconds counter, so given it doesnt exist how would DLM determine a means of moving an interleaved line back to fast path? I dont think it just uses CRC/ES as I have seen instances where there is no ES but moderate levels of FEC, then a higher noise margin is applied, the levels of FEC go down, then DLM acts to switch it back to fast path.
On your homeplugs example, are you saying you seen them cause FEC on an interleaved line but when that "same" line is on fast path they generate no CRC? Or rather they cause no CRC on interleaved lines, there is a clear difference between the two. Because I can tell you I have seen homeplugs most definitely cause huge amounts of CRC, and then the line becomes interleaved and they then cause FEC instead.
-
Line has had interleaving put on again. The line, for whatever reason, obviously can't do fastpath at the sync speed it achieves whilst on it, so would it be worth capping the line at roughly the sync rate that I get when interleaved? Sure I've seen this mentioned before. If so, should I do that now or if/when it goes back to fastpath?
-
Yep I think it was j0hn who mentioned that and it seems a reasonable idea to me.
-
Do it now and the line won't resync at the higher speed when Interleaving is removed.
The lower the the sync on your line = the lower the ES the line generates.
I had to call my sync to maintain fastpath. It would always revert to interleaved within a few days of full sync on fastpath. It works very well.
-
Yep I think it was j0hn who mentioned that and it seems a reasonable idea to me.
I got the idea from the sticky in the FTTC section. The credit goes to les-70.
-
Thanks. I'm getting around 66.5/15 sync at the moment, so would the command be: adsl configure --maxDataRate 66000 15000 100000? Or can I set the upstream to 20000? Even though I'm unlikely to get anywhere near that.
-
Set the US to a figure above what you expect it won’t have any effect.
-
I've capped the sync rate, see if it helps once the line returns to fastpath.
-
Still no movement on removing the interleaving. ES are about 0.4/hour, don't think I could realistically get it much lower.
-
Whenever I have had fast path removed by DLM, the ES were extremely low, probably less then 2-3 a day if I remember correctly. It should be viewable in my historical stats in my sig, but I cannot remember specifically when I was last temporarily interleaved.