CRC Errors / Errored Seconrs (frustrated)

[James_D]

I wonder if you guys can help, Since the WBC (ADSL2+) regrade I've been having problems like many of you, the obvious one being that my sync rate is actually less than it was pre regrade 3500 down to 3000 but I know theres more to a connection than just it's speed and its not that that is bothering me.

Whats getting to me is that every 15 minutes I'm getting an average of 10 errored seconds sometimes more, this is presented when playing online games I lag for a few seconds then resync, when I host game servers all my friends that are joined lag for a few seconds then resync.

Ive tried 3 routers with 3 chipsets (TI / Connexant / Broadcom) the best line stats coming from the broadcom but all 3 have the same issue so I know its not the router, I'm plugged into the test socket, have tried 4 filters, 3 RJ11 wires(router to filter) have no bell, a latest version NTE5 and nothing else plugged into the line.

CRC errors are also present on the line 8000 over 3 days when I wasnt here and 2000 in the past 24 hours, CRC errors are always going to be present on a line but its the number thats worrying.

Ok to the point, how many of you monitoring your lines after WBC regrades are experiencing these type of issues? or is it that most of you just have the sync rate / throughput problems?

Downstream
Sync is 3066
SNR is 15.5DB
Attenuation is 49DB

Upstream
Sync is 895
SNR is 7DB
Attenuation is 28DB

I'm on fast path at my request and the first step were taking is to interleave and see what happens but feel this may just partly mask an issue with the line / BT's equipment.

Several phone calls to IDnet and I honestly dont fell well looked after, every time in the past I have been pleased and supprised at the level of service but this time feel somewhat frustrated at being told its something at my end, when I explained the lag issues the person (who I wont name) they couldnt understand why that was a problem.

Any help to maintain my sanity would be much appreciated.

[Sebby]

Your downstream target SNRM has been raised to 15dB, suggesting instability. I suspect the error count is related. Unfortunately, you've done everything you can, I'd say, to ensure it's not your side of things. Although you've got nothing else connected, have you tried the test socket? Also, have you tried a different filter?

Have you been in touch with IDNet? They can test your line.

[Rik]

Not being a gamer, James, I won't be a lot of help. Since going to WBC, my pings have been stable at more or less the same level as on Max. The line, however, has not, and the error count is definitely higher. It almost seems to me that WBC is trying to go too fast for my line as it will frequently re-sync at 08:00-08:30, at higher speeds than it can then sustain.

That said, such problems almost always are 'local', ie noise pickup between you and the exchange. It's a matter of eliminating your own equipment and wiring, collecting data and then, if necessary, getting a BT engineer out in the knowledge that they may raise a charge if they find no fault, or find it on your side of the test socket.

[Glenn]

Have you been in touch with IDNet? They can test your line.

You missed this bit Seb

Several phone calls to IDnet and I honestly dont fell well looked after, every time in the past I have been pleased and supprised at the level of service but this time feel somewhat frustrated at being told its something at my end, when I explained the lag issues the person (who I wont name) they couldnt understand why that was a problem.

[Sebby]

That'll teach me.

[James_D]

II'm plugged into the test socket, have tried 4 filters, 3 RJ11 wires(router to filter) have no bell, a latest version NTE5 and nothing else plugged into the line.

I understand its a bit of a mess of grammar / wall of text so will forgive you ;-)

Interleaving was just activated (30mins not bad!) and they turned down the SNR from 15 to 9DB so im sat at 8.8DB SNR on the down at the moment. I cant explain it but theres a bit more db per tone in places there wasnt before so here's hoping.

Just have a nagging feeling somethings wrong.

[Rik]

 

Keep us posted will you, James?

[Simon_idnet]

Hi James

Now that we have interleaving enabled on your line we'll be monitoring it for a while in the hope and expectation that this will help alleviate the CRC error problem for you.

Regards
Simon

[rireed3]

Interleaving was just activated

James,

My line is a lot shorter, but I think my experience was similar.

The difference was that I never asked for interleaving to be turned off.

I noticed that my target margin had been reduced to 4db.   I was getting 50 CRCs per hour, and thought that the sync was too aggressive.

Then I was re-synced and I was quite surprised to see FEC numbers rapidly piling up, as I had never had interleaving.
I was also surprised that my pings stayed quite low.  The interleaving was very effective, and now I get maybe 2 or 3 'severe' error seconds in several days and about 8 CRCs per hour.  The pings are still low.

Hopefully, this unaccustomed arrangement will work for your line.

Richard

[Rik]

WBC does seem to be 'aggressive' compared to Max in trying to obtain a high sync speed, doesn't it, Richard.

[James_D]

Well that didnt take long  

Within the first hour of interleaved

Counters        Down            Up
SF:             238939          224910
SFErr:          14              0
RS:             212895468               3838768
RSCorr:         11927           0
RSUnCorr:       349             0

HEC:            13              0
OCD:            0               0
LCD:            0               0
Total Cells:    27674208                390476
Data Cells:     1324647         57030
Drop Cells:     0
Bit Errors:     0               0

ES:             7               0
SES:            0               0
UAS:            33              0
AS:             3824

INP:            5.01            2.14
PER:            16.00           17.92
delay:          6.89            7.96
OR:             49.49           8.03

Bitswap:        14              11

Total time = 1hours 4min 34sec
SF  = 238939   CRC = 14
LOS = 0   LOF = 0   ES = 7

Not too bad but 7 Errored seconds on downstream, it's less than it was but still not good. then the bad things happened.

Counters        Down            Up

SF:             629192          592296
SFErr:          120             0
RS:             560610492               1519376
RSCorr:         32891           0
RSUnCorr:       65772           0

HEC:            29              0
OCD:            3               0
LCD:            0               0
Total Cells:    72864558                284389
Data Cells:     30150983                74188
Drop Cells:     0
Bit Errors:     0               0

ES:             18              0
SES:            2               0
UAS:            33              0
AS:             10069

INP:            5.01            2.14
PER:            16.00           17.92
delay:          6.89            7.96
OR:             49.49           8.03

Bitswap:        190             19

Total time = 2hours 48min 39sec
SF  = 629192   CRC = 120
LOS = 2   LOF = 0   ES = 18

Latest 1 day time = 2hours 48min 39sec
SF  = 629192   CRC = 120
LOS = 2   LOF = 0   ES = 18

80 of those CRC errors were in a 15 Minute period, this included 2 losses of signal and 4 errored seconds. the losses of signal are new, thats not happened in over a month, SNR has dropped to 6.9DB but I'm putting that down to the time of evening, I'm a sad panda but at least I get to say I told you so to someone

The below grpahs show the intemmitent nature of this problem.

18:00 -> 19:00


19:00 -> 20:00


FYI the tone graphs dont really have that many gaps in them its just the result of JPG compression and image compression on the forums!

[rireed3]

WBC does seem to be 'aggressive' compared to Max in trying to obtain a high sync speed, doesn't it, Richard.

I can't complain, but it makes me tense 

[rireed3]

Within the first hour of interleaved

I think there is significant noise affecting you.

You should be syncing higher for 49db attenuation.  It's fairly stable, but the low sync and varying noise margin point to interference.

I see you did all the right things for your house wiring -- assume you looked at noise margin and sync rate at every step.  If so, the noise has to be coming from outside the house or possibly the mains.

Edit:   All, is that interleave depth 384 right? -- seems to be a v4 Netgear -- that would also mean significant noise.
Richard

[James_D]

Its a DG834PN so is very similar to a DG834V4 but I have also used a DG834GV3 and a Zoom X4.

I've said all along (2 years) I should get higher sync from 49DB ATT but when mentioned I normally get asked by support "how do you figure that"

SNR on the DG834PN has been a solid 15.5 till yesterday, the support guy requested maximum interleaving depth and a new target margin of 9DB that probably allowed me to maintin a 3000 sync with interleaving rather than falling to 2000-2500.

But yea the problem had never seem'd to be noise, I've got spare noise up the wazzu! when I get these lag spikes theres no reduction in SNR so its not like a spike in interferance. I'll be running router stats tonight for a good few hours to proove this.

If there was a significant ammount of noise on the line I'd expect the bit graph to look worse than that for a 3.6km line but something must be upsetting things.

[Rik]

I wish I could suggest something, but this seems so obscure.

[James_D]

Indeed in over 10 years as a computer engineer I've seen some things but this one escapes me, of course all BT's "line tests" come back as fine (do they ever not?)

My thoughts at the moment are it may be a bit of signal leak on the line, I dont use the home phone all that much so cant say I've noticed any echo or voices on the line, when a corded phone is plugged straight into the line it seems crystal clear though again I've not sat there for hours listening 

One small thing I noticed on the way back from home last night I just remembered is there was a BT engineer working on the cabinet in the middle of town (think we only have a few!) maybe they know of a fault. being in a small town the exchange is in the next town along which isnt ideal again makes me think they may be cramming us in down the line, only hope there would be the FTTC rollout but can't see that happening for a long long time.

Gah I am going to resume pulling my hair out for now, really dont see why I'm the one doing the leg work here when somethings clearly wrong and I've done all I can at my end.

[James_D]

HHMMMMM things are getting worse, Since interleaving was turned on I've had my first OCD, SES and UAS (much the same thing only worse) errors, see below.

Counters        Down            Up

SF:             5187174         4882530
SFErr:          220             1
RS:             326805014               1730653
RSCorr:         303061          543
RSUnCorr:       124734          0

HEC:            44              1
OCD:            5               0
LCD:            0               0
Total Cells:    600714704               351607
Data Cells:     44155886                2320
Drop Cells:     0
Bit Errors:     0               25

ES:             35              1
SES:            4               0
UAS:            33              0
AS:             83004

INP:            5.01            2.14
PER:            16.00           17.92
delay:          6.89            7.96
OR:             49.49           8.03

Bitswap:        9828            114

Total time = 23hours 4min 14sec
SF  = 5187174   CRC = 220
LOS = 2   LOF = 0   ES = 35

Latest 1 day time = 23hours 4min 14sec
SF  = 5187174   CRC = 220
LOS = 2   LOF = 0   ES = 35

[Rik]

I wish I had an answer for you, James.

[James_D]

Ok I think I'm one step closer to answering this / figuring it out and I'm man enough to admit when I've made a mistake, this evening while watching the errored seconds rack up again I phoned support, my handler was busy so he rang me back, as he rang me back 30 mins later he asked me to enable pings on my router, to my supprise my router wasnt responding, no matter flicked over to my back up router, still no response,,, hhmmm this is strange.

Told support guy I'd e-mail when it was all sorted as it was clearly an internal network issue (I have quite a lot of PC's and experimental servers and a complicated topology here) 5 minutes later a reboot of my vista pc and all was well,,,,,, for another 5 mins, it seem'd every 5 minutes my vista pc would fall over to limited connectivity (non wireless connection) which is most odd, renewing the adapter fix's it for another 5 minutes. after 4 hours of uninstalling / reinstalling / updating drivers / checking re-checking settings / changing settings I hooked a monitor onto one of my servers and found it was happily browsing the net and everything was fine so I had to finally accept that my on board 1GB network card has died. wether it's been going for a few days or what I cant tell now but the 1GB Intel Pro at the moment is working!

So I have the conclusion that there is still an issue with my line, the problems I was getting i.e. errored seconds are still an issue between router and the outside world, what we had was 2 issues hence such strange symptoms.

At the time of writing I've been connected for 15 minutes wich is 10 more than I've managed all night! 

[Simon]

Sounds like you're at least half way to sorting your problems.  Do let IDNet know, so that they can concentrate on your line issues.  

[Sebby]

I see what you mean, James. So you think the errors may be local?

[James_D]

I see what you mean, James. So you think the errors may be local?

I very much doubt all the errors are local, Errored seconds / CRC errors are being logged on the router with the exchange & the issue of losing connection for a few seconds is apparent on all the networked PC's, in it's most basic form the router is plugged into the test socket & a 10M cable goes to a 10/100/1000 switch that seperates the "home" network from my "test" network.

The home network plugs directly into that switch and the test network resides another switch the two being connected via Fibre Link (I test virtual machines etc hence the need for such hardware)

It is possible but very minutely that a faulty nic could collide with the traffic heading to the router but it's doubtful, remembering this is just 1 faulty nic / PC and the loss of internet for a few seconds has effected the servers too.

Either way I'll know more tonight after 24 hours of logging I'll have some data to dig through, the 2 loss of signals may have been engineers working in the area, theres no work logged but I've seen an openreach van working on a roadside cabinet a few streets away.

*edit*

Forgot to add before someone ask's yes I did connect the home pc directly into the router to isolate the network but ofc with a faulty nic the problems persisted.