Print

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

[waltergmw]

Gentlefolk,

I was called back to one of our notoriously lousy 21CN lines recorded as 6588 m long last Tuesday (4th October) as the line had gone down yet again. This was after nearly a month of 50 % intermittent service and with the bRAS on 911 Kbps. After the usual nauseia of button pushing, then waiting as they were so very busy, I eventually found myself speaking to Sierra India Delta. Fortunately, after the usual pleasantries of describing the SSFP, Sid agreed to test my line thoroughly and that he would ring me back within 30 minutes. Imagine my surprise when he actually did so !

Clearly he too observed that the line had lost sync so he told me he would request his supervisor to Reset the SNR after he had raised the fault formally. He told me the line speed might vary for several days before settling down to a higher level and that he would ring me again in a few days. Within about 5 minutes of the line test I found the line had regained sync. Sure enough the fault record could be seen on-line about an hour later showing a possible fault at the "BT Retail Servers". Upon checking the bRAS I discovered it had shot up to an all time record of 1879 Kbps. The next day I checked the fault status to see it had been closed without a word of explanation.

Quite remarkably the bRAS has remained stable at 1879 Kbps ever since. Quite clearly somebody within BT W or BT O have done something quite different from resetting the SNR. Perhaps they have increased the transmit power so I shall do a complete check as soon as I can to record the new 2Wire diagnostic data.

I do wish that BT would:-

a)  Provide a cogent explanation of precisely what had been changed.

b)  Do their own line monitoring and correct observed anomalies without waiting for the end user to complain with all the attendent frustrations.

Kind regards,
Walter

[roseway]

I've always suspected that BTOR often prefer to fix a problem quietly if they can, rather than admit that there was an actual line fault. It's very frustrating for knowledgeable users, but I suppose that the majority of users are just happy to have the fault go away. Providing a cogent explanation would mean admitting to there having been a fault.

Proactive line monitoring and fault correcting is probably expecting too much from such a defensive organisation, but it certainly would make a big improvement in their services.

[waltergmw]

Gentlefolk,

I can now report that the normal situation has returned.

It seems obvious that all BT did was to reset the training period.
Although the speed recently has peaked at 1913 Kbps it has now returned to 1292 Kbps, just a little faster than a longer line of 7414 m in the same cable I monitor at 1184 Kbps.
Both these lines have 2Wire 2700HGV modems.

It is quite preposterous that BT will not train their Asian call centre staff to clearly state precisely what they are doing.
Furthermore there is every chance that the cause of intermittency is still present and is likely to cause further difficulties in the future.

The sooner Ofcom insist that a spade is called a spade the better.
Perhaps then the nation might wake up to the fact that a fibre communications network is urgently needed throughout the country.

Kind regards,
Walter

[sevenlayermuddle]

Good Morning Walter.

I'd just add that, in my opinion, the last thing you want to do with a line that may have a fault is to reset the training period before fixing the fault.  Doing may result in a new (lower) Fault Threshold Rate.  The lower FTR may then enable the call centre to decline to investigate next time the user complains of poor performance, arguing it is 'within acceptable limits', whereas it may not have been within limits of the earlier FTR.  Cynically I could suggest that's why they do it, but I suspect a simple lack of understanding may be more likely.

That said, I believe BTw refer to the retraining requests as 'Customer Controlled SNR' (CCSNR) resets, which is a pity as I think that wording conveys the wrong impression.  I haven't dealt with BT's Indians, but if they are anything like Demon's Indians then they are not real techies, and simply won't actually have a clue what the SNR Margin does, let alone understand the distinction between CCSNR and old-style manual SNR reset.

[waltergmw]

Thanks 7LM,

I had overlooked the FTR even though I'm well aware of that trick. However you can't argue with the "experts" can you ?

Perversely the sync rate stayed higher so I suspect the MTR ? FTR is actually higher.

I'm now deviating on to another problem on a very short line probably of 200 or 300 m. All modems sync perfectly usually at top rate but the throughput and ping vary widely on the THSE exchange which is not recorded as having any capacity problems. The 2700 HGV picture is included below.

Does anybody know if short lines require special attenuation in the exchange and if so how that should be done ?

However what infuriates me is that BT won't tell you what problem they are observing. When on site the BT speedtest always fails with a message to call your service provider. When you do the standard on line bt.com line test from a remote site, it tells you that they are unable to test the line and please phone 0800800151. When you have taken 45 minutes to India explaining the situation and tell them you're not interested in seeing the modem's performance stats as you know they're perfect. Then you are told you must attend site so that tests can be done (no doubt via the test socket) BEFORE they will raise a fault and wait 4 days to get the site visit.

The REALLY preposterous thing is that if the service was via BT wholesale to an ISP such as Zen or AAISP, using almost all of the same equipment, they would listen, understand and take action immediately.

Yet again why should "the proud guardians of the local loop" invent such ridiculous rules instead of using common sense as their wholesale competitors do ? Why don't they see that some customers are actually able to help solve problems ? Why don't they give the customer an idea of the fault they are obviously observing ?

Kind regards,
Walter

[attachment deleted by admin]

[jeffbb]

Hi
quote:Does anybody know if short lines require special attenuation in the exchange and if so how that should be done ?
As far as I know the only thing that is required is to lower the TX power ,which seems to have been done .One of the side effects of very short lines is that the SNR margin goes up because the SNR is so good. . You will also notice the beginning of the bit loading is like an inverted u ,

quote from Kitz:Also note the smooth curves at on the upstream causing a inverted 'U' and also at the beginning of the downstream. This is likely due to spectral masks applied at the DSLAM which ensures that power levels are cut back to reduce the likely hood of cross-talk on those frequencies.

Read more: http://www.kitz.co.uk/adsl/adsl_technology.htm#ixzz1axUzV8RQ
 
Regards Jeff