sunnuntai 27. kesäkuuta 2010

Evidence based decision making

Between Saturday 26 June and Sunday 27 June I carried out the following test: I moved the cable modem to the cellar and plugged it directly in to Eltrona's coaxial splitter, with a piece of coax with plugs fixed by Eltrona at the time when the first installation was made. (The WiFI function of the cable modem is useless in the cellar, that' why I wanted it to be installed in my office on the 1st floor.)

The logic behind this exercise of evidence based decision making was as follows: If there are no interruptions while the modem is in the cellar, the problem is either my 75 ohm RG-6U cable or the connectors I had screwed to the two ends of the cable. If, however, interruptions still appeared while the modem was in the cellar, then changing the cable and fittings would be of no use.

After the installation, the signal levels looked even better than upstairs in a stable situation. The downstream signal was amplified by 5 dBmV, and 42 dBmV was the sufficient signal level to communicate upstream. On the other hand, in the stable functioning of the connection, the +/-0 dBmVa and 46 dBmV seem to be sufficient. Signal to noise ration did not improve from 34/35 dB.

After checking the logs in the next morning, the same old T3's were there: From 1.30 to 1.40 Luxembourgish time, the service was down again.

* Saturday - Sunday 26/27 June 2010

The conclusion is that changing the cable and connectors will not solve the problems. Either the disturbances come from the cable modem (which I doubt) or they come from Eltrona's side of the network.

lauantai 26. kesäkuuta 2010

On 25 June, a friendly employee from Eltrona called me. He confirmed that Eltrona has investigated into all components of their network and that they can see the service breaks, but that the breaks are not caused by their installations.

What can I say? I believe them, they should know. Hence the problem must be in my house, either the modem, the connectors or the cable -there are no other possibilities. Eltrona even suggested a on-site visit, which I had to postpone until September, the end of a busy period at work and some holidays.

To do something, I replaced to connectors at the two ends of the cable, to assure a tighter connection. Immediately after the change, the transmission signal levels were perfect, but then the service went down again. And the breaks continued all evening. In the morning of 26, everything was perfect again, at least for some time.

Next week, I plan to go to Eltrona shop and buy a new "double blindage" coax to replace the current one, bought from Auchan satellite receiver department a couple of years age. That MUST solve all the problems!

torstai 24. kesäkuuta 2010

Offline, again

Today again, I sent an e-mail to the helpdesk, reporting yesterday's service breaks. Previously, they phoned me back, not today. I have not received an e-mail response for months, telling that they have (1) received my problem statement, (2) telling the follow-up number for the "call" and (3) promising to do something - this does not seem to be part of Luxembourg Online's business culture.

What seems to lack in this holy union of Luxembourg Online and Eltrona is any sort of transparency. If it is so difficult to provide a reasonable service between 18-22 on working days, then just tell it!

As the log files of today tell, the service was down again around 17.30 local time. Who knows what follows this evening... And as so often, the connection fell again while I was writing this report!
Signal to Noise Ratio 22.5 dB
Power Level -13.3 dBmV
Frequency 0 Hz Down


  • Thursday 24 June 2010
  • keskiviikko 23. kesäkuuta 2010

    Luxembourg Online - offline records

    And so it started...


    It is particularly annoying to have a failed internet service for the following reasons:
  • Bank transactions are interrupted before the transaction is properly finished.

  • VoIP conversations are cut by the failing ISP - correspondents think that I am living in a banana republic with smoke signals communications dominating the national data comunications market.

  • Service interruptions from 12 June 2010 onwards (if only the service availability allows me to update the site)


  • Saturday 12 June 2010

  • Sunday 13 June 2010

  • Monday 14 June 2010

  • Luxembourg Online and the cable service Eltrona reacted quickly and sent a technician for the morning of 15 June 2010.
    Measurements and installation of an amplifier took 2 hours of the technician's working - and the client's free time. As a result, measurements showed acceptable signal levels: upstream 46 dBmV and downstream -3 dBmV and my connection worked. However, I rarely had problems in the morning...
    The next day it started again: a service failure while paying my holiday trip. The upstream connection was lost. Downstream signal was -12.6 dBmV.
    Why is the signal level changing so much? Is it the poor cabling I have inhouse (which is in a fairly constant temperature) or Eltrona's infallible cable outside? Or is it because the Motorola SGB900E is scanning different frequencies and at a busier time, the frequency found is more damped out than the one found during the off-peak period? Not as far as I can see: These two shapshots show that the signal level changes very quickly - and that the frequency for both channels is the same.

    Reporting continues after the amplifier installation


  • Tuesday 15 June 2010

  • Wednesday 16 June 2010

  • Thursday 17 June 2010 seems to be a good day: No T3's or T4's, as far as I can see!

  • Friday 18 June 2010 and Saturday 19 June morning: the connection works since a couple of days as it should: snapshot signal levels are optimal (DS 46, US 1 and SNR 34), just one T3

  • Saturday 18 June in the afternoon

  • In the morning, everything was perfect. However, at 2 pm, the signal levels were much worse, even if there were no time-outs (yet): Downstream signal was damped by 9.8 dBmV and the modem was screaming at 55 dBmV upstream.
    In the morning, it seems that I had the whole bandwidth at my disposal a speed test showed >7 megabits downstream speed, but in the afternoon, I had to share the capacity with another user.
    Does this mean that my connection works properly only if I am alone on the line? Not only the speed is eaten by other users, but also the connection becomes instable.
    Despite some moments when the downstream signal level was down to -12.9 dBmV, SNR close to 30 dB and the modem was pushing upstream at 55 dBmV, there have been no service breaks during the week-end. Still, the fluctuating signal levels are a concern...
  • Monday 21 June 2010

  • During Tuesday 22 June 2010 from 18.30 to 19.00 another service breakdown appeared. Could this be caused by the Luxembourg Online/Eltrona staff celebrating the Eve of the National Day - and ignoring to send some smoke signals, or another essential part of the hidden communication network between the client and the Internet?
  • Tuesday 22 June 2010

  • The party at Luxembourg Online/Eltrona seems ho have resumed: Several interruptions of the service have taken place during the National Day (see the link below). At 20.20, the signal levels are poor and the connection was lost. Happily I was not caught in the middle of a phone conversation this time. The simple conclusion is: This Luxembourg Online cable modem connection is far too often offline and it is very unreliable! The service breaks take place during the hours when it is mostly used (18-21 in the evening) and the helpdesk is also "offline".
  • Wednesday 23 June 2010