Print

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

[MaximusPrime]

@tbailey2,

it would be better if DSL Stats uploaded the info from the disconnection to the reconnection so the circumstances of the disconnection are recorded imo

[roseway]

That would be my job, but it would be quite difficult to implement, and limited time means that I won't be able to do it in the near future.

[MaximusPrime]

@roseway,

is the upload delta 5 minutes? If the upload is missed the 5 minutes data is lost? Could it be reduced to 1 minute if so? I suppose that would introduce performance issues?

[tbailey2]

@tbailey2,

it would be better if DSL Stats uploaded the info from the disconnection to the reconnection so the circumstances of the disconnection are recorded imo

Likely to be very difficult I'm afraid due to the workload of people voluntary involved in the project. You have to bear in mind that all of this system is provided totally free by three different people who have other things to do in life.

In my own case, I estimate that if I had coded and supported this as a commercial venture over the past two years instead of for free, I'd be at least £75,000 better off... The voluntary donations purely go towards recovering actual costs since the project was started - and this is still in deficit.

EDIT:
Sorry missed Eric's post. The delta is one minute currently.

[MaximusPrime]

@tbailey2,

very impressive for 3 people, congrats on that. DSL Stats written in C++ probably across multiple platforms would be a difficult & then hooking it up to the web component would be extra difficult.

So basically in future the enhancement would be to upload the DSL Stats information from when the broadband connection is reestablished to when DSL Stats reestablishes connection to MDWS.

I'll make a donation soon to get the upgraded MDWS. Planning a Raspberry Pi at the moment but it's more complex than I thought to get a good rig set up & more expensive too. I think I'm better sticking with my old Linux box in the meantime.

[MaximusPrime]

@roseway,

thought about the problem, did some pseudo code if you are interested:


vector<Upload> pendingUploads;

while (running) {
   if (currentTime - lastUpdateTime >= delta) {
      Upload newUpload(data, lastUpdateTime, delta);
      
      if (dslConnected && MDWSConnected) {
         for (Upload pendingUpload : pendingUploads) {
            uploadToMDWS(pendingUpload);
         }
         uploadToMDWS(newUpload);
      }
      else if (dslConnected && !MDWSConnected) {
         pendingUploads.append(mewUpload);
      }

      lastUpdateTime = currentTime;
   }
}

[William Grimsley]

Should be:

Code: [Select]

pendingUploads.append(newUpload)

[roseway]

Thank you for the suggestion, but I'm afraid there's a lot more to it than that. I really don't have the necessary time available.

[MaximusPrime]

@roseway,

I was thinking it would be more difficult but was hoping there'd be an easy solution

[MaximusPrime]

I have long suspected the noise on my line at night is caused by street lights which run over the underground telephone cable for 8oo metres

I got a lot of errors last night, attached

There is 1 broken street light. Was the 120,000 FEC's caused by a SHINE burst?

[MaximusPrime]

Hi Guys,

today's MDWS emails tell me my disconnections are due to:

- 8000 Loss of Service
- 1 Remote Defect Indicator/DLM

What's the difference between these?

[NewtronStar]

A street light effecting broadband my dropwire is close to street lamp less than 4 meters away when there was a fault with this lamp it would not come on and when it did it starts up then dims in a few minutes and the cycle continues until the council with there cheery picker arrives to replace the bulb.

During these events it has not caused any issue on my sensitive broadband circuit   

[MaximusPrime]

@NewtronStar,

take a look at my MDWS, past 2 nights there's been a lot of noise

[NEXUS2345]

I would personally say that looking at the last 24 hours, streetlights would have very little to do with it. The last resync was at around 15:00, a time when streetlights should be completely inactive. I cannot comment on what it might be as I am very inexperienced in this matter, but the evidence shown suggests nothing to do with streetlights, at least, not unless you live in a scandinavian country.

[MaximusPrime]

@NewtronStar,

the resync at 15:00 was caused by me installing a new high speed RJ11 cable to reduce noise & attenuation which it did slightly, resulted in higher SNR margin as expected which was good