[colug-432] talking to your ISP

Rick Hornsby richardjhornsby at gmail.com
Sun Dec 30 11:17:07 EST 2012


Worked with Bill and we got a script going that can handle both of our different modems.  If anyone else would find it useful, it is now a self-contained ruby script.

http://ktzr.net/websvn/pub/filedetails.php?repname=KTZR+Public+SVN&path=%2Fcable_modem%2Fget_modem_stats.rb

Sorry the ruby is kind of sloppy. It is the first real thing I've written in Ruby.  It ain't your grandma's Perl.

If anyone wants me to add their modem to the script, just send me (off-list) the URL where the SNR, tx/rx power figures are at on your modem, and a sample file of the HTML source.

I've been able to get some interesting data so far: http://ktzr.net/files/modem_rx_graph.png

-rick



On Dec 28, 2012, at 11:20 , yanovich.1 at osu.edu wrote:

> I think the internal IP address may be different depending on the model
> of modem they issue you. I have had two different models, and the
> 192.168.0.1 as worked for both. I don't have the make/models available
> at this moment to do a comparison.
> 
> The power level for the downstream looks good. I'm not sure what the
> ideal range is for the upstream power level is though.
> 
> From a snapshot I have from a while ago when I was taking screenshots,
> one of my old modems was a WebSTAR DPC2100R2 and reported the following
> numbers:
> 
> downstream Power Level -5.6 dBmV
> upstream Power Level 55.0 dBmV
> 
> At the time I was having problems with the modem not connecting to TWC's
> hub and providing access. They eventually came out and rewired the
> connection. The contractor said it's best to put or split the wire that
> goes to the modem closest to the main hookup in the house, and to avoid
> using coaxial connectors along the way to the modem. This makes sense in
> eliminating more faults and reducing possible loss of signal.
> 
> On 12/28/2012 12:13 PM, davelist at mac.com wrote:
>> 
>> For TimeWarner, this is the link that usually works for me.
>> 
>> http://192.168.100.1/signal.html
>> 
>> As a data point, for me this shows:
>> 
>> downstream
>> S/N 36 dB
>> Power Level 3 dBmV
>> 
>> upstream
>> Power Level 41 dBmV
>> 
>> Dave
>> 
>> 
>> 
>> On Dec 28, 2012, at 10:10 AM, yanovich.1 at osu.edu wrote:
>> 
>>> Another metric you may want to look at is the Power Level being reported
>>> by your cable modem from TWC.
>>> 
>>> In the past I've been able to access it by using this address when
>>> behind the cable modem: http://192.168.0.1/
>>> 
>>> Look for a status page that contains some of this information:
>>> 
>>> Downstream Status
>>> Channel ID
>>> Downstream Frequency
>>> Modulation
>>> Bit Rate
>>> Power Level
>>> Signal to Noise Ratio
>>> ----------
>>> Upstream Status
>>> Channel ID
>>> Upstream Frequency
>>> Modulation
>>> Bit Rate
>>> Power Level
>>> 
>>> 
>>> Specifically, from my experience with TWC and according to the numerous
>>> techs I've talked to when dealing with crappy speeds the number to look
>>> for is the Downstream Power Level. This number is in units of dBmV and
>>> should be between -7.0 and +7.0, if it is outside of this range CALL
>>> THEM UP. As then you can put the blame on them for the crappy speeds, as
>>> this means (from how I understand what the techs explained to me) is
>>> that the signal from their nearest hub to your house is not reaching
>>> your house with enough power and is degrading the signal. If the number
>>> is within -10.0 and +10.0 they will probably try brushing this off as
>>> being nothing, but if it is outside +14.0 and -14.0 then start
>>> complaining ever more, as that is entirely their fault for not having an
>>> adequate line down your street. Either way, I would recommend keeping a
>>> log of this Downstream Power Level number.
>>> 
>>> At first, they will probably send somebody out to your house to look at
>>> the quality of lines to your house and inside your house. They'll
>>> probably replace a few coaxial cables, replace the hookup out on the
>>> street, or they may even install a signal booster. Then they'll probably
>>> do some tests with the device they carry around that plugs directly into
>>> the coaxial cable and ask you to use it for a few days and call back in
>>> if there are any more problems.
>>> 
>>> If the problems persist, they might need to boost the signal down your
>>> street, which in turn will fix your neighbours problems too, assuming
>>> they even notice. I've never had them do this, but lots of people
>>> running off of the hub you are running off of will need to complain
>>> before they'll probably do this, as I imagine this is expensive and
>>> takes more time from them than just sending out a contractor to replace
>>> a few wires outside/in your house.
>>> 
>>> Good luck!
>>> 
>>> On 12/28/2012 01:00 AM, Rick Hornsby wrote:
>>>> 
>>>> On Dec 27, 2012, at 23:53 , Dan <dcarruth2 at columbus.rr.com> wrote:
>>>> 
>>>>> Did you stop to think that the slow response may be from a slow or busy server? 
>>>> 
>>>> I did.  I said as much, right here:
>>>> 
>>>> "I know there are a lot of factors that go into a download speed, and that download speed itself isn't everything.  I know that some servers can be overloaded, and that some links can get saturated."
>>>> 
>>>> That's why I took numbers from multiple servers in multiple cities, and multiple "speed test" websites.  That's also why I targeted servers locally.  Not just because they were local, but because unlike Chicago, New York, etc a server in Kansas City is probably not on very many national radars for a speed test.  That 3 of 4 unrelated KC servers are all "busy" seems a little odd.
>>>> _______________________________________________
>>>> colug-432 mailing list
>>>> colug-432 at colug.net
>>>> http://lists.colug.net/mailman/listinfo/colug-432
>>>> 
>>>> 
>>> 
>>> -- 
>>> Michael Yanovich
>>> 
>>> 
>>> _______________________________________________
>>> colug-432 mailing list
>>> colug-432 at colug.net
>>> http://lists.colug.net/mailman/listinfo/colug-432
>> 
>> _______________________________________________
>> colug-432 mailing list
>> colug-432 at colug.net
>> http://lists.colug.net/mailman/listinfo/colug-432
>> 
>> 
> 
> -- 
> Michael Yanovich
> 
> _______________________________________________
> colug-432 mailing list
> colug-432 at colug.net
> http://lists.colug.net/mailman/listinfo/colug-432




More information about the colug-432 mailing list