What is included in the traffic count?
In connection to SL's data caps, they had Netforecast conduct a review of their meters. In Netforecast's document detailing their independent assessment of the meters, they indicate that SL's count will include...
"All subscriber traffic to and from the Internet (known as payload), including Suddenlink
management and voice over IP service traffic, will be counted, along with protocol
overhead required to transfer the payload over the cable modem access line." (Emphasis is mine)
Now, I also subscribe to SL's home phone service. Does this mean THAT traffic is included as well?
I emailed tech support and they said, "no," since it comes under a different IP address. But then what is the line above referencing?
VoIP covers any telephone service that travels over your internet line, such as Vonage or Magic Jack.
Short of unhooking your network/computer from the modem for a whole day and making a few hours of phone calls, I'm not sure how you could test if SL phone service is being charged or not. I'll try and get ahold of Pete to see if we can clarify that a bit.
|reply to dhchenault |
I got ahold of Pete and he is checking into clarifying what exactly is going on there, if SL phone service counts or not. I'll post back again when he gets back to me!
|reply to dhchenault |
Pete just got back on this question.
If you are using the phone and internet combo modem, then the SL phone traffic WILL be counted. He passed on that ~1500 minutes of phone talk time would equal ~1 gig of data (or 300 minutes ~ 200 megs). The metering system does this because it watches the traffic at the modem level (I assume the meter is MAC based after this)
It only makes sense to monitor at the modem level, they want to count everything that comes down your pipe.
Should that usage count include unrequested broadcast packets, pings, port probes, etc? Or should it only include traffic initiated on the client side of the modem? Ever notice how much your router's data activity light blinks even when your computers are all off?
If you've got ICMP/ping response enabled (which is default for a lot of routers), anyone can ping through your modem with 64K packets that will faithfully be re-transmited back to requestor, again and again and again. It would be an interesting experiment just to see how much total traffic this could count against a modem's usage limit in a single day. And all off that traffic can be generated outside the modem without the knowledge or consent of the subscriber, yet they have to pay the excess usage penalty?
Personally I'm not against caps or pay-per-usage, but it seems to me a little more thought into what counts against the cap is necessary for it to be fair and accurate.
If the cell phone company sends you a text message, it does not count against your total messages. Calling 611 to talk to your cell phone provider does not count against your minutes. There is certainly precedent for excluding certain traffic from counting toward higher usage charges.
A raw count of bits to and/or from the modem might be a good indicator of heavy users, but using that raw number for billing thresholds seems a bit unfair.
Should is subjective.
The report that SL had done and published (link on the faw site) shows that basically the only thing that doesn't count is Docsis overhead. That said, if you were getting flooded, your router should by default block it after a short while. Even if you don't have any flood control, for a "ping attack" to register on the meter to any significant amount of data being used, you should be aware of this as your connection would suffer.
Comparing SL meter to the phone company meter is sort of apples and oranges. Cell phones have been metering, well, since they came out. SL is just attempting to start doing this. Also, you did used to be charged for all service related contact, text or phone ins. Cell meters and service have both evolved over time, a large portion of plans don't even have meters or limits on voice and sms service beyond customer interest.
Geot, I was just reporting back what I found out. The initial question and answer provided by a SL rep turns out to be wrong.
50/3 Suddenlink : Current
5/1 CMA : Old
15/2 TWC : Old
So what this tells me is use something other than Suddenlink phone. Not only does it count against my data (counting against it is still counting against it, no matter how little) but it costs me about $20 a month for phone service I can get elsewhere since I not only have the pay the phone bill fee, but also have to rent the modem to actually use their service. That easily comes out to around $240 a year if not more for a service I can be paying 1/10th the price for.
As an aside, this also has the added effect of making their meter inaccurate compared to other meters due to it counting traffic that other meters cannot measure, so it really is a "believe us, we know what's best" situation. This actually throws a bucket of water on their "independent" study saying their meters are accurate since there is no way they were able to measure both Suddenlink's internet and their phone service in their testing without having some kind of traffic meter being hooked up before the traffic reaches the modem, and I doubt they even bothered to test the meter in regards to the phone service anyway. Lest we not forget, I already tore apart why the testing was flawed previously since the possible rounding errors in the test could have hidden a discrepancy of easily 10% or more.