I wonder if states may get involved. Most (All?) states have a weights and measures office that checks fuel pumps, scales, etc. This feels like an area that may get more regulation in the near future.
Certainly, and it should.
However in this case, the meter itself wasn’t broken, as the title implies. The wrong meter was being read. This happens all the time in the utility world (electric, water, gas), and usually has a quick resolution without additional state involvement.
I don’t necessarily agree. The costs involved would be prohibitive and not much would be gained versus “we assume a 1% error in favor of the customer”
How would you like to see a 3rd party and/or government agency measure counting octets transferred over broadband networks?
One thing the law could provide is an escalation pathway. The subject of this article would have had no recourse but an expensive lawsuit, if the media hadn’t chosen to get involved - that’s just wrong.
tl;dr: Comcast incorrectly recorded the MAC address of the customer’s modem, and was charging this person for someone else’s usage. Also learned these measurements are happening at the CMTS, which I assume uses the router’s SNMP byte counters.
To check that I understand: So this is CMTS, right? And are you suggesting these numbers are self-reported by the router, thus open to manipulation by the customer? (To say nothing of changing the router’s MAC to that of another customer.)
A CMTS is an ISP router and looks like what you linked. The upstream connections face the carrier’s transport network (e.g. 10 gig Ethernet) and the downstream faces cable modems (via DOCSIS usually as Hybrid Fiber Coax / HFC). If Comcast’s CMTSes are similar to what I used to work on, there are standard interface statistics like bytes and packets sent and received and errored packets, broken out by DOCSIS client (CM). Its not manipulable by customers and represents the ground truth of that link. You can’t change the modem’s MAC, and indeed that is the primary subscriber identifier used in DOCSIS.
Thanks for breaking that down and tossing in those nicely searchable keywords. :)
I don’t know. I’m not taking sides, but simply not being home isn’t a really great defense. It’s not like you have to be sitting there watching bit-torrent run. And same thing with malware and viruses. And it’s pretty normal to connect via ssh, etc.
And “disconnecting it from his home network” is only convincing if he he means the modem was completely powered off, which isn’t mentioned.
That’s fair. The article does mention that the company finally discovered a typo in their record of his MAC address, so presumably he was right. But it doesn’t go into enough detail to answer whether he was sufficiently careful at measurement. I read it thinking that perhaps the modem doesn’t have built-in wifi, so that disconnecting it from the router was possible, and was what he meant.
“disconnecting it from his home network” is only convincing if he he means the modem was completely powered off, which isn’t mentioned.
That actually is mentioned in the pastebin linked from the original article. Comcast registered 50 GB of traffic while the modem itself was physically unplugged.
He measured it at his router, too.