I have been battling perplexing and annoying Internet connectivity problems for nearly a month. Occasionally my digital local line drops out, and quite frequently, my Internet access has high latency or comes screeching to a full stop. For those that are into such things, a description of the problem and initial troubleshooting is below the cut.
Today, during a particularly bad patch of connectivity when I was trying to participate in a conference call & webcast, I got annoyed and managed to send a note out via twitter:
Nasty internet connection latency plus repeated digital phone service crashes means yet another call to Comcast. Isn’t telecom fun?
Much to my surprise, a Comcast technician replied to my tweet within 15 minutes with an offer to help.
This could have seemed a “big brother” type response, but the technician made both a non-intrusive open-ended offer to help and her twitter account was set up with obvious care – her name, a picture and a profile that indicated she was a real person.
Through @replies and DMs, she did proceed to help, rather successfully. It appears she provides support through this model all day long and is rather pretty darn effective at it.
By the end of the day, I had a call from a Tier 4 local plant technician who looked at my stats and records and agreed that there was indeed a problem with my Comcast connection. (Darn it, I wasn’t taping…) I will be getting a call from a field supervisor to set up an onsite trouble call early next week.
What Comcast didn’t know is that I work in call center technology for a similar Alaskan company, and am always interested in effective use of alternate support channels.
This is a nice model, a positive application of Twitter and the near-real time capabilities of the Internet, and I’d love to see our company adopt something similar in its technical support center.
Nicely done, Comcast!
Now for the technical details:
My ping times to various resources (Comcast, GCI, etc) were quite often in the 1.0 – 1.3 sec range with 10% to 50% packet loss. A traceroute indicated the problem was in the hop between my wireless router and the local Comcast router.
I’ve had Comcast out twice to work on the issue. The first tech indicated some neighborhood node issues; the second tech found nothing, but I persuaded her to replace my VOIP cable modem.
I also tried replacing my wireless router, thinking that might be the problem. I now have a lovely a dual band Airport Extreme N router. It’s connected by Cat5 cable to my cable modem.
Resetting the cable modem temporarily resolves the problem, but all too toon thereafter it’s back. There’s no pattern in time of day or association with machines on the network.
Today I decided to work direct connected to the cable modem to see if the problem continued in that configuration. (This was after trying and failing to buy tickets from Ticketmaster because of the latency.)
Not only did my access intermittently slow way down, and latency jump way up, but I also saw my cable modem reset itself multiple times over the course of a couple of hours, making land line calls and web conferences impossible.
The Comcast tech I originally spoke with had several excellent suggestions about my local network, but after she looked at the node, escalated the issue to her local plant technicians.
The tier 4 local technician looked at my traceroute data and correlated times of day with me and were convinced that there was indeed a node problem.
I’m looking forward to seeing Comcast onsite early next week.