[ale] Update on cablemodem problems, considering DSL

[Ken Cochran]

Maybe bringing things some more back on-topic... :)

Y'all are gonna love this...

The cablemodem outage happens hourly, at 10 minutes before the
hour and lasts around 3-20 minutes, at which time IP transport
returns to "normal."  This has been happening since this past
Thursday, roughly midnight.  (Hmm, this sure sounds like a
software change of some kind that went in on 3rd shift...)

Ok, so the *line* tech shows up Friday morning, per my support
request (regular home tech is a waste of time for both me
& them).  *His own* (personal/home) location is experiencing
the same phenomenon.  However I can show him such things as
pings & traceroutes exhibiting the problem, especially since
this failure mode exhibits itself while he's here onsite.
Example: traceroute -n fails right at the 1st hop (which I
see wherever there is cablemodem service here is 10.108.0.1 -
not sure just what that is but I *think* it's the CMTS at the
head-end; I'm fairly certain that it is *not* the cablemodem -
it has a 192.168 address).

I then traceroute "back" to my ip-address from another place &
that fails at the very last hop (mine).  Cablemodem signal:noise
ratios & signal levels are fine, so there's really nothing
for the line tech to do besides report it.

Problem persists throughout the weekend.  Monday I visited the
cableco office personally to get a status update & to report
continuing problems.

Me:  "My Internet still isn't working right."
Lady at the counter: "Losing your Internet 10 minutes before every hour?"

:-O  Indeed...
Turns out another customer had been in earlier reporting the
same thing that the line tech had the previous Friday.  Also,
the gentleman behind me in line was experiencing the same &
thought maybe he needed a new cablemodem.  *And* the local
cableco *manager* is experiencing this.  That's 5 confirmed
outage locations, spread out over different neighborhoods.

As of now (10:15am Central time Tuesday) the problem persists.
This is not the 1st time for this (symptom) sequence; I (we
locally?) encounter this situation roughly 3 times a year.

Any idea(s) as to what it might be?  IMO if it were a hardware
problem somewhere the problem would manifest itself randomly.
These failures being just about clockwork indicates (to me)
that it's a software problem, most likely at the cableco's
local head-end.

-kc