Verizon FiOS - is BGP an option?
mysidia at gmail.com
Sat Aug 4 16:41:13 UTC 2012
On 8/4/12, Chris Adams <cmadams at hiwaay.net> wrote:
> long-haul fiber all stayed up. Most of the other problems were because
> of the power failure (some local fiber rings dropped, especially one
> CLEC's that puts their nodes in customer premises and were broken by
> customers' power failures).
I would go as far as to say most electric utilities I know of
specialize in safe efficient long-distance transmission for a mass
consumption audience (massive number of users, very few users with
specific high reliability requirements who can't accept a 5 day outage
at least), and favor safety and efficiency over reliability, using
many components that are not buried or heavily shielded, and are
highly susceptible to weather events, trucks knocking poles over,
lightning, tornados, solar flares, etc, and that their answer to
outage prevention is to let it fail, or shut it down in case of
damage, and then repair later.
If a telco provider's answer to powering remote comms facilities is to
just let the electric company bring in AC, to charge a battery which
will last a short time, then disaster survivability was not the
driving design goal for that selection. Possibly because their
customers or their local government weren't demanding it, or weren't
willing to pay enough for them to install suitably designed power
distribution and backup.
If you really care about building a reliable communications
infrastructure, then you have at least two independent paths and
sources for communications, and at least two independent paths and
sources for power, to each major component of the system, so you
provide your own power distribution, favoring reliability.
That increases the price of the service, and if the consumer doesn't
want it so badly that they'll pay significantly more, then it could
be a waste, financially; most of the time,
people won't notice extra fault tolerance measures versus a
service that isn't so resilient in disasters.
More information about the NANOG