It's also interesting to note that, at least by some estimates,
the brief power outage in L.A. yesterday took down more networks
than Hurrucane Katrina:
http://www.techweb.com/showArticle.jhtml?articleID=170702966
Of course, So. California is pretty "network-dense", but what does
that say about the level of seriousness that network operators place
I think there is a difference as to network going down for 3 hours
and network going down for 3 months...
BTW - care to speculate what will happen if cat5 hurricane hits LA? 
Or maybe we should be thinking of 8+ earthquake ....
folx,
It's also interesting to note that, at least by some estimates,
the brief power outage in L.A. yesterday took down more networks
than Hurrucane Katrina:http://www.techweb.com/showArticle.jhtml?articleID=170702966
fyi, yes, during the power outages in Los Angeles, at their peak,
there were 301 outages (highly localised partitionings with the
Internet on one side and the rest of the world on the other :-),
according to our peerset at renesys. other views may vary, but
probably not by much.
[insert aimless debate about the meaning of 'outage' here.]
that's a significant, and visible outage, but it's not outrageous.
tranmission to and through the region doesn't appear to have blipped
at all, as the majority of power redundancy worked.
Of course, So. California is pretty "network-dense", but what does
that say about the level of seriousness that network operators place
on their "uptime"?
i don't know. a *big* chunk of the visible failures were caused by a
very small number of facilities with supposedly redundant power where
generators and UPSes failed. people who are in those faciliites are,
no doubt, working with building management to obtain RFOs, request
SLA credits, and consider breaking leases in the extreme cases. i
wish the affected parties luck with those efforts.
most stuff just stayed up.
t.
BTW - care to speculate what will happen if cat5 hurricane hits LA?
Or maybe we should be thinking of 8+ earthquake ....--
William Leibzon
threat models for huricanes are different that earthquakes.
(or is that one of those "disaster+geography" equations?)
--bill
And reported Oct 2004:
<http://www3.nationalgeographic.com/ngm/0410/feature5/?fs=www7.nationalgeographic.com>
-Hank
threat models for huricanes are different that earthquakes.
(or is that one of those "disaster+geography" equations?)
For one thing, if you use natural-gas powered gensets,
you are virtually guaranteed to go off-line after an
earthquake. For another, after an earthquake you will
have to be prepared to run without utilities for many weeks
unlike most hurricanes. New Orleans is not an exception
since the hurricane is not what caused the flooding.
Another difference is that underground cabling is protected
from hurrican damage but is not protected from earthquake
damage.
I wonder if there are any resources that cover planning
for different disaster types. It is pretty clear that
telecom companies need to do their own planning, not rely
on government agencies.
--Michael Dillon