[FADCA] Some observations RE: Charlie.

[Chuck Hast]

Folks,
Having sat at the Hillsborough, Co EOC thinking I was going to return
to something similar
to what we see down in Charlotte, Co. I went over network issues in my
mind and also some
issues we discovered here in Tampa that had never surfaced before.

Looking at the damage and wind speeds, we really need to think about
how large of a size
the radiators need to be in order to provide the services we want to
provide but at the same
time survive such a event.  Watching the footage from the passage of
the storm through
Charlotte Co was a stark reminder that the same forces that can left a
large aircraft and
allow it to fly also can totally destroy anything that is in the wind
stream of such a storm
I know that small is not  something we associate with the antennas of
our network devices.
But we need to think about both antennas and bands used for both links
and access points.

Also we need to look hard at HF and tropo as alternative routes to get
data out of the
disaster areas, indeed HF is being worked over quite well, and I had a
chance to test it
out using  'Airmail' while at the  EOC (Bud we tried to connect to you
but no joy) our
link was through a station in TX, but regardless it did the job and we
were there and ready
if needed.

Now, on to something that I have beat on every so often and that is
tropo links. For some
reason I can not figure out why, but we seem to avoid this one like a
plague yet I have
found that unplanned for sold tropo paths have in many cases been a
after installation
issue in many data systems I have worked on. If these things without
planning them
appear and work solidly without planning, why can't we take advantage
of them and
make them work for medium haul links out of say targeted sites on a area by area
basis.  As I see it the normal paths will be point to point and will
of course link neighboring
switch sites. I believe if we had enough tropo links that we could use
them to cover
distances that HF may not handle well at higher speeds and with much
higher reliability.

Now the question is how to get started on some testing of such links. 
Obviously many
sites will not be able to support a tropo link, but key sites will.
The idea would be to
space them such that when a even such as the last storm takes place,
we will loose
some of these sites or the tropo piece due to the size of the
antennas, but if the link
antennas are small enough to get the data out of the immediate area,
then it can be
routed to a site that has a functional tropo installation that can
then in one or two hops
get the data on to the final location such as the state EOC or
whatever. That combined
with the Telpac access points should give us a very well covered and
linked network.

We really do need to look at what is needed to harden a site so that
it will live through
a even like the one we just lived through. At least the local access
port, the links to the
neighboring switches and associated Telpac access point. Add to that
enough tropo
and HF links and we would have a very robust network that would be
able to pass data.

Each working link out of a trouble area represents bandwidth, and the
more of them we
can keep alive the better we will be respected in this as a viable
solution for events such
as that we just went through.

Another point that needs to be really looked at in depth is power, our
Tampa site in
the county building down town went off the air  due to a programmed
power outage in
preparation for a 14 foot surge which would have submerged all of
downtown Tampa.
TECO rather than let the underground infrastructure fill with water
and short out pulled
started pulling the plug on the downtown area, the building in
question has power plants,
but the whole of downtown was being evacuated including the people who
would operate
the power plant, so a system that should have had backup was not so.
This we hope
will make you think about other sites that may be 24X7 emergency, but
does not take
into account such issues as the one we ran into which curtailed 24X7 operation.

Yesterday at the EOC we were discussing how to correct this problem,
either moving
the equipment to another site or build a emergency power system for
all the gear up
on that building roof. The locaton and access make it quite attractive
and as we found
out we are not the only ones affected, and I believe that in this case
we have something
we can work together with the county to correct (the county WAN also
went down as
all of the routing and access is based out of that building.

-- 
Chuck Hast 
To paraphrase my flight instructor;
"the only dumb question is the one you DID NOT ask resulting in my going
out and having to identify your bits and pieces in the midst of torn
and twisted metal."