Sorry for the sensational subject, but hey, I had to get some attention
amidst the raging linux/solaris debate on ldm-users, which I'm staying
out of. However, this is very important I think...
Buried deep in a newsletter from our campus communications service office,
I found a reference to a proposed "funding plan" for campus networking
services. It almost looked too boring to bother typing in the URL to learn
more. I'm glad I did though! Actually, not glad at all. What I
found was a proposal to start charging departments for the
internet traffic they use (in addition to monthly charges per network
jack, IP address and other things)
I've been sitting here doing some back of the envelope calculations
and based on our current traffic rates attributable solely to LDM,
it'd cost us an estimated $1100 monthly just in network usage charges
for LDM according to their listed rates. That's over $13K annually.
Since we don't pay a dime for it now, that definitely not something
in our budget.
Of that, 19% is due to incoming data and 81% is due to outgoing
downstream relaying.
This is at current rates of IDD data flow. The number and size of
IDD delivered products however is forecast to grow considerably
over the next year. The NNEXRAD feed for example will more than
double in size.
As you can see, it quickly becomes cost prohibitive for us to relay.
And beyond that it ultimately becomes more cost effective to
just purchase a NOAAPORT receiver and not participate in IDD at all.
In statistics from this past summer, our LDM server was the second
largest traffic source on our entire campus (behind UIarchive, a
massive FTP mirror site) And we're a big campus with lots of
bandwidth hungry people!
For those of you at smaller institutions, you may find that your
IDD servers comprise an even larger percentage of your campus'
traffic.
And don't think those who operate your networks don't know that.
Of course, this is only a proposal and might not (and hopefully won't)
happen at all, but it could. And if it can happen here, it'll likely
start happening elsewhere. The Big10 schools like us, like doing everything
the same, football rivalries aside.
IDD has been very successful so far, largely due to all the "free"
bandwidth available to the participants. (Of course it's not free,
but in most cases, the funds don't come directly from the wallets of
those using the data). If it is no longer free, or reasonably cheap,
then IDD is no longer successful.
Just a though to darken your day. :-)
--
--------------------------------------------------------
David Wojtowicz, Sr. Research Programmer
Department of Atmospheric Sciences Computer Services
University of Illinois at Urbana-Champaign
email: davidw@xxxxxxxx phone: (217)333-8390
--------------------------------------------------------
