[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[IDD #QBK-381226]: access to NEXRAD/Level II radar data (LDM)
- Subject: [IDD #QBK-381226]: access to NEXRAD/Level II radar data (LDM)
- Date: Thu, 20 Jan 2011 15:34:49 -0700
>
> Jeff,
>
> Do you have any update on this?
>
> Thank you!
Hi Karen,
Apologies for the delay, things have been busy :)
Looking at the list of machines you found (on an old, now orphaned page)...none
of them are currently
running an LDM. Our site contact list:
http://www.unidata.ucar.edu/software/idd/sitelist.html
..has only one .mil...and it no longer resolves. Tightened security, as you
know, has impacted the ability for .mils to share info and/or report stats. Our
stats page indicates no .mils feeding, but we know they are out there:
http://www.unidata.ucar.edu/cgi-bin/rtstats/siteindex
Soooooooo, where do we go from here? Unidata's cluster is stable and will
maintain IP addresses as long as possible. Since this data access is part of a
project at RAL, we would prefer for those sites to still get the data from you.
It scales better this way rather than having all projects and their data
requests hit our distribution cluster.
I suspect you would like to have more liberty on what machine(s) you use, and
to be able to do hot swaps on machines etc.., hence the request for a new
topology. Not sure how you have your system is configured, but we would be
happy to help out to show how to maintain IP address while changing out
machines so that the .mils would not need to get new permissions for access. I
may be missing the mark here, please correct me if I have taken this down the
wrong road :) It may be beneficial to have the .mils you serve, also add our IP
block so that in the case of RAL going down, they could fail over to our
cluster until RAL returns. In previous inteactions with .mils, they have, at
times, been hesitant to allow a "block" of IP's, but since we run a 15+ machine
cluster, it is needed to have the block approved so that new connections to
different nodes in the cluster happen seamlessly.
Thoughts?
Jeff
We are open to any other suggestions you might have to smooth this out
Jeff Weber
Unidata User Support
http://www.unidata.ucar.edu
Ticket Details
===================
Ticket ID: QBK-381226
Department: Support IDD
Priority: Emergency
Status: Open