[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

20030819: LDM-6 upgrade for maelstrom.meteo.mcgill.ca



>From: Unidata User Support <address@hidden>
>Organization: Unidata Program Center/UCAR
>Keywords: 200307021823.h62INMLd020225 LDM-6 upgrade

Hi John (and Alan),

Last month, maelstrom2 was successfully upgraded to LDM-6 and
configured to request data from atm.geo.nsf.gov.  After completing the
upgrade, I asked Alan about upgrading maelstrom to a current LDM-6
release (it is currently running LDM-5.0.6).  Since I see that
maelstrom has not been upgraded, I am revisiting the issue of whether
you intend to replace data IDD data ingestion on maelstrom with that on
maelstrom2.  If not, I would like to encourage you to upgrade the LDM
on maelstrom to LDM-6 as it is much more efficient than LDM-5.

We are willing to do the upgrade for you if given appropriate logins.
Please let us know if you would like us to proceed.

As far as the LDM-6 installation on maelstrom2 is concerned, the real
time statistics that it is reporting back to Unidata show that its
clock is not being set correctly.  The latency plot for the IDS|DDPLUS
data effectively shows the clock drift:

http://www.unidata.ucar.edu/cgi-bin/rtstats/iddstats_nc?IDS|DDPLUS+maelstrom2.meteo.mcgill.ca

For reliable IDD data ingestion, it is important to keep one's clock
set accurately.  As the clock difference between the receiver and
upstream host(s) increases, so does the likelihood that data will not
be delivered correctly (the LDM embodies the notion that data more than
one hour old is not useful, so it is not sent).  As I recall, setting
the clock on maelstrom2 with ntpd or ntpdate has problems due to
firewall issues (?).

Cheers,

Tom Yoksas