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

[CONDUIT #PNI-592429]: LDM config statement for new GFS 0.25 data

Hi Chris,

re:
> I've verified that my upstream host (vortex.esc.brockport.edu) is not
> receiving CONDUIT data.

OK.

re:
> Apparently we are receiving standard model data
> from another type of feed (UNIDATA?)

It is most likely that you are receiving the NGRID and possibly the
HDS feeds.

re:
> and have been missing out on the
> hi-res stuff all along.  How do I go about setting up a feed for CONDUIT
> data?

The top level relays for CONDUIT data have historically been:

idd.meteo.psu.edu    - Penn State
flood.atmos.uiuc.edu - UIllinois
idd.aos.wisc.edu     - UWisconsin/AOS
idd.tamu.edu         - Texas A&M
idd.unidata.ucar.edu - Unidata Program Center/UCAR

Currently, only Penn State and us appear to be relaying the newly expanded
CONDUIT feed.  The other sites may not be able to relay the full
CONDUIT feed due to the inability to increase their LDM queue sizes,
network limitations, etc.  The same limitations are likely to be
the case with non-relay sites.  We will learn who can and who can
not ingest the full CONDUIT feed on a case-by-case basis.

I suggest waiting a bit before diving into the CONDUIT "deep end".
The first things to investigate on your side are:

- is your network capable of handing all of the data you are
  currently REQUESTing AND the full CONDUIT feed?

- does your LDM machine have enough RAM so that you can increase
  your queue size to be able to handle all of the feeds that
  you have been receiving AND the full CONDUIT feed?

- does your machine have enough disk space to do something
  meaningful with the CONDUIT data?

  FYI: The volume of all of the 0.25 degree GFS data is 20 GB for each
  model run.

- is your machine fast enough to process the CONDUIT data you
  are interested in using?

  I seem to recall that you are a GEMPAK site.  This would mean that
  you would be decoding the CONDUIT data using the GEMPAK 'dcgrib2'
  decoder.  It is our experience that this decoder is very demanding
  of computer resources, hence the caution.

When all of the above questions have been answered in the affirmative,
and when the smoke has cleared wrt relay of the full CONDUIT stream,
we should revisit who you should be REQUESTing CONDUIT from.

Cheers,

Tom
--
****************************************************************************
Unidata User Support                                    UCAR Unidata Program
(303) 497-8642                                                 P.O. Box 3000
address@hidden                                   Boulder, CO 80307
----------------------------------------------------------------------------
Unidata HomePage                       http://www.unidata.ucar.edu
****************************************************************************


Ticket Details
===================
Ticket ID: PNI-592429
Department: Support CONDUIT
Priority: Normal
Status: Closed