Re: [conduit] Missing GFS files on conduit this morning


somebody just asked where the files are when they looked in this directory
http://modelweather.com/noaaport/dish/ncep/model/gfs/

my fault.. I didn't explain very well.. sorry!

since GEMPAK is being phased out, I am writing a custom ingest package for our internal uses, and filing the products as each source sends them over NOAAPort

for example, GFS which was brought up today comes over the dish as:

71550 20180220035934.747355 NGRID 19629065 YRPK40 KWBC 200000 !grib2/ncep/GFS/#003/201802200000F072/RELH/400 hPa PRES

where:
'ncep' is the source
'GFS' is the model
'#003' is the grid

so I chose to file things as
../root/www/dir/noaaport/dish/'source'/'model'/'grid'/

and then instead of a crazy filename.. simply pipe everything into
YYYYMMDD.HHHH.{F}HHH.grib

so all variables for each product are filed into the corresponding date, hour, and forecast hour by model type.

then with NIDS data for example, I file all products by:
../root/www/dir/noaaport/dish/nids/'source'/'radome'/'prod'/
YYYYMMDD.HHMM.nid

so to answer the unspoken person's question, while this is a work in progress for me internally, of course I don't mind sharing with the unidata community if you should need to grab something... unidata is good people Eventually I hope to file every single product that comes over noaaport, because what I'm doing is instead of copying from pqact.gempak, writing everything manually by parsing the notifyme output of every feed.. it just takes awhile.. but this way, products will be able to be used natively with MetPy and other new apps being developed


cheers,

--patrick

-----------------------
Patrick L. Francis
AerisWeather.com
modelweather.com

--

PNG image

PNG image

PNG image