If I understand correctly, I can use either the global decoder entry OR the
per type entries. Or is it possible to be using the global for default
decoding/filing and the per type for the few I want decoded/filed
seperately?
Anyway, right now, I still have the global decoder uncommented in my
pqact.conf, and then I uncommented/edited a similar entry to the one below.
So far, no data has made it to the place specified in the new ETA model
entry. And I'm seeing some errors. My pqact.conf entry:
HRS ^[YZ].[BRU].*/mETA
PIPE decoders/dcgrib2 -d data/gempak/logs/dcgrib_ETA.log
-e GEMTBL=@GEMTBL@
data/gempak/model/www/YYYYMMDDHH_eta@@@.gem
and I'm getting errors in my dcgrib_ETA.log
[15019] 031119/1600 [DC 3] Version 5.6.l.1
[15019] 031119/1600 [NA -1] cntrgrib1.tbl
[15019] 031119/1600 [DCGRIB -51] can't open grib tables [NA -51], set GEMTBL
[15019] 031119/1600 [DC 5]
[15019] 031119/1600 [DC 2] Number of bulletins read and processed: 0
[15019] 031119/1600 [DC 6]
Also, in my ldmd.log:
Nov 19 21:42:17 thunder pqact[13788]: pipe_prodput: trying again
Nov 19 21:42:17 thunder pqact[13788]: pbuf_flush (17) write: Broken pipe
Nov 19 21:42:17 thunder pqact[13788]: pipe_dbufput:
decoders/dcgrib2-ddata/gempak/logs/dcgrib_ETA.log-eGEMTBL=@GEMTBL@data/gempa
k/model/www/YYYYMMDDHH_eta@@@.gem write error
So something is awry with my setup. BTW, the /www directory has 755
permissions, so ldm should be able to write in it, and it IS writing in the
dcgrib_ETA.log, so the write error can't be either of those.
Patrick
Message -----
Cc: <ldm-users@xxxxxxxxxxxxxxxx>
Sent: Wednesday, November 19, 2003 11:36 AM
>
> Patrick,
>
> You do not need to use the global decoding pqact entry. I also provide in
the
> decoders example, pqact entries on a per type basis. Also, if you provide
a file
> name, the gribkey.tbl is not needed. For example, to decode
> the ETA grids, I provide in the pqact.gempak_decoders file:
>
> # ETA model output
> # -------------------
> # NOAAport ETA grids
> # Grid #211 80km CONUS: ^[YZ].Q.*/mETA
> # Grid #212 40km CONUS: ^[YZ].R.*/mETA
> # Grid #214 47.5km Alaska: ^[YZ].T.*/mETA
> # Grid #215 20km CONUS: ^[YZ].U.*/mETA
> # Grid #216 45km Alaska: ^[YZ].V.*/mETA
> # Grid #217 22.5km Alaska: ^[YZ].Y.*/mETA
> # Grid #218 12km CONUS: ^[YZ].B.*/mETA
> # Grid #242 11.25km Alaska:^[YZ].S.*/mETA
> # Select any/all grids desired from [QRTUVYBS]
>
> HDS ^[YZ].[QRTUVYBS].*/mETA
> PIPE decoders/dcgrib2 -d data/gempak/logs/dcgrib2_ETA.log
> -e GEMTBL=@GEMTBL@
> data/gempak/model/YYYYMMDDHH_eta@@@.gem
>
> In the above, select which grids you want, and modify the list [QRTUVYBS]
> accordingly. The @@@ in the output file name will be replaced by the
> grid number found in the data. You can modify the path on the output file
> to anywhere independent of the gribkey.tbl entry.
>
>
> Steve Chiswell
>
>
> On Wed, 19 Nov 2003, Patrick O'Reilly wrote:
>
> >
> > Hi again,
> > We at UNI were having mail server problems yesterday, and seem to be
lingering
> > today, therefore none of your responses are getting through to me, I
have to
> > read them from ldm-users archive. Anyway, Ben Cotton's reply about the
two
> > file entries for one product in pqact.conf is adequate for data that you
> > simply file, but for the global grid data decoder I have:
>
> >
> > # Global GRIB decoding action. Products stored maintained in
> > $GEMTBL/grid/gribkey.tbl
> > #
> > HDS|NMC2 ^([HOYZ]|/afs)
> > PIPE decoders/dcgrib2 -v 1 -d data/gempak/logs/dcgrib.log
> > -e GEMTBL=/home/gempak/GEMPAK5.6/gempak/tables
> >
> > where the table in GEMPAK tells the ldm where to file and what to name
the
> > data if I'm not mistaken. How can I tell the ldm in a similar entry to
file
> > only certain types of data in another directory? I can figure out the
> > feedtype/reg exp in the first line, but the rest is not clear. Do I
need to
> > pipe the subset of grids into a script of some sort? Thanks again for
your
> > suggestions.
> >
> > Patrick
> > _______________________________________
> > Patrick O'Reilly
> > Meteorological Decision Support Scientist
> > The STORM Project - University of Northern Iowa
> > patrick.oreilly@xxxxxxx ~ ph: 319-273-3789
> > http://www.uni.edu/storm
> >
> > "No trees were killed in the making of this e-mail...however,
> > a large number of electrons were horribly inconvenienced."
> >
