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

20010123: installation of XCD at Plymouth (cont.)

>From: Jim Koermer <address@hidden>
>Organization: Plymouth State College
>Keywords: 200101080056.f080uCo08251 Unidata McIDAS-XCD installation 
>configuration

Jim,

>LDM is indeed running on cyclone. Here are the pids:
>
>cyclone# ps -aux | grep ldmd
>ldm      270  0.0  0.0 202824  300  ??  Ss   11Dec00   0:32.36 rpc.ldmd
>-q /data/ldm/ldm.pq /home/ldm/ldm-5.0.6/etc/ldmd.conf
>ldm      275  0.0  0.1 203208 1164  ??  I    11Dec00   7:42.78 rpc.ldmd
>-q /data/ldm/ldm.pq /home/ldm/ldm-5.0.6/etc/ldmd.conf
>ldm      276  0.0  0.3 203752 2928  ??  S    11Dec00 436:01.42 rpc.ldmd
>-q /data/ldm/ldm.pq /home/ldm/ldm-5.0.6/etc/ldmd.conf
>ldm      277  0.0  0.0 202800  124  ??  S    11Dec00   0:32.35 rpc.ldmd
>-q /data/ldm/ldm.pq /home/ldm/ldm-5.0.6/etc/ldmd.conf

Hmm...  I guess I got fooled by the output from ps:

<ldm>uname -a
FreeBSD cyclone.plymouth.edu 4.2-STABLE FreeBSD 4.2-STABLE #2: Mon Dec 11 
19:34:01 GMT 2000     address@hidden:/huge/bsd-src/sys/compile/MYKERNEL  i386

<ldm>ps -u ldm
USER   PID %CPU %MEM   VSZ  RSS  TT  STAT STARTED      TIME COMMAND
ldm  57737  0.0  0.1  1392  912  p0  Ss    9:46AM   0:00.09 -csh (csh)
ldm  57831  0.0  0.0   396  188  p0  R+    9:46AM   0:00.00 ps -u ldm

>It is putting the mcidas files in /data/mcidas-c and some of these files
>are copied to /data/mcidas.

OK.  I got caught by the different ps syntax.

>We would like to be able to run mcidas-x pointing to xcd files from
>either cyclone or snow, since they share the same data files via nfs.

OK.  This is easy to setup in McIDAS once the data gets decoded.

>Is there a way to redirect (perhaps without xcd or with a pseudo-xcd
>process) to my NOAAPORT surface, upper-air, and model files as an
>alternative along the lines of the GINI files?

The XCD data monitors read through daily spool files to extract the
products they decode.  The daily spool files are created/maintained by
XCD the ingestors ingetext.k (for textual (e.g., DDPLUS|IDS) products)
and ingebin.k (for binary products like GRIB messages).  The data
decoders (which are subroutines of data monitor routines) simply read
through those spool files and decode whatever their configuration files
have told them to decode.  The trick is that the daily spool files are
composed of the original NOAAPORT stream products; the product
delimiters still must be intact so the data decoders can identify them
in the spool.

XCD was originally designed for the data ingestors to read from serial
ports like those provided by NOAAPORT receivers.  The trick in this
case would be to figure out how to pipe the raw products into the stdin
of the ingetext.k and ingebin.k.  The LDM pqact.conf setup was a
natural for doing just this.  All I had to add to the ingestors was the
ability to read from stdin (in addition to reading from the serial 
port device).

The way that some sites are connecting their NOAAPORT receivers with
GEMPAK/McIDAS decoders is by running an LDM on their NOAAPORT machine
and pqing the raw products as they come in into the LDM queue.  That
LDM then feeds an LDM running on some other machine (it could run
locally on the NOAAPORT machine I guess) where decoding will take
place.

So, the question before us know is how to wire up the data receipt
on your NOAAPORT ingest system so that products are sent to the XCD
ingest processes.  Any ideas on this?  I will touch base with Chiz
tomorrow about how we are doing this on our NOAAPORT ingest systems
and get back to you.

Tom