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

20030520: met-62 upgrade to LDM-6.0.11; ldm-mcidas decoding of UNIWISC images



>From: "Robert J. Ballentine" <address@hidden>
>Organization: SUNY Oswego
>Keywords: 200305202137.h4KLb1Ld017781

Hi Bob,

>Ideally, I would like to have everything on met-62.oswego.edu, but
>what confuses me is the ldm-mcidas support for decoding the area files.

OK.

>As I mentioned, we are not able to ingest GOES-12 images.  We do not
>have ldm-mcidas on met-62, but you installed it on met-05 running
>Red Hat 5.2.  I would like to eliminiate met-05 as a server, install
>Red Hat 7.3 or higher on it, and make it another workstation available
>to students in the lab.  However, we would then lose all of our
>satellite data.
>
>   Would it be possible to install area decode capability on met-62?

Yes, absolutely.

>I assume this would require installing ldm-mcidas.  If you have time
>to do that, then the satellite files could go in
>                  /var/data/ldm/gempak/images/sat
>on met-62.

OK, this is the same directory as ~ldm/data/gempak/images/sat, so it
matches the setup on met-05 nicely.

>It would be nice to have everything on one machine.  We
>have been using the ldm-mcidas that you installed on met-05 because
>I never learned how to set up ldm-mcidas/area file decoding on our
>newer machine (met-62).

I logged onto met-62 using ssh, not telnet.  You should always use
ssh and never telnet, by the way, since passwords to a telnet login
go over the net as clear text.  Passwords in an ssh transaction are
encrypted.

Here is what I have done so far:

1) installed LDM-6.0.11 from a binary distribution that I had to
   create on one of our RedHat 7.3 Linux systems.  I wanted to
   build the LDM from source, but your development environment
   is missing 'yacc'.  It would be nice if you installed the
   rest of a development environment on met-62 so things like
   the LDM could be built on there from source.

2) modified all of the ldmd.conf files in your ~ldm/etc directory.
   The changes I made were:

   - add the reporting of rtstats in all of them

   - add an allow line for atm.geo.nsf.gov in all of them.  atm
     is the top level IDD node for the East Coast now.  It is an amply
     equipped machine housed in the ATM offices of NSF in Arlington,
     VA, and it running the Unidata 4-port NOAAPORT ingestion system,
     and is totally under our control.

   - made a backup copy of the ldmd.conf file you had setup to
     request IDD products from snow.nrcc.cornell.edu.  Cornell is
     moving to a pay-per-byte pricing for internet use, so they have
     asked us to find other feed hosts for folks that have been feeding
     from them.  We are contacting sites that have used snow for data
     and asking them to move to atm.geo.nsf.gov.

   - I created a new ldmd.conf file for atm.geo.nsf.gov in which
     all of your feed requests except NLDN is made from it.  This
     is the machine that met-62 is now ingesting data from and
     everything appears to be working smoothly.

   - added a request for all UNIWISC (Unidata-Wisconsin aka MCIDAS)
     imagery in all ldmd.conf files.  You are now receiving those
     images on met-62.

   - after met-62 started reporting real time statistics, I was able
     to see that its clock was off.  Since I did not see xntpd running,
     I added the running of /usr/sbin/ntpdate to 'root's crontab.  The
     time server I setup met-62 to use in setting its clock is housed
     at Columbia University.

   - added entries to ~ldm/etc/pqact.conf for decoding of all of the
     UNIWISC (Unidata-Wisconsin aka MCIDAS) images.  I am waiting
     for images to come in to see if the entries are working as
     you indicated above.

     Images are being stored in directories under ~ldm/data/gempak/images/sat
     and the logging of the ldm-mcidas decoder can be found in 
     ~ldm/data/gempak/logs/pnga2area.log.

     The current setup matched the one on met-05, but I tidied up and
     corrected the pqact.conf entries on met-62.

Image ingestion/decoding is looking good.  So far, you have received
one Cloud Top Pressure (CTP) product, one OZONE product, one Manually
Digitized Radar (MDR) product, one GOES-10 IR image, one GOES-10 VIS
image, and one GOES-10 WV image.  These were all decoded with
apparently no errors by the ldm-mcidas PNG compressed AREA file
decoder, pnga2area.  Since the GOES-West images come in first, and
since you had been having problems decoding GOES-East (GOES-12) images,
I will stay logged on until I see some GOES-East images (until about
half past the hour).

Now, as to the scouring of the images being written, I seem to recall
that you had met-05 setup to do the scouring from the 'root' login,
not from 'ldm'.  Given this, I will leave the setup of scouring on
met-62 up to you.

Cheers,

Tom

>From address@hidden Wed May 21 17:29:35 2003

Tom:
   Thank-you very much for installing LDM-6.0.11 and LDM-McIDAS
on our server met-62.oswego.edu and for tuning ldmd.conf.  I was
really worried about getting this done correctly since I have
limited experience with system administration.  I am glad to get
everything (especially the satellite data) on one machine with a
relatively robust version of Red Hat linux.  As far as I can tell,
everything seems to be running smoothly now.
   Thanks again
   Bob