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

20040210: ldm-mcidas upgrade on bergeron (cont.)

>From: "Market, Patrick S." <address@hidden>
>Organization: University of Missouri-Columbia
>Keywords: 200402101807.i1AI7qp1003819

Hi Pat,

re: ldm-mcidas upgrade
>Thank you very much, Tom.  This is a huge help.

No worries.

>I'll tuck my responses in below.

OK.  My responses to questions will follow your questions.

re: looks like the half-hourly images are now being correctly decoded
>>>>
>>>> This all looks good.  Thanks!
>>>>

You are welcome.

re: non-standard naming for certain images
>>>>
>>>> I wasn't aware that the naming was non-standard, but our GARP is set up
>>>> to display those properly.  Should we change it?
>>>>

If things are working for you, then there is no reason to change anything.
I was just concerned that bergeron was doing some decoding work that
was not being used.

re: dcgrib errors in ~ldm/logs/ldmd.log
>>>>
>>>> I confess that I am not aware of them.  Without local support I have
>>>> attended to the system when a data stream disappears.  Because we seem
>>>> to be getting most/all of the HDS that we request, I didn't want to
>>>> look inside to find out.  I hate to share my ignorance with the world,
>>>> but that's the case...  What is the nature of the errors, and what
>>>> can I do to correct them?
>>>>

The errors look like:

Feb 10 06:00:13 bergeron.snr.missouri.edu dcgrib[9174]:  Error writing grid 
data:  -11
Feb 10 06:00:13 bergeron.snr.missouri.edu dcgrib[9174]: 852 skipped 040210/0000 
        Grid ID  240   7 100   0 200
Feb 10 06:00:14 bergeron.snr.missouri.edu dcgrib[9174]:  Error writing grid 
data:  -11
Feb 10 06:00:14 bergeron.snr.missouri.edu dcgrib[9174]: 853 skipped 040210/0000 
        Grid ID  240  52 100   3 182
Feb 10 06:00:14 bergeron.snr.missouri.edu dcgrib[9174]:  Error writing grid 
data:  -11
Feb 10 06:00:14 bergeron.snr.missouri.edu dcgrib[9174]: 854 skipped 040210/0000 
        Grid ID  240  52 100   2 188
Feb 10 06:00:14 bergeron.snr.missouri.edu dcgrib[9174]:  Error writing grid 
data:  -11

The solution is most likely to use dcgrib2, but I am not the GEMPAK
expert here, Steve Chiswell is.

re: old version of McIDAS loaded and trying to run XCD decoding
>>>>
>>>> Can you tell me what file that is in?  Is it in ldmd.conf?  
>>>>

The entries are:

~ldm/etc/ldmd.conf:
exec    "xcd_run MONITOR"

~ldm/etc/pqact.conf:
DDPLUS|IDS    ^.*     PIPE
       xcd_run DDS
HRS   ^.*     PIPE
      xcd_run HRS

>>>>
>>>> I would love to upgrade, but I fear the extra admin that I will need
>>>> to do (and not get done).  I think we should leave McIDAS dormant here
>>>> for now.
>>>>

OK.  Given this I commented out the ldmd.conf and pqact.conf entries for
McIDAS-XCD.  I then stopped and tried to restart your LDM:

<as 'ldm'>
cd ~ldm
ldmadmin stop
ldmadmin start

The 'ldmadmin start' failed with the message that the LDM queue did
not exist, and, indeed, it did not:

I am LDM =>ls /res/met/data/ldm
app        ddp        difax      forecasts  fronts     gempak     grids      
logs       mcidas     nldn       stl        surface    upperair

This was WEIRD since the LDM was running OK when I stopped it!!

Anyway, I remade the queue and restarted the LDM:

I am LDM =>ldmadmin mkqueue
I am LDM =>ldmadmin start
Checking the pqact(1) configuration-file "/usr/local/ldm/etc/pqact.conf"...
/usr/local/ldm/etc/pqact.conf is syntactically correct
Starting the LDM server...

It is now running without the McIDAS-XCD decoding, so you should save
some CPU cycles.

re: you are now rotating the ldm-mcidas log files

>>>>
>>>> Thank you.
>>>>

re: Please let me know if you see anything amiss with the current
UNIWISC image decoding.

>>>>
>>>> Will do.  Thanks again,
>>>> Pat
>>>>

No worries.

Cheers,

Tom
--
NOTE: All email exchanges with Unidata User Support are recorded in the
Unidata inquiry tracking system and then made publically available
through the web.  If you do not want to have your interactions made
available in this way, you must let us know in each email you send to us.

>From address@hidden Thu Feb 12 07:04:15 2004

Tom:

Thanks again.  Bergeron is 1-2 years from retirement (from hard labor
anyway), so any strain that can be taken off is welcome.  I will write
to Steve about dcgrib2.

I, too, have experienced that strangeness with stopping/starting the
ldm.  Of course it's fixed with the outdated
stop-clean-delqueue-mkqueue-start approach, but I thought that was
supposed to be a thing of the past.  No bother.  A few extra keystrokes
can't kill me.

Regards,
Pat

Dr. Patrick S. Market
Dept. of Soil, Environ. & Atmospheric Sci.
University of Missouri-Columbia
Columbia, MO  65211
Voice: 573-882-1496