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

RE: [Support #FIR-658958]: RE: 20060221: 20060216: problems with CONDUIT feed from idd.unidata.ucar.edu



Title: RE: [Support #FIR-658958]: RE: 20060221: 20060216: problems with CONDUIT feed from idd.unidata.ucar.edu

FYI..I made the changes (every 6 hours..at least I think I did it right..and split the sequence numbers) on our machine in New Mexico.  It is running LDM v. 6.3.  I tried to build 6.4.4 but I some sort of stupid Linux error (couldn't find some include files during configure) and I had no patience to try to slog through that, so I will leave that alone.  The older Linux box back here (newpsn) is now running ldm-6.4.4 upgraded from 6.0.14.  I will see how it does getting data from New Mexico.

BTW, you should try running (if you haven't) GARP/NMAP2 on a Solaris 10 x64 box with an nVidia card that is running the commercial nVidia Solaris driver.  It is absolutely amazing the difference in performance over the Xorg/Xfree86 drivers under Linux or Solaris.  And unlike under Linux the nVidia driver doesn't make the box unstable (because of Linux issues).   I had 10 triple headed boxes like that at ImpactWeather...and am starting that conversion here.


-----Original Message-----
From: Unidata CONDUIT Support [mailto:address@hidden]
Sent: Wed 2/22/2006 11:10 AM
To: Robert Mullenax
Subject: [Support #FIR-658958]: RE: 20060221: 20060216: problems with CONDUIT feed from idd.unidata.ucar.edu

Robert,

To just request the 0Z and 12Z one degree CONDUIT grids, you can use a request line pattern:
MT.gfs_CY.(00|12).*onedeg

To split the feeds 2 ways, you can do odd/even on the sequence number at the end of the name:
MT.gfs_CY.(00|12).*onedeg.*[02468]$
MT.gfs_CY.(00|12).*onedeg.*[13579]$

Your request below matching on all DF.gr1 will include the 2.5 degree grids from
192 through 384 hours. If you want those (separated out from the onedeg, then add the pattern:
MT.gfs_CY.(00|12).*2p5deg

Also, just to throw in 2 cents here, your LDM version of 6.0.14 is quite old. Is there a reason you are not
updating (I know you are just returning, so it might be on your todo list)? The current distribution is much
more efficient according to Steve.

Steve Chiswell
Unidata User Support

Ticket Details
===================
Ticket ID: FIR-658958
Department: Support CONDUIT
Priority: Normal
Status: Closed