[ncl-talk] OPEnDAP issues with large files (reading all zeros)

[Tyle, Kevin R]

Hi Bill,

I noticed this as well a month or so ago, and emailed ncl-talk about it. 

I suspect, based on other reports, that this may be tied to a problem with the netcdf-c library which is causing timeouts when reading large datasets via opendap.  I can replicate similar misbehavior with Python tools.

A workaround is to subset the request so as to limit the amount of data that gets returned by the OpenDAP server.

--Kevin

_____________________________________________
Kevin Tyle, M.S., Manager of Departmental Computing
Dept. of Atmospheric & Environmental Sciences   
University at Albany
Earth Science 235, 1400 Washington Avenue                        
Albany, NY 12222
Email: ktyle at albany.edu
Phone: 518-442-4578                             
_____________________________________________


-----Original Message-----
From: ncl-talk [mailto:ncl-talk-bounces at ucar.edu] On Behalf Of Capehart, William J
Sent: Friday, October 26, 2018 1:28 PM
To: ncl-talk at ucar.edu
Subject: [ncl-talk] OPEnDAP issues with large files (reading all zeros)

Hi All.

We just noticed an issue in reading large GRIB and NetCDF data from unidata’s Jetstream TDS server in NCL.  

We also copied the file to our local file system, and our local TDS server.

For example: 

http://thredds-jetstream.unidata.ucar.edu/thredds/dodsC/grib/NCEP/GEFS/Global_1p0deg_Ensemble/members/GEFS_Global_1p0deg_Ensemble_20181025_1200.grib2

for variables such as "Geopotential_height_isobaric_ens"

Things work well in other software such as ncdump and Panoply but when working with NCL, it touches the file, pulls the metadata as well as the coordinate variables for any given variable.  But all values are read in as zero.

Is anyone else seeing this?

Thanks
Bill


_______________________________________________
ncl-talk mailing list
ncl-talk at ucar.edu
List instructions, subscriber options, unsubscribe:
http://mailman.ucar.edu/mailman/listinfo/ncl-talk