Greetings Tom!
I'm not sure which version of netCDF-C the docker image uses, so that might
be an issue (Julien, do you know off the top of your head)?. More likely is
that there is a bug related to netCDF-4 file writing in the netCDF-Java
library, which relies on the netCDF-C library underneath. When you switch
to accept=netcdf, you are then using a pure java netCDF file writer to
produce a netCDF-3 file, and so the problematic code path isn't used.
Dennis - any ideas?
Cheers,
Sean
On Sun, Aug 12, 2018 at 12:00 PM Thomas Cook <tmcook@xxxxxxxx> wrote:
> HI
> I'm having an issue with Netcdf Subset and Netcdf4
> For instance, when I do the following
>
> http://hfrnet-tds.ucsd.edu/thredds/ncss/HFR/USWC/6km/hourly/RTV/HFRADAR_US_West_Coast_6km_Resolution_Hourly_RTV_best.ncd?var=u&var=v&north=35&west=-120&east=-116&south=31&horizStride=1&time=2018-08-12T15%3A00%3A00Z&addLatLon=true&accept=netcdf4
>
> I get
>
> NetCDF: Bad chunk sizes. nc_def_var_chunking on variable u
>
> If I replace "accept=netcdf" in the above URL, I get a valid dataset and
> all is good.
>
> I'm using the thredds-docker Version 4.6.12-SNAPSHOT -
> 2018-01-26T10:29:32+0000 image and not quite sure the best way at
> troubleshooting this.
>
> Thanks
> Tom
>
> _______________________________________________
> NOTE: All exchanges posted to Unidata maintained email lists are
> recorded in the Unidata inquiry tracking system and made publicly
> available through the web. Users who post to any of the lists we
> maintain are reminded to remove any personal information that they
> do not want to be made public.
>
>
> thredds mailing list
> thredds@xxxxxxxxxxxxxxxx
> For list information or to unsubscribe, visit:
> http://www.unidata.ucar.edu/mailing_lists/
