Thanks for the advice. The problem turned out to be a memory leak in my
application. Once I fixed that, the error stopped.
Robert Palumbo
palumr@xxxxxxxxx
On Jul 14, 2012 4:45 PM, "Dave Allured" <dave.allured@xxxxxxxx> wrote:
> Rob,
>
> 1. Try renaming your variable "stat". In examples in the Netcdf
> documentation, they use the var name "status". There may be a
> conflict between the var name "stat" and a standard C library function
> by the same name, from stat.h.
>
> 2. IIRC, zero is *never* a legal returned value for ncid, the Netcdf
> file ID number. If you see zero, stop right there and find out what
> else is wrong.
>
> 3. Ensure that ncid and status are both properly declared as type int
> in your program.
>
> Just guessing here, but these are a few things I noticed.
>
> --Dave
>
> On Wed, Jul 11, 2012 at 7:29 PM, Robert Palumbo <palumr@xxxxxxxxx> wrote:
> > Hey Everyone,
> >
> > I have an application that is writing consecutive NetCDF files
> > sequentially (i.e. it's a single-threaded process). I am using NetCDF
> > 4.1.3 on a Windows 7 OS with MSVC 10. Each NetCDF file is written
> > using the following code:
> >
> > // Create NetCDF File
> > stat = nc_create(filename, NC_NOCLOBBER, &ncid);
> > check_err(stat,__LINE__,__FILE__);
> >
> > // Define data dimenssions
> > stat=nc_def_dim(ncid,"Radial",alg_par->num_beam[0],&radial_dim);
> > check_err(stat,__LINE__,__FILE__);
> >
> > ...
> >
> > >From there, I continue to define my data dimensions and input data
> > before closing via 'nc_close()'. The process is then repeated a number
> > of times. I implement check_err() via the suggested format:
> >
> > void check_err(const int stat, const int line, const char *file)
> > {
> > if(stat!=NC_NOERR) {
> > fprintf(stderr,"line %d of %s:
> %s\n",line,file,nc_strerror(stat));
> > fflush(stderr);
> > exit(-1);
> > }
> > return;
> > }
> >
> > I am getting an intermittent NC_EBADID error from the first
> > nc_def_dim() call after the file is created. It always seems to happen
> > when the NetCDF ID ('ncid' above) is set to zero (ncid == 0).
> >
> > Has anyone out there seen this? If the nc_create() call returns with
> > NC_NOERR set, I'm confused as to why the first 'nc_def_dim' call fails
> > with an error code.
> >
> > Any suggestions/help is greatly appreciated,
> >
> > Rob
> >
> > --
> > Robert A Palumbo
> > Graduate Student
> > University of Massachusetts - Amherst
> > palumr@xxxxxxxxx
> >
> > _______________________________________________
> > netcdfgroup mailing list
> > netcdfgroup@xxxxxxxxxxxxxxxx
> > For list information or to unsubscribe, visit:
> http://www.unidata.ucar.edu/mailing_lists/
>
