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

[netCDFPerl #DPK-277567]: netcdf-perl and netCDF 4

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: [netCDFPerl #DPK-277567]: netcdf-perl and netCDF 4
  • Date: Fri, 02 Mar 2012 09:06:12 -0700
Andrew,

> We have installed the netcdf-perl module and netCDF-4 on one of our RH
> Linux machines.  Your web site claims that this perl module will work with
> netCDF-3, but I could not find any statement on expected compatibility with
> netCDF-4.

My understanding is that the netCDF-4 is intended to be backward-compatible 
with netCDF-3, so the NetCDF-Perl module should work.

> In testing, we have found that when we  open, read, then close  a netCDF
> file, and then do that again-and-again through a LARGE loop, eventually we
> run out of positive NetCDF IDs (ncids), and they become negative.  Even
> though each file is closed after reading, and there is at most one file
> open at a time, the software does not seem to re-use NetCDF IDs.  Here is
> what's interesting:  the returned NetCDF IDs become negative after we have
> read-in 32767 files, which is the max value for a signed  (16-bit) short
> integer.

Interesting! I did not know that.

The NetCDF-Perl module simply returns the value from the ncopen(3) function 
(which returns an "int") as an "int" -- so this "rollover" must be due to the 
netCDF package.

> So, having explained this behavior, does (or should) this perl module work
> with netCDF-4?

It should.

> Could we have an installation problem or other issue on our end?

I doubt it.

> Are negative netCDF IDs really indicative of a problem?

I don't know. I've forwarded this issue to the netCDF team. Hopefully, you'll 
hear from them soon.

> From other applications, we have confidence that all of our files are
> well-formed netCDF files containing good data.
> 
> We have never seen this behavior using this perl module along with netCDF-3.
> 
> Thank you for your help,
> 
> -Andy
> 
> ---
> address@hidden
> NOAA/OAR/ESRL/GSD/ACE  R/GSD5
> 325 Broadway  Boulder, CO  80305-3328
> phone:303-497-6211   fax:303-497-6301

Regards,
Steve Emmerson

Ticket Details
===================
Ticket ID: DPK-277567
Department: Support netCDF
Priority: Normal
Status: Closed