Bug#749859: netcdf: Please increase size of netcdf models for exodus

Alastair McKinstry mckinstry at debian.org
Mon Nov 9 06:50:41 UTC 2015


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256

Hi Bas

I don't know how many 'large' models there are.

As the documentation states, the netcdf library does not statically
define anything based on these arbitrary limits, so I believe there is
no impact on netcdf itself. Certainly exodus works fine when netcdf is
changed to meet these limits.
If there is a negative impact, it may be due to certain codes that use
classic netcdf statically allocating in order to meet these potential
limits, but I've never heard or seen such issues.

regards
Alastair



On 08/11/2015 7:52 p.m., Sebastiaan Couwenberg wrote:
> notfound 749859 4.2.1+
> found 749859 netcdf/1:4.3.3.1+ds-1~exp1
> tags 749859 upstream
> thanks
>
> Hi Alastair,
>
> On 30-05-14 11:36, Alastair McKinstry wrote:
>> The netcdf library must be modified slightly in order to handle the
>> exodus datamodel. The specific changes are:
>>
>> 1. include/netcdf.h -- Modify the following defines:
>>
>> #define NC_MAX_DIMS     65536    /* max dimensions per file */
>> #define NC_MAX_VARS     524288   /* max variables per file */
>> #define NC_MAX_VAR_DIMS 8        /* max per variable dimensions */
>>
>> If these changes are not made, the exodus library will still function
>> for "normal" size models, but it will fail for "large" models. The
>> error message from exodus/netcdf should be somewhat clear when this
>> happens...
>
> As you know, we have NetCDF 4.4.0~rc2 in testing/unstable for some time
> now. While that meets the 4.2.1 version requirements, but not the
> NC_MAX_* values:
>
> /**
> Maximum for classic library.
>
> In the classic netCDF model there are maximum values for the number of
> dimensions in the file (\ref NC_MAX_DIMS), the number of global or per
> variable attributes (\ref NC_MAX_ATTRS), the number of variables in
> the file (\ref NC_MAX_VARS), and the length of a name (\ref
> NC_MAX_NAME).
>
> These maximums are enforced by the interface, to facilitate writing
> applications and utilities.  However, nothing is statically allocated
> to these sizes internally.
>
> These maximums are not used for netCDF-4/HDF5 files unless they were
> created with the ::NC_CLASSIC_MODEL flag.
>
> As a rule, NC_MAX_VAR_DIMS <= NC_MAX_DIMS.
> */
> /**@{*/
> #define NC_MAX_DIMS     1024
> #define NC_MAX_ATTRS    8192
> #define NC_MAX_VARS     8192
> #define NC_MAX_NAME     256
> #define NC_MAX_VAR_DIMS 1024 /**< max per variable dimensions */
> /**@}*/
>
>
> I'm not comfortable changing these values, because I cannot oversee the
> impact.
>
> How common are the "large" models? Is it not a very specialised use case?
>
> Kind Regards,
>
> Bas
>

- -- 
Alastair McKinstry, <alastair at sceal.ie>, <mckinstry at debian.org>,
https://diaspora.sceal.ie/u/amckinstry
Misentropy: doubting that the Universe is becoming more disordered.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2

iQIcBAEBCAAGBQJWQEJBAAoJEMvmu05dmtOlF6MP/39dDi4Pb2QoB8aMHct5eTt/
0xwMLSx5k23G+WwtLn2ZBrmQHhpVhJda+cOjGDKhd+eetpWm3gWYjo8Z3Qfi2Axi
MuP1knVRzrulw0pnSyRDeZTjF0XOpdDeCIaglnqjw5dHTWsmskloWy4d87BHccnS
njZxJ6fd/rXiIg42w5eRbi1mNxoW56M+5dHtSIBEG4R43HzqQzuVIhYWoxybw86I
9OqD/C0JN95itmwy1GuaWMvb/0it9cA5ubGX2kfK9mgFTbubztiWSpraoyk+a4C8
tqDjCB3aPZF+hGmwAKjsgBsN3RL6jkwEP1NZjAzsnWc3f/si0TXpRAhDbHw2Ykqx
c+ZHWTzET8y0eC3FEec9ckZYU38T/RpNmoas62pkUMk7YyMlumipAXTar0+R2vfF
qtTkjAS2QRpjlwRC7NE9CCloH/PCTDPPNH2NMAPdFsMFKHtfXfzaRx3bLGbVL+qZ
2ZRzDChgMxS7Yj2ZSCes28QEjP1GvFd/J7ZlgJ9oukKLeBYEWQOSChWszLVgNqvJ
nY+yThn5K5rD5dB+KCOYQ8M7puJp3PY4goUavTz4eAWil5zC1WD2yPuiutXqIlqQ
2VPs/LUP5tJNTcKMWpubI1CghmlcHq67yCllihs9na51ewjWaRIwZqLEoYgvni7m
1/RmcJDxeXQRthGTiNOl
=7hzg
-----END PGP SIGNATURE-----



More information about the Pkg-grass-devel mailing list