[MOAB-dev] problems reading meshes in MOAB in the 'read_part' mode

[Dmitry Karpeev]

Yes, I think the file is "broken" in the sense you just described.
The reason I mentioned it is that, I think, the way 'read_part' fails
to read this file
indicates a problem with 'read_part'.  Ordinarily, I would expect the
failure to have
something to do with the aforementioned bad partition (that's what
'bcast_delete'
complains about). Instead, 'read_part' fails on this file the same way
it fails on other,
properly partitioned files.

Dmitry.

On Thu, Apr 1, 2010 at 12:21 PM, Jason Kraftcheck <kraftche at cae.wisc.edu> wrote:
> Dmitry Karpeev wrote:
>> I've run several cases of mbparallelcomm_test with 3 different files
>> on my laptop,
>> and I conclude from the results sketched below that the 'read_part'
>> mode is currently
>> broken.
>>
>> For these runs I used MOAB/trunk rev 3733 and three files:
>> 64bricks_1khex_256.h5m
>
> I looked at this file.  It contains a definition for a tag named
> 'PARALLEL_PARTITION', but the tag is not set for any entity.  It does have a
> default and global value.  The latter is rather odd.
>
> - jason
>
>