[MITgcm-support] Can we create NetCDF-enabled binaries... no
Christopher L. Wolfe
clwolfe at ucsd.edu
Tue Mar 17 13:32:46 EDT 2009
Jody,
I've gotten by with adding a FLUSH statement to the end of EEDIE. Then
you it least get some output if the system dies "on purpose."
Otherwise, you're left with adding FLUSH statements after every WRITE
statement in the code, which is tedious at best. Seems like there
ought to be a #define statement which enables automatic output
flushing, but I can't figure out what it is. Surely one of the MITgcm
gurus knows some trick ...
Christopher
-----------------------------------------------------------
Dr. Christopher L. Wolfe 858-534-4560
Physical Oceanography Research Division OAR 357
Scripps Institution of Oceanography, UCSD clwolfe at ucsd.edu
-----------------------------------------------------------
On Mar 17, 2009, at 8:03 AM, Klymak Jody wrote:
> Hi Martin et al,
>
> On Mar 17, 2009, at 7:52 AM, Martin Losch wrote:
>
>> such as the end of the output, just before it crashes.
>
>
>
> OK, I'm sure I'm being a doofus when I compile, but when my jobs crash
> using MPI the STDOUT.000* files are clearly not flushed so the end of
> the output is not available. Is there a compiler flag or some other
> flag to force stdout to flush or at least do so with smaller block
> sizes? I can usually tell what I've done wrong in the code if I can
> tell where it crashes.
>
> Thanks, Jody
>
>
> --
> Jody Klymak
> http://web.uvic.ca/~jklymak/
>
>
>
>
>
>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support
More information about the MITgcm-support
mailing list