[MITgcm-support] error in PROFILES_INIT_FIXED
Matthew Mazloff
mmazloff at mit.edu
Tue Jun 9 16:44:24 EDT 2009
Could this possibly have anything to do with fiddata never being
initialized? perhaps an issue with fiddata = 0 causing confusion --
just a guess?
On Jun 9, 2009, at 1:31 PM, Matthew Mazloff wrote:
> Hello,
>
> So I am getting the following error:
>
> (PID.TID 0029.0001) Opening num_file 5 dir0029/
> aaiw05_ctddata0029.nc
> (PID.TID 0029.0001) fid, num_file, ProfNo, ProfDepthNo
> 0 5******************
> (PID.TID 0029.0001) fid dimid ProfNo 0 -1 5 0
>
>
> This has nothing to do with file dir0029/aaiw05_ctddata0029.nc (e.g.
> if I remove this file from data.profiles another file will cause the
> error), I think it has to do with the number of processors and them
> tripping over themselves. I was able to get by this in my small
> setup by giving all files there own name -- hence processor 29 is
> opening aaiw05_ctddata0029.nc. On my larger setup having there own
> name (or even directory) doesn't help.
>
> Any ideas as to why the model is tripping over the profile open call
> err = NF_OPEN(fnamedatanc, 0, fiddata(num_file,bi,bj))
> I am running on Ranger
> Any idea of what to try?
>
> -Thanks
> Matt
>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support
More information about the MITgcm-support
mailing list