[MPICH] Behavour if MPI_File_Open fails on some nodes

James S Perrin james.s.perrin at manchester.ac.uk
Mon Aug 20 08:57:55 CDT 2007

Hi Rob,

	Case 2 would be fine, however MPI_File_Open (rank 0) is not returning 
at all in the situation I've described.

	BTW since I don't know if the file is shared or duplicated I'm  only 
using MPI_File_Open, MPI_File_view, MPI_File_read and MPI_File_close. ie 
nothing that should depended on the shared file pointer.


Robert Latham wrote:
> On Mon, Aug 20, 2007 at 11:40:39AM +0100, James S Perrin wrote:
>> If the file does not exist for one of the processes I'd expect either:
>> 1. The specific process to return an ERROR that that file can't be opened
>> 2. All processes to return an ERROR
>> I assumed case 1) and I do an Allgather so that all the processes know 
>> if all the other processes were successful and to proceed accordingly. 
>> This seems to be behaviour I have with other MPI implementations and is 
>> only causing problems if the directory path to the file doesn't exist i.e.
> In addition to Rajeev's comments, we treat errors as in case 2: an
> error from any process should result in MPI_File_open returning the
> apropriate error as part of the return code.   
> ==rob

James S. Perrin,                  | email: james.perrin at manchester.ac.uk
Research Computing Services,      | web:   www.mc.manchester.ac.uk
Kilburn Building, The University, | tel:   +44 161 275 6945
Manchester, England. M13 9PL.     | fax:   +44 161 275 0637
"The test of intellect is the refusal to belabour the obvious"
- Alfred Bester

More information about the mpich-discuss mailing list