[MITgcm-support] strange problems: "No. of procs= 1 not equal to nPx*nPy= 4"

李志远 oceanlizy at 163.com
Wed Oct 8 18:28:17 EDT 2014


From: Li zhi <oceanlizy at 163.com>
To: MITgcm-support at mitgcm.org
Subject : strange problems: "No. of procs=     1 not equal to nPx*nPy=     4"




hello,all;
    I have met a strange problem recently : I have been using MIT for half a year without any problems ,but just a few days ago , when I use it , it occured such errors:
/opt/gridview/pbs/dispatcher/mom_priv/jobs/6009.node200.SC: line 1: !/bin/bash: No such file or directory
EEBOOT_MINIMAL: No. of procs=     1 not equal to nPx*nPy=     4
EEDIE: earlier error in multi-proc/thread setting
PROGRAM MAIN: ends with fatal Error
ABNORMAL END: PROGRAM MAIN


It puzzled me so much ,because  I changed nothing ,but it did not work well any more !!!!!    From the error ,it seemed that I did not compile with MPI ,but the fact is , I set the right path in /public/userspace/allenson3/MITgcm/tools/build_options ,  here is the MPI path in my linux_amd64_pgf90+mpi_xd1 file,and I add /dcfs2/program/mpi/mvapich2/1.8/pgcc.pgf90/bin/ to my PATH variable .
FC='/dcfs2/program/mpi/mvapich2/1.8/pgcc.pgf90/bin/mpif90'
CC='/dcfs2/program/mpi/mvapich2/1.8/pgcc.pgf90/bin/mpicc'
LINK='/dcfs2/program/mpi/mvapich2/1.8/pgcc.pgf90/bin/mpif90'


MPI='true'


DEFINES='-DWORDLENGTH=4'
CPP='/usr/bin/cpp -P -traditional'
EXTENDED_SRC_FLAG='-Mextend'
INCLUDEDIRS='/dcfs2/program/mpi/mvapich2/1.8/pgcc.pgf90/include'
INCLUDES='-I/dcfs2/program/mpi/mvapich2/1.8/pgcc.pgf90/include'
LIBS='-L/dcfs2/program/mpi/mvapich2/1.8/pgcc.pgf90/lib'


I compile with :../../../tools/genmake2 -mpi -mods=../code -of=../../../tools/build_options/linux_amd64_pgf90+mpi_xd1 ,  then make depend ,make ,it produced the mitgcmuv file ,so I think there is nothing wrong
with my MPI setting .but once I run it , it occured the error as above .  I do not think I am a new mitgcm user ,because I have used the model and run some results I think is good . Just as I say , it suddenly did not work ....
I made a test in the lab_sea directory . below is tha code files :  
SIZE.h file :
     &           sNx =  10,
     &           sNy =   8,
     &           OLx =   4,
     &           OLy =   4,
     &           nSx =   1,
     &           nSy =   1,
     &           nPx =   2,
     &           nPy =   2,
     &           Nx  = sNx*nSx*nPx,
     &           Ny  = sNy*nSy*nPy,
     &           Nr  =  23)


I guess there maybe something wrong in other place , such as the eedata.mth file as below ??
 # Example "eedata" file  for multi-threaded test
#   (copy "eedata.mth" to "eedata" to use it)
# Lines beginning "#" are comments
# nTx - No. threads per process in X
# nTy - No. threads per process in Y
 &EEPARMS
 nTx=1,
 nTy=2,
 &
# Note: Some systems use & as the namelist terminator (as shown here).
#       Other systems use a / character.


did some one met the same problem before ?


thank you for your quick reply !


LI ZHI 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20141009/8512868b/attachment.htm>


More information about the MITgcm-support mailing list