The ABINIT Users Mailing List ( CLOSED )

[Michel Cote <Michel.Cote@umontreal.ca>]

This looks like the problem with the MPI_GROUP. Xavier might have fix this problem in the latest version of the code. Could you try it?

Michel Côté

Le 07-10-11 à 11:58, hyello25@gmail.com a écrit :

Dear abinit users,

I met a weird problem when putting the el-ph calculation of mgb2. The calculation cannot continue when ndtset exceeds 6 or 7 in the job. It  stopped

when reading the WFKs from the 1st and previous nscf results. And the log file stopped at 

-------------------------------------------------------------------------------

 mkfilename : getwfk/=0, take file _WFK from output of DATASET   1.

 mkfilename : getwfq/=0, take file _WFQ from output of DATASET  2. 

 getdim_nloc : deduce lmnmax  =   4, lnmax  =   2,

                      lmnmaxso=   4, lnmaxso=   2.

  distrb2: enter

  mpi_enreg%parareel=           0

  mpi_enreg%paralbd=           1

  mpi_enreg%paral_compil_respfn=           0 

  distrb2: exit

------------------------------------------------------------------------------

and never continue again even after hours. While, the cpus were still running 100% load all the time. 

  Does this problem relate with my compile process?

  My os is Rocks 4.3, and the version of abinit is 5.35. using ifort 9 and OpenMPI 1.2.4

  Besides, does the warning below matter?

------------------------------------------------------------------------------ 

-P-0000  hdr_check: WARNING -

-P-0000   Restart of self-consistent calculation need translated wavefunctions.

-P-0000   Indeed, critical differences between current calculation and

-P-0000   restart file have been detected in: 

-P-0000         * the number, position, or weight of k-points

-P-0000         * the format of wavefunctions (istwfk)

-P-0000 ==============================================================================

Thanks a lot.

_____________________________________

Michel Cote

Departement de physique

Universite de Montreal

emai: Michel.Cote@umontreal.ca

web: http://www.phys.umontreal.ca/~michel_cote